Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
listlengths 1
91
| keywords
listlengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
The early stages of the thermal unfolding of apoflavodoxin have been determined by using atomistic multi microsecond-scale molecular dynamics ( MD ) simulations complemented with a variety of experimental techniques . Results strongly suggest that the intermediate is reached very early in the thermal unfolding process and that it has the properties of an “activated” form of the native state , where thermal fluctuations in the loops break loop-loop contacts . The unrestrained loops gain then kinetic energy corrupting short secondary structure elements without corrupting the core of the protein . The MD-derived ensembles agree with experimental observables and draw a picture of the intermediate state inconsistent with a well-defined structure and characteristic of a typical partially disordered protein . Our results allow us to speculate that proteins with a well packed core connected by long loops might behave as partially disordered proteins under native conditions , or alternatively behave as three state folders . Small details in the sequence , easily tunable by evolution , can yield to one or the other type of proteins .
In addition to the folded and unfolded states , many proteins may adopt stable conformations that display mixed properties of the native and denatured states . These conformations , usually known as intermediates , may appear under unusual external conditions ( i . e . non-physiological pH , pressure or temperature ) , in the presence of high concentrations of certain cosolutes ( denaturants , salts ) , or as a consequence of mutations [1]–[9] , and are supposed to be populated during folding ( and unfolding ) , especially in the case of medium or large proteins [10] . Such folding intermediates may be on-pathway , facilitating the reaction , or off-pathway , acting as traps that may lead to missfolding and even aggregation [11]–[12] . The occurrence of equilibrium intermediates is often associated with stress phenomena and can trigger pathological effects , such as spongiform encephalopathy and other types of amyloidosis [13]–[14] . This explains the existence of many physiological mechanisms designed to reduce their harmful effects , mainly by reducing the life time of these potentially dangerous conformations [15]–[17] . For some proteins , however , physiological roles have been postulated for their intermediate conformations and such possibility might be more common than originally believed [8] , [18]–[21] . All these reasons explain the interest in understanding the nature of intermediates and the atomistic details that favours the transition from native to intermediate structures . Unfortunately , the study of intermediate conformations is much more difficult than that of native forms . Equilibrium intermediates can be detected in vitro as a deviation from two-state behaviour , i . e . , non-coincident protein unfolding curves obtained with different experimental techniques [22] , although their structures and energetic properties are more difficult to probe . Folding intermediates are elusive to X-ray crystallography and their normally small population and the extensive signal broadening compared to the native state difficult their analysis by means of NMR techniques [23]–[28] . As a consequence , structural information on intermediates is often obtained by using low-resolution techniques , often based on Φ-analysis [29] , [30]–[31] , or low-resolution spectroscopic or scattering data , which can give clues on the general shape of the protein , but not atomistic information . This explains the need and frequent use of simulation techniques , particularly molecular dynamics ( MD ) to try to gain atomistic details that are unreachable to experimental techniques [32]–[44] . Flavodoxins are a family of proteins essential for the survival of many human pathogens that has become one of the most studied models for protein folding and unfolding . They are mono-domain α/β proteins , with a parallel five-stranded β-sheet surrounded by five α-helices , and they carry a non-covalently bound FMN group which can be reversibly removed [45] . Several experimental studies on apoflavodoxins from the Anabaena [46]–[47] , Azotobacter and several Desulfovibrio [48]–[50] strains have demonstrated the thermal unfolding of this protein follows a three-state mechanism , where a partly unfolded intermediate accumulates at moderately high temperatures . Using a variety of techniques applied to wild type and mutant proteins Sancho's group arrived to a low resolution picture of the thermal intermediate of the apoflavodoxin from Anabaena PCC 7119 [51]–[52] finding evidences that the intermediate is in fact close to the native structure , with the two hydrophobic cores well preserved , and with distortions probably located mostly in the loops and in one of the β-strands [53] . The overall dimensions of this thermal intermediate were characterized by small-angle X-ray scattering analysis , which suggests that the intermediate is slightly more extended than the native form , but clearly far from the expected situation of a random coil [54] . In this paper , we present a massive molecular dynamics ( MD ) effort for the study of the early stages of thermal unfolding of apoflavodoxin from Anabaena and for the characterization of its thermal intermediate . The study is especially challenging , since the slow folding dynamics of this protein ( average transition times in the order of 101–102 millisecond makes impossible the use of pure force-approaches based on atomistic potentials , which would require second-scale trajectories . Furthermore , many attempts to use coarse-grained potentials failed to sample structures reproducing experimentally known intermediate properties and unfolding pathways , while equilibrium dynamics obtained from coarse-grained potential seems stiffer , but qualitatively similar to that expected from MD simulations ( data not shown but available upon request ) . Accordingly , we decided to use a hybrid approach , based on the use of microsecond scale atomistic MD , supplemented by low- resolution spectroscopic and scattering data and previously derived Φ-analysis . The approach allowed us to characterize with atomistic detail the ensemble of conformations that define the intermediate as experimentally detected in melting experiments . With this synergistic approach the mechanism that drives the transition from native to intermediate and most likely the early stages of the thermal unfolding of the protein were explored .
The crystal structure of Anabaena apoflavodoxin deposited in the Protein Data Bank with reference 1FTG [51] was used as starting conformation for our simulations . Crystallographic SO42− was conserved and simulated together with the protein [in the X-ray structure of Anabaena apoflavodoxin ( crystallized in high ammonium sulphate concentration ) , a sulphate ion is bound , mimicking the FMN phosphate , which opens the possibility that the native conformation in this region is a consequence of the binding of the ion] , the rest of ions 24 Na+ and 6 Cl− which are needed to neutralize large values of electrostatic potential around the protein were added using CMIP calculations implementing Poisson-Boltzman potentials [55] . The resulting systems were then solvated by around 7600 TIP3P water molecules [56] , partially optimized , thermalized to 300 K ( Nose-Hoover thermostat ) and equilibrated using our standard protocol [57] , followed by additional 50 ns of post-equilibration . Ten randomly selected snapshots ( separated by at least 1 ns ) were selected from the last 20 ns of the equilibration trajectory to generate the starting coordinates of ten replicas of the protein in water at T = 300 K . To increase diversity in the ensemble of the native form velocities were randomized and each replica was re-equilibrated for 5 ns prior to 0 . 2 µs isothermic-isobaric production simulations ( T = 300 K , P = 1 atm ) . The structure obtained at the end of the 50 ns equilibration at T = 300 K was heated slowly ( 0 . 5 ns ) to 368 K and equilibrated at this temperature for additional 10 ns , followed by 2 µs simulation using isothermic-isobaric conditions ( T = 368 K , P = 1 atm ) . Periodic boundary conditions and Particle Mesh Ewald calculations were used to deal with long-range effects [58] . RESPA ( Multiple time step ) [59] with a minimum time step of 1 fs was used in conjunction with RATTLE [60] algorithms for maintaining bonds involving hydrogen atoms at equilibrium distances . Multi-microsecond trajectory at high temperature suggests that under the simulation conditions the unfolding trajectory reaches conformations which reproduce known properties of the thermal intermediate in less than 200 ns ( see Results ) . Thus , to enrich our trajectories with the intermediate sate we performed 50 independent simulations starting from 50 different snapshots of the solvated protein extracted every nanosecond during the first 50 ns of the long T = 368 K simulation . Velocities in each snapshot were randomized and after 5 ns re-equilibration the 50 independent trajectories were followed for 0 . 2 µs using identical simulation conditions , representing an aggregate time in the replicas of 12 µs . Such meta-trajectory was analyzed to determine the nature of the intermediate by confronting collected structures with experimental observables of the intermediate state . All MD simulations were carried out with NAMD 2 . 6 [61]–[62] computer program using the CHARMM27 [63]–[64] force field using the MareNostrum supercomputer at the Barcelona Supercomputer Center . Snapshots were saved every picosecond and submitted to a large variety of analyses . Basic geometrical descriptors were determined using the ptraj module of AMBER9 [65]–[67] , clustering was done in function of the RMSd of the clustered structures using the MMTSB Tool set [68] and representative structures of the clusters were determined as those closer to the centroid of each cluster . Secondary structure assignment and solvent accessibility of the representative structures of each cluster were calculated independently using the program PROCHECK [69] . Theoretical changes in the UV spectrum of the protein related to unfolding were determined by analysing the solvent accessible surface of the four Trp ( SASTrp ) and using four references: i ) the crystal structure , ii ) the ensemble obtained in MD simulations at room temperature , iii ) four isolated Trp and iv ) the protein after 50 ns of MD simulation at T = 500 K ( where it reaches RMSd>15 Å from X-ray structure and all structural signatures are lost ) . SAS were computed using the NACCESS [70] program with standard values for protein and solvent particles . Essential dynamics ( ED ) [71] was done to determine the nature of the easiest deformation movements in the native and intermediate states of the protein and to determine the overlap between the essential deformation modes of the protein and the native<$>\raster="rg1"<$>intermediate transition vectors . For this purpose covariance matrices were calculated for the native and intermediate ensembles ( using a common reference system defined by the structurally conserved regions of the protein ) . Such covariance matrix was diagonalized to obtain a set of eigenvectors ( the essential deformation modes ) and the associated eigenvalues ( the amount of variance associated to each eigenvector ) . The similarity between the essential space of native and intermediate was compared using Hess metrics [72]–[73] taking 50 eigenvectors as a common essential space ( at least 90% of variance explained in each ensemble ) : ( 1 ) where n is the dimension of the essential space , A and B are two ensembles and stands for the eigenvectors . Considering the relative size of the protein and the essential space , any >0 . 1 signals a statistically significant similarity [74] . The relative similarity between two essential deformation spaces was computed using [73]: ( 2 ) where the self-similarity indexes where obtained by comparing two different parts of the ensemble . Relative similarity index corrects absolute metrics by the intrinsic noise of MD simulations . A value close or even greater than 1 indicates that considering the noise of the trajectories the two ensembles are identical . The transition from the intermediate to the native states was obtained by taking the first eigenvector calculated by principal component analysis of a meta-ensemble obtained by mixing an equal number of snapshots of the intermediate and the native state . The overlap between the intermediate essential dynamics and the intermediate→native transition vector was determined as: ( 3 ) where Ov is the overlap ( maximum equal to one ) , r is the transition vector and int stands here for the intermediate ensemble . Experimental φ-values profiles were taken from a previous work by Sancho's group [75] , [52] , [53] . Theoretical estimates were derived by individual φicalc values ( i stands for a residue ) determined as the fraction of native contacts , Ni , made by that residue in the MD with respect to those found in the crystal structure , Ninat i . e . , φicalc = Ni/Ninat [76] . Comparison between experimental and simulated φ values was extended to all residues with φi<1 except for residues in helix 3 , where experimental uncertainties in the determinations were large [75] . The ability of a structural ensemble to satisfy the experimental Φ-value profile was studied by analyzing the sum ( over all residues ) of the difference between predicted and simulated Φ-values: ( 4 ) SAXS experiments were performed on the high brilliance beamline ID02 at the European Synchrotron Radiation Facility ( ESRF , Grenoble , France ) . An apoflavodoxin sample at 1 mg/ml concentration was prepared in 50 mM Mops buffer at pH 7 . Several SAXS curves were acquired with a momentum transfer range of 0 . 07<s<0 . 31 Å−1 at a broad range of temperatures ( 6–67°C ) . Solutions were pushed in a capillary into the chamber where they were equilibrated for five minutes . An equivalent protocol was applied to measure buffer profiles . Ten successive frames of 1 s each were acquired for both sample and buffer . Each frame was inspected and the presence of protein damage was discarded . The different scans at each temperature were averaged and subtracted from their buffer counterpart using standard protocols with PRIMUS [77] . The forward scattering , I ( 0 ) , and the effective radius of gyration , Rg , was obtained from the scattering profiles using the Guinier's approximation [78] assuming that , at very small angles ( s<1 . 3/Rg ) , the intensity can be represented as I ( s ) = I ( 0 ) exp ( − ( sRg ) 2/3 ) . SAXS curve measured at 26°C was used to evaluate MD trajectories in native conditions . The evaluation of trajectories in denaturing conditions was performed with the curve obtained from the Multivariate Curve Resolution by Alternating Least Squares ( MCR-ALS ) analysis of the SAXS dataset measured at the complete range of temperatures used to follow thermal denaturation of apoflavodoxin [54] . Principal Component Analysis ( PCA ) of the temperature variation SAXS dataset identified three components in the apoflavodoxin denaturation process that were assigned to the native the unfolded , and an intermediate states . MCR aims at finding the pure SAXS curves of these coexisting species in solution as well as the evolution of the relative concentration of these species upon environmental changes . The decomposition is obtained by solving the matrix equation ( 5 ) where D is the SAXS data matrix , C is the matrix describing the contributions of the N components , ST is the matrix describing the instrumental responses of these N components , and R accounts for the residuals of the fitting . Details of MCR-ALS approach and its application to SAXS data can be found in the original publications . [79]–[82] . Due to the post-processing nature of the SAXS profile of the intermediate , no experimental errors are associated to the derived intensities . A homogeneous 7% of error was assumed for each of the intensities of the curve . The agreement of SAXS profiles with three-dimensional structures of the MD trajectories was evaluated with CRYSOL [83] using default parameters . The χ-value of the fitting between experimental and theoretical curves is used as a measure of the quality of fitting ( the smaller the χ-value , the better the agreement ) . Note that due to the de-convolution process and the use of a small homogeneous error in the intermediate , larger χ-values are expected in the fitting of the intermediate than to that of the native state . Near-UV absorbance spectra of apoflavodoxin [51] at different temperatures were recorded from 250 to 310 nm in a Chirascan spectropolarimeter ( from Applied-Photophysics ) using 30 µM protein solutions in 50 mM Mops , pH 7 in a 4 mm path-length cuvette . The absorbance spectra of native , intermediate and unfolded Anabaena apoflavodoxin were then determined by deconvolution of spectra recorded at different temperatures , using equation: ( 6 ) where the observed absorbance value at a given wavelength and temperature , Y ( λ , T ) , is a linear combination of the values of the different states , Yi ( λ , T ) and of their populations , Xi ( T ) [84] . On the other hand , the populations are calculated at each temperature from the free energy values ΔG1 and ΔG2 previously obtained by global fitting to the sequential three-state model of unfolding curves recorded using absorbance , fluorescence and circular dichroism [53] .
Ten independent 200 ns long MD simulations suggest that the equilibrium structure of the protein in solution is close to that found in the crystal , without any clear unfolding tendency ( Figure 1 ) . The RMSd of trajectory from the crystal structure are always below 3 Å for all replicas , and seems quite stable after the first 10–40 ns where protein relax from lattice contacts existing in the crystal structure ( see Figure 1 ) . The general shape of the protein and the structural core is fully maintained ( see Tm-score plot in Figure 1 ) and the most significant deviation from crystal state is a small expansion of the protein as a result of the removal of lattice constraints , a behaviour very commonly found in massive MD simulations of the proteome [85]–[86] , which is visible in a small ( in average around 0 . 5 Å see Figure 1 ) increase in radii of gyration ( which happens already in the post-equilibration phase ) as well as in an increase around 13% in the solvent accessible surface without changes in the polar/apolar SAS ( see Figure 1 and Suppl . Figure S1 ) . This slight increase in the size of the protein when liberated from lattice constraints is reflected in a small increase in the Trp accessibility , a parameter that correlates with the UV spectra [87] of the protein ( see Suppl . Figure S2 ) . However , all changes in size and shape of the protein upon transferring from crystal to solution are small . Not surprisingly then , the scattering properties computed from the 2 µs ensembles agree very well with the experimental SAXS curve , and also math the conformational preferences indicated by the X-ray structure ( see Suppl . Figure S3 ) . Contacts between residues are massively preserved ( see Suppl . Figure S1 ) and the few native contacts which are transiently lost are typically replaced by alternative contacts with neighbouring residues ( see Suppl . Figure S1 ) . Both α helices and β sheet elements remain fixed at crystal values , while there is a conversion of a portion of residues in β turn into coil conformations ( see Suppl . Figure S1 ) , which is localized in the loop regions . In fact , analysis of B-factors ( Figure 2 ) obtained in the 2 µs meta-trajectory reveals that the regions of larger flexibility are located around the loops of the protein . It is worth noting that most of these loops appear with large B-factors in the crystal structure . However , two loops which are flexible in the simulation ( loop 90–100 , contributing to binding the cofactor; and loop 120–135 , characteristic of long-chain flavodoxins and involved in the binding of partner proteins ) appear with low B-factors in the crystal . Analysis of different crystal structures of this protein in PDB ( including 1FTG used here as starting conformation ) reveals that in the crystal all these loops are directly or indirectly constrained by intermolecular packing contacts , which suggests that the largest mobility found in our simulations cannot be considered a simulation artefact ( see Suppl . Figure S4 ) . Cartesian cluster analysis ( Figure 3 ) reveals that around 91% of the time trajectories are sampling the same conformational basin , which is very close to the crystal structure ( RMSd to the crystal 1 . 5–2 . 5Å ) . The trajectory also populates two alternative basins ( two clusters with population 4% each; RMSds to crystal 2 . 0–2 . 5Å ) that only differ in the conformation of the long loop characteristic of the long-chain flavodoxin family ( including β6 and β7; positions 120–135 ) and , at a minor extent , in the 90–100 loop ( that connecting β4 and α4 ) . Conformational changes in the loops yield to a marginal loss of native contacts in the region ( see Figures 2 and 3 ) , without further changes in the global structure . In summary , extended MD simulations demonstrate that selected force-field and simulation conditions are able to represent the folded form of the protein , which seems to be quite rigid except for local movements in the aforementioned loops . It is never clear what is the effective temperature in a classical MD simulation , since it is force-field dependent [88]–[89] . It is then almost impossible to define a simulation temperature as to guarantee that a finite time simulation will populate the experimentally characterized thermal intermediate . Thus , as described in Methods , we decided to locate ( by comparison with experimental data ) the intermediate as a transient conformational ensemble populated during unfolding at high temperature ( below water boiling point ) . The increase in the temperature does not lead to complete protein unfolding in 2 µs ( see Figures 4–5 ) , something that could be expected only in very fast-folder proteins , typically small proteins with simple kinetic folding mechanisms . The maintenance of TM-score and the hydrophobic solvent accessible surface demonstrate that the protein core is preserved even until 2 µs of trajectory at high temperature . However , although the general fold is maintained , structural distortions from native structure are significant at the end of the simulation ( as noted in the large RMSd ) and affect key elements of α and β secondary structure ( Figures 4–5 ) . Major distortions are first located at the loops , as expected from native simulations ( see above ) , but are later propagated to the neighbouring elements of secondary structure ( see Figures 4–5 ) . Thus , the large movements of loop 90–100 lead to distortions in neighbouring helix α4 , which is shortened in 0 . 2–0 . 5 µs part of the trajectory and is almost completely lost at the end of it . Similarly , distortions in the long loop 120–135 produce early in the unfolding trajectory the disruption of small β-sheet elements β6 , β7 and β5b and the shortening of terminal helix α5 . Large movements of other smaller loops like 53–62 and 75–80 lead also to distortions of neighbouring secondary elements ( for example helix α3 ) , but this happens late in the trajectory and is less dramatic than those noted above . Clearly , our long simulation has not statistical power to describe the intermediate , but suggests a general picture where the perturbation in the loops corrupts in a first step short elements of secondary structure , which has no impact in global structure , but later the α-helices segments are compromised which should eventually yield to the complete unfolding of the protein in longer time scales . Cartesian cluster analysis reveals significant population ( more than 100 ns ) of 5 structural families along the 2 µs trajectories ( Figure 5 ) , which illustrates the increasing level of deformation gained along the simulation . It is tempting to assign the most populated family ( cluster 4 ) as the putative intermediate , but as discussed above there is no guarantee that effective microscopic simulation temperature matches the experimental macroscopic temperature at which the intermediate is detected . Accordingly , we cannot be sure which family represents better the intermediate ensemble and we do not know at which time frame intermediate is populated during our MD unfolding simulations . Clearly , comparison with experimental observables can help to locate the intermediate in our ensemble . The UV spectra determined experimentally for the intermediate ( see Methods ) is very similar to that of the native state , without the blue shift in the spectra which is clear in the unfolded state ( see Suppl . Figure S5 ) . Thus , we can be quite sure that the exposure of Trp side chains has not changed much from native to intermediate state . Based on this criteria the intermediate is detected during the beginning of the simulation ( around 0 . 2 µs; Figure 6 ) , while structures sampled at the second half of the trajectory yield too exposed Trp to justify experimental spectra . The SAXS spectra of the intermediate is well reproduced in the region 0 . 1–0 . 3 µs and later in the second half of the trajectory ( as noted in χ values in Figure 6 ) . Finally , the Φ–profile ( see Methods ) computed experimentally is well reproduced in the 0 . 1–0 . 2 µs region , while structures collected before are too “native-like” and those collected later have advanced too much in the unfolding pathway . In summary , comparison with experimental data strongly suggests that the intermediate is going to be closer to clusters 1–3 than to the most populated cluster 4 ( see Figure 6 ) , and that it is reached quite fast ( around 0 . 2 µs ) during our unfolding simulation . Following the findings obtained from the analysis of the 2 µs trajectory , which suggested that native→intermediate transition happens early in the simulation , we performed 50 independent 0 . 2 µs trajectories , which combined provide us a 10 µs ensemble enriched in the intermediate state . All the different trajectories advance towards protein denaturation ( see Figure 7 ) , with a range of velocities that show a normal distribution with unfolding velocities ranging from 0 . 4 to 0 . 8 nm RMSd/0 . 2 µs . The lack of unusually slow or fast unfolding pathways [90] suggests the existence of a unique mechanism for the transition from folded to intermediate state under the selected simulation conditions , which is characterized by first a focalization of structural deformations in loops ( Figure 7 ) and later a transfer of such perturbation to the surrounding elements of secondary structure ( see Figure 8 ) , matching the general unfolding trends found in the 2 µs trajectory . Cartesian clustering of the 10 µs meta-trajectory allowed us to detect six major “states ( clusters ) ” , four of them with populations above 5% . Not surprisingly , the most populated one ( 69% of meta-trajectory ) is that describing a near-native conformation , which appears populated in the beginning of all the individual trajectories . As the unfolding progresses , partially unfolded conformations , characterized by distorted loop conformations and partial losses of neighbouring secondary structure become populated ( Figure 8 ) . Thus , in structures assigned to cluster 2 ( 10% meta-trajectory , populated in 55% trajectories ) the large movements of the long loop ( 120–135 ) have led to the loss of short β strand elements β6 and β7 . Ensembles represented by clusters 3 and 4 ( 12% and 7% meta-trajectory , populated in 65% and 45% individual trajectories ) are characterized by an advance in the distortion produced by loop oscillations , either to the helix α4 ( cluster 3 ) or the helix α3 ( cluster 4 ) . Finally , the minor clusters 5 and 6 represent much more distorted conformations , where a significant amount of secondary structure is lost and the departure from native basin is quite evident ( Figure 8 ) . Clusters 5 and 6 account for less than 1% of the entire meta-trajectory and are sampled only in two of the individual trajectories ( one for each ) , which suggest that they do not fit the experimental requirements of the intermediate . It is very tempting to try to identify one of the above mentioned clusters with the thermal intermediate , but analysis of the individual trajectories show that in reality clusters 2–4 and part of structures assigned to cluster 1 interchange in a fast way and share many common characteristics , with a well conserved central core and largely distorted loop regions . The fast and large movement of such loops ( and neighbouring secondary elements ) generates a large dispersion in the structures when projected into the Cartesian space , which is reflected in the different assignment of structures to different clusters , when they share many key structural characteristics . It is also worth to note that structures which are within the same cluster can yield very different values of some experimental observables ( see Figure 9 ) , while structures very distant in terms of RMSd , and accordingly assigned to the different clusters can be indistinguishable in terms of experimental observables ( see Figure 9 ) . In summary , it seems that the intermediate cannot be represented as a small ensemble defined as a narrow basin centered in a well-defined structure , but as a wide ensemble of conformations that cover a wide range of Cartesian space , but that share a common conformational core .
We interrogate our 10 µs ensemble to determine how many of these structures fulfill all the experimental requirements of the ensemble known experimentally for the thermal intermediate . Considering a loosely criteria ( SASTrp between 100 and 300 Å2 , fitting the SAXs curve with a χ below 1 . 5 and fitting the Φ-value profile with absolute accumulated error below 2 ) almost 30% of the ensemble is annotated as intermediate . If we assume that experimental measurements for the intermediate are very accurate and use a much more restrictive criterion ( SASTrp between 100 and 300 Å2 , χ<1 . 0 and Φ-error<1 . 0 ) the intermediate ensemble is reduced to around 10% of the meta-trajectory . Such an ensemble is contributed by all individual trajectories and is proportionally enriched with structures assigned to clusters 2–4 , with no contribution of clusters 5 and 6 . When analyzed , the intermediate sampling shows a quite interesting picture of the structure that is transiently populated during thermal unfolding of the protein ( Figures 10–11 ) . The structure has enlarged with respect to the solution ensemble and hydrophobic solvent accessible surface has increased significantly , a fingerprint of a partially unfolded structure . A significant number of native contacts ( defined as those present in the solution ensemble ) are lost , especially those involving the protein loops , which have disappeared completely ( Suppl . Figure S6 ) . However , the structure maintains still many native inter-residue contacts , mostly located in the central core , where the amount of secondary structure has decreased , but is still quite significant ( Figures 10–11 ) . Clearly , analysis of the results demonstrates that the intermediate is not an alternative structure of the protein , but has to be represented as a wide ensemble ( average RMSd between structures in the ensemble is around 0 . 6 nm; Figure 11 ) . Two broad regions can be easily recognized in the protein: the central core , where the native fold is well preserved and the loops ( including the long loop hosting a small β-sheet encompassing strands 6 and 7 ) , which adopt a canonical random coil confirmation ( Figure 11 ) . It is very interesting to realize that the large flexibility movements governing the essential dynamics in the intermediate ensemble are already a maximization of the intrinsic deformation pattern of the native state of the protein ( absolute similarity ( γ ) = 0 . 52; relative similarity ( κ ) = 0 . 76 , see eqs . 1 and 2 ) , as it was already suggested by B-factor distributions ( see Figure 2 ) . Altogether , the intermediate fits perfectly in the definition of a partially disordered protein with a solid-like core and a liquid-like external loop core . It is very encouraging that such a representation of the intermediate fits well with the picture derived from the analysis of the NMR spectra of a mutant , which is believed to adopt intermediate-like conformation under native conditions [91] . Our MD simulations suggest a quite complete picture of the initial stages of the thermal unfolding of apoflavodoxin , which might be common to other proteins having long loops stabilized by weak contacts . Thus , under native conditions the protein has an intrinsic tendency to become a partially disordered protein , but several loop-loop contact keep the potentially flexible part of the protein reasonably organized . When the temperature increases these loops gain kinetic energy and in a quite short period of time become random coils ( see Figure 2 ) . The anchoring points of the loops , with the exception of short β-sheet elements , are very stable and held together the core of the structure defining the experimentally detected intermediate . Additional thermal energy will be then concentrated in the anchoring points of the loops , particularly in the helices 3 and 4 , which are the Achilles' heel of the apoflavodoxin core . The distortion of these helices opens the structure and should lead to the final disruption of the three dimensional structure of the protein in longer time scales . Under this general picture , the lack of intermediate when denaturing agent is urea [92] can be easily rationalized , since urea will attack directly the core of the protein [39] , eliminating the resistance points that stop the thermal unfolding pathways in a partially disordered conformation . Under native conditions the thermal intermediate acts as an “in-path” stationary state , since the essential deformations of the intermediate implicitly code the intermediate→native transition , as noted in the high overlap ( Ov = 0 . 63; see eq . 3 ) between the intermediate essential deformation subspace and the intermediate→native transition vector . This finding strongly suggests that the intermediate should be considered as an “activated-high entropy” form of the native state , ( see RMSd oscillations in Figure 11 ) , with properties of partially disordered protein , which acts as an attractor of folding routes toward a state that in the absence of an excess of kinetic energy will converge in a down-hill manner to the native form . We can hypothesize that a non-negligible number of partially disordered proteins , which adopt a well-defined three dimensional structure only in the presence of partner , can be considered as generalized examples of three-state folder proteins , which in native conditions populates conformations containing well-structured cores and very mobile regions . The flexibility pattern of such intermediates should favour a down-hill transition to a well-defined three dimensional structure in the presence of interactions stabilizing the disordered region ( in these case binding partners ) .
|
A simplistic view of protein structure tends to emphasize the opposition between the native state and the denatured ensemble of unfolded conformations . In addition to these extreme conformations , proteins subjected to a variety of perturbations often populate alternative partly unfolded conformations , some of which are close in energy to the native state and , accordingly , can be populated under native or quasi-native conditions . There is increasing evidence that these “perturbed” conformations participate in protein function or , in some cases , are related to the outcome of folding diseases . We have used the “state of the art” molecular dynamics combined with a variety of experimental techniques to characterize for the first time , to our knowledge , the thermal intermediate of a three-state folding protein ( apoflavodoxin ) . Based on our results we have been able to suggest a general mechanism of thermal unfolding in complex proteins and to determine interesting links between thermal intermediates and partially unfolded proteins .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"physics",
"computer",
"science",
"chemistry",
"biology"
] |
2012
|
Defining the Nature of Thermal Intermediate in 3 State Folding Proteins: Apoflavodoxin, a Study Case
|
Inappropriate recollections and responses in stressful conditions are hallmarks of post-traumatic stress disorder and other anxiety and mood disorders , but how stress contributes to the disorders is unclear . Here we show that stress itself reactivates memories even if the memory is unrelated to the stressful experience . Forced-swim stress one day after learning enhanced memory recall . One-day post-learning amnestic treatments were ineffective unless administered soon after the swim , indicating that a stressful experience itself can reactivate unrelated consolidated memories . The swim also triggered inter-hemispheric transfer of a lateralized memory , confirming stress reactivates stable memories . These novel effects of stress on memory required the hippocampus although the memories themselves did not , indicating hippocampus-dependent modulation of extrahippocampal memories . These findings that a stressful experience itself can activate memory suggest the novel hypothesis that traumatic stress reactivates pre-trauma memories , linking them to memory for the trauma and pathological facilitation of post-traumatic recall .
Inappropriate negative responses in emotionally neutral circumstances and the development of negative associations to harmless stimuli are core , debilitating features of post-traumatic stress disorder ( PTSD ) , depression , and a host of anxiety and mood disorders . The possibility that stress itself might promote inappropriate associations between unrelated memories and events has not been explored , although a central role for stress and memory in these disorders is established [1]–[3] . Here we demonstrate that a single stressful experience can activate already consolidated memories outside of their appropriate context . These findings provide the basis for a novel hypothesis: by triggering out-of-context activation of memories stressful events themselves create opportunities for inappropriate associations to form , thereby promoting and perpetuating anxiety and mood disorders .
Stress modulates memories that are undergoing cellular consolidation [4] and may play a role in the swim-induced enhancement of memory . Experiments 4 and 5 were designed to test if the day-old memory is undergoing cellular consolidation at the time of swim . If the memories were consolidating , then amnesic treatments by electro-convulsive shock ( ECS ) [5] or the beta-adrenergic antagonist Propranolol [6] 24 h after conditioning should impair retention . Temporary inactivation of hippocampus with long-acting ( 6–10 h ) tetrodotoxin ( TTX ) [24] or short-acting ( ∼30 min ) lidocaine [25] was used to test if hippocampus is important for OCAM . The TTX injection was determined to block neural activity in both the dorsal and the ventral hippocampi [24] .
The results of this set of experiments suggest that stress can activate memory , even if the memory is unrelated to the stressful experience . We use the term “memory activation” in the established sense that the term is used in the consolidation and reconsolidation literatures , to mean that memory is in a labile ( “active” ) state rather than an inert ( “inactive” ) state [9] . We provided multiple lines of evidence that a stressful swim returned a consolidated memory to a labile state . As a result , expression of the memory was strengthened , and if the memory was lateralized , the swim triggered its interhemispheric transfer . The activation by forced-swim stress was independent of the conditioned and external contextual stimuli that were present during learning , leading us to call the phenomenon OCAM ( alternative interpretations are considered and rejected in the section that follows ) . OCAM seems to be a general phenomenon that does not depend on whether the conditioned response is rapidly extinguished ( Experiment 1 ) or persistent ( Experiments 2–10 ) or whether the activated memory is acquired during single ( Experiment 5 ) or multiple ( all other experiments ) appetitively ( Experiment 1 ) or aversively ( all other experiments ) conditioned trials that reinforce an inhibitory ( Experiment 5 ) or an active ( all other experiments ) conditioned response . OCAM also seems to occur whether or not memory expression depends on the hippocampus ( inhibitory avoidance ) or the neocortex ( left/right discrimination ) . Both beta-adrenergic activation ( Figure 3A ) and Dexamethasone-suppressible HPA activity ( Figure 3B ) were required for the OCAM effect , indicating a central role for stress and two key stress mediators . This is consistent with the idea that stress and arousal act together to modulate memory mechanisms [26]–[28] . The forced swim modified a consolidated memory and no anterograde learning effects were detected ( Experiment 3b ) . The results of the first three experiments that assayed enhanced memory expression did not distinguish between whether the forced swim had its effect on memory storage or the process of memory retrieval , including for example by inducing perseveration during the reversal tests for memory strength . However , the results of Experiments 4 and 5 with amnestic agents indicate the swim-induced modifications only occurred soon after the swim but not after a 5-h delay . This strongly suggests the effect of swim was not on retrieval itself , which occurred a day later . The results of Experiment 2b are also consistent with an effect on storage rather than retrieval , because 6 d after the swim , we also observed the enhanced expression of intensively conditioned left/right discrimination memory ( Figure 1D ) . Swim-induced increases in circulating hormones are unlikely to persist for 6 d ( corticosterone returns to baseline levels within a day of the forced swim; Figure 1B ) . This is additional evidence that the effect of the swim was not on the retrieval process itself . Because IHT is conventionally interpreted as indicating memory formation in a “naïve” brain site , perhaps the strongest evidence that the swim altered memory storage and not retrieval is that IHT was induced during the swim ( Experiments 7 and 10 ) . An effect on storage rather than retrieval would be consistent with the effects of consolidation , reconsolidation , and protein synthesis inhibition [31] , which are all also believed to affect memory storage . The forced swim activated consolidated memories that were 24 h old , to the best of our knowledge mimicking the basic phenomenon of reconsolidation , possibly with an important distinction . Reconsolidation is said to occur when a consolidated memory is retrieved and the activation converts the memory from a biochemically stable state to a labile state [10] , [32] that is characterized by additional memory formation [33] in which the original memory can be modified , strengthened , or changed ( see [34] for review ) . Curiously , we did not observe any memory disruption due to the forced swim stress , which seemed to activate , strengthen , or expand the localization of established memories . Further work is necessary to determine whether the stress-induced activation of memory we observed is biochemically identical to consolidation or reconsolidation , a pair of related but biochemically distinguishable phenomena [35]–[38] . It is important in the present context to point out that both consolidation and reconsolidation are specific to the memory that was directly activated by learning or retrieval [39] , whereas we observed that the stressful forced swim activated several different memories , none of which were related to the stressful experience . This distinguishes the stress-induced activation of memory phenomenon we describe from conscious , recall-triggered activation of memory . For example , after CS2 → CS1 → US second-order-conditioning , recall elicited by CS2 causes the directly activated CS2 → CS1 association to become labile without altering the indirectly activated CS1 → US association [39] . We investigated the OCAM effect in several memories , but we only assayed each memory in isolation , so whether stress activates all or a subset of the rat's memories remains an open question . We suspect that the answer will be complex because whether and how a memory is modified after retrieval depends on the strength and age of the memory [40] , the brain regions involved in information storage [41] , as well as the duration of the reactivation and whether extinction occurred [42] . A model of memory that attempts to synthesize the consolidation and reconsolidation literature [35] states that learning creates a memory trace , and both learning and reactivation evokes memory modulation events . The stabilization of memory is a graded function of the amount of modulation for each memory . This view predicts that forced swim will be more likely to activate recent memories than remote ones . Providing evidence for this hypothesis will require extensive experiments that manipulate both the strength of the memory and the interval between learning and forced swim . However , regardless of whether or not there is a restricted time window during which the stressful swim can cause the modification of a once consolidated memory , the present data demonstrate that OCAM occurs at the very least for consolidated memories that are 24 h old . Blocking hippocampal activity during the swim prevented both the swim-induced memory enhancement and the swim-induced IHT of lateralized memory for left/right discrimination , the learning or expression of which is insensitive to hippocampal inactivation . This suggests that hippocampal activity during the swim was necessary for the out-of-context activation of an extra-hippocampal memory . The results do not indicate whether the role of hippocampus was only to mediate the response to stress or whether hippocampal memories were specifically activated . The data demonstrate the hippocampus plays a role in memory beyond its role in associative memory storage [43]–[45] , adding to the evidence that hippocampus modifies recent memories that are stored elsewhere in the brain [8] , [46] and is a site along with amygdala for the combined roles of stress and arousal in mediating memory modulation [26] , [28] . The results of Experiments 8 , 9 , and 10 suggest that OCAM is a hippocampus-dependent process that appears to alter memory in extrahippocampal sites . OCAM is a common feature of human conscious recollection , but despite a recent suggestion that hippocampus is important for recollection in rats [47] , there are alternative interpretations of those data and whether rats recollect remains controversial [48] . To our knowledge OCAM has never been described in non-humans . Future studies will determine whether swim-induced OCAM in rats is related to human out-of-context recollection in part by investigating whether the same hippocampal-neocortical networks are engaged . The electrophysiological re-expression of recently expressed hippocampal and neocortical activity patterns has been recorded from monkey [49] and rat during sleep [50] , [51] and during conscious human recall [52] . It is of substantial interest whether such electrophysiological reactivation is the expression of memory and whether it will occur during swim-induced OCAM . Indeed , our extensive use of the intensive aversive L/R discrimination protocol was motivated in part by the fact that it generates stereotyped behavior that is amenable to searching during the stress for replay of the place cell ensemble activity sequences that are expressed during memory formation , as the rat runs up the start arm to the choice point . The findings presented here indicate that under acute stress , the hippocampus is involved in activating a set of arbitrary memories that can be stored at both hippocampal [41] and extrahippocampal sites . Although we evaluated the effect of stress on single memories , one at a time , we assume that stress can concurrently activate many memories for two reasons . First , the forced swim had little in common with the learning and retrieval experiences we investigated , suggesting stress affects memory in general rather than just memories of specific , stress-related experiences . Second , we observed that the stressful swim enhanced a variety of associations that included a weak appetitively L/R discrimination ( Experiment 1A ) , as well as more persistent aversively conditioned L/R discrimination and inhibitory avoidance responses . These findings extend our understanding of the consequences of memory consolidation and reconsolidation , which our data demonstrate can be modified by stress . While further investigations of the effects of stress on multiple , concurrent memories are warranted , our observations indicate that stressful experience alters diverse associative memories . We only found evidence of memory enhancement , for both weak and strong associations; it however remains possible that other forms of memory that we did not test were weakened by the stress . Nonetheless , at this point , our observations suggest that in stress-induced OCAM , stress acts to generally strengthen memory rather than acting to strengthen some and weaken others . If confirmed , this may help understand the memory dysfunction in PTSD and other stress-related mood disorders . We hypothesize that stress-triggered memory activation creates a condition where multiple memories coactivate , and through mechanisms of synaptic plasticity [53] that include both long-term potentiation and depression [54]–[57] , consolidation and reconsolidation , their subsequent expression is enhanced . We point out that there is evidence that recall which activates a consolidated memory can cause additional information to become incorporated into that memory via the molecular events associated with consolidation [58] but not reconsolidation [59] . According to our hypothesis , already strong traumatic memories or the stress itself can become inappropriately associated with other memories of everyday experience , making the subsequent experience and recall of everyday events more likely to trigger unwanted recall of the traumatic memory .
The experiments were conducted in accordance with Institutional ( SUNY , IACUC 07-197-05 ) and NIH guidelines , and the directive of the European Communities Council ( 6/609/EEC ) . Male rats of the Long-Evans strain weighing 350–450 g were used . The experiments were performed during the light period ( 07:00 to 19:00 ) of a 12 h:12 h cycle . Rats were habituated to handling by the experimenter for 3–5 d prior to behavioral testing . Trunk blood was collected under Halothane anesthesia . After overnight storage , the blood was centrifuged at 4 , 000 rpm for 10 min; the supernatant was withdrawn and then stored frozen until assayed by radioimmunoassay . More than 10 experiments were performed requiring the use of a large number of behavioral and experimental manipulations . Here in the Methods we describe the procedures themselves , and to optimize the clarity of the report , we describe each experiment's protocol in an introduction to the individual experiment in the Results . Rats were placed in a plastic holding cage next to the forced swim bucket . A pair of electrodes was clipped to the ears and an ECS ( 50 mA , 50 Hz , 1 s ) was delivered . After the treatment , the rats were returned to their home cage to recover . Average measures ± SEM are reported . Significant differences confirmed by ANOVA were followed by Newman-Keuls post hoc tests . The results of these pair-wise comparisons are reported in the main text , and the statistical details are given in the corresponding figure legends . Chi-square and t tests were also used as indicated in the text .
|
This work identifies a powerful effect of stressful experiences on memories . We report that a single intensely stressful experience can activate memories in a situation that has essentially no physical or motivational relationship to the stressful experience . Using a forced-swim test as a stressor in rats , we find that this treatment was able to activate unrelated memories formed 24 hours earlier . We also find that the hippocampus of the brain is required for this effect of stress but that recall of the memories themselves does not . The ability of stress to activate memories that are unrelated to the stressful event may help to explain how memories can sometimes become pathological and uncontrollable following traumatic events , as in post-traumatic stress disorder . Our findings suggest the novel hypothesis that the stress of the traumatic event activates neutral , unrelated memories , which then become associated with the traumatic event . Subsequent normal recall of the neutral memories can more easily trigger inappropriate recall of the traumatic event , initiating another bout of stress and inappropriate associations of neutral and traumatic memories .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"neuroscience/behavioral",
"neuroscience",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration",
"neuroscience/animal",
"cognition",
"neuroscience/experimental",
"psychology"
] |
2010
|
Stress-Induced Out-of-Context Activation of Memory
|
The MglA protein is the only known regulator of virulence gene expression in Francisella tularensis , yet it is unclear how it functions . F . tularensis also contains an MglA-like protein called SspA . Here , we show that MglA and SspA cooperate with one another to control virulence gene expression in F . tularensis . Using a directed proteomic approach , we show that both MglA and SspA associate with RNA polymerase ( RNAP ) in F . tularensis , and that SspA is required for MglA to associate with RNAP . Furthermore , bacterial two-hybrid and biochemical assays indicate that MglA and SspA interact with one another directly . Finally , through genome-wide expression analyses , we demonstrate that MglA and SspA regulate the same set of genes . Our results suggest that a complex involving both MglA and SspA associates with RNAP to positively control virulence gene expression in F . tularensis . The F . tularensis genome is unusual in that it contains two genes encoding different α subunits of RNAP , and we show here that these two α subunits are incorporated into RNAP . Thus , as well as identifying SspA as a second critical regulator of virulence gene expression in F . tularensis , our findings provide a framework for understanding the mechanistic basis for virulence gene control in a bacterium whose transcription apparatus is unique .
Francisella tularensis is a Gram-negative , facultative intracellular pathogen and the aetiological agent of tularemia , a disease that can be fatal in humans . Although outbreaks of tularemia are thought to be rare , infections caused by F . tularensis have become a public health issue because of the potential for using this organism as a bioweapon . As few as ten organisms can constitute an infectious dose , and the pneumonic form of tularemia , which has a particularly high mortality rate , can be acquired when the organism is aerosolized [1 , 2] . Relatively little is known regarding the molecular mechanisms of F . tularensis pathogenesis , in part because of the genetic intractability of this organism . Evidence suggests that the pathogenicity of F . tularensis depends on its ability to survive and replicate within macrophages . However , only a handful of genes that are required for intramacrophage survival have been identified [2–4] . One of these encodes a putative transcription regulator called macrophage growth locus A ( MglA ) that is responsible for controlling the expression of multiple virulence factors that are themselves required for survival within macrophages [3 , 5 , 6] . In particular , in F . tularensis subspecies novicida , MglA positively regulates the expression of multiple virulence genes , some of which are located on a pathogenicity island [5–7] . MglA also regulates the expression of many other genes that may or may not be implicated in virulence [5 , 6] . However , the mechanism of this MglA-dependent regulation is unknown . MglA is an ortholog of the stringent starvation protein A ( SspA ) from Escherichia coli ( the two proteins share 21% identity and 34% similarity at the primary amino acid level ) . In E . coli , SspA is a RNA polymerase ( RNAP ) –associated protein that is thought to regulate the expression of a subset of genes under conditions of stress [8 , 9] . SspA in E . coli is also essential for the lytic growth of bacteriophage P1 [10] . Specifically , Hansen and colleagues have shown that E . coli SspA functions as an activator of P1 late gene expression , acting in concert with a phage-encoded transcription activator called Lpa [11] . SspA homologs have also been found in a number of pathogens other than F . tularensis , and several of these homologs appear to play important roles in virulence gene regulation [12–14] . Recently , the genomes of two different strains of F . tularensis have been sequenced ( [15]; GenBank accession number NC_007880 ) . One of these strains is Schu S4 , a highly virulent strain of subspecies tularensis , and the other is the live vaccine strain ( LVS ) , an attenuated derivative of one of the subspecies holarctica strains . The genome sequences reveal that F . tularensis contains an MglA-like protein , which has been annotated as SspA [15] . In addition , the sequences reveal that , unlike any other bacterial genome sequenced to date , the F . tularensis genome contains two genes , designated rpoA1 ( or FTT0350 in the annotated Schu S4 genome sequence ) and rpoA2 ( FTT1442 ) , that encode different α subunits of RNAP . The α subunit of bacterial RNAP forms a dimer in the RNAP complex ( subunit composition α2ββ′ωσ ) . The two putative α subunits in F . tularensis differ from one another in regions that are predicted to be critical for dimer formation , promoter recognition , and activator interaction . Given the fundamental roles the α subunit plays in transcription regulation [16] , this may have far-reaching implications for how gene expression is controlled in F . tularensis . Here , we show that both MglA and SspA associate with RNAP and positively control virulence gene expression in F . tularensis . Furthermore , we present evidence that the ability of MglA to control virulence gene expression is linked to its ability to associate with RNAP . We also show that the two distinct α subunits encoded by the F . tularensis genome are both incorporated into RNAP . Our results provide a framework for understanding the mechanistic basis for virulence gene control in this organism .
To address the question of whether MglA and SspA are associated with RNAP in F . tularensis , we took a directed proteomic approach in which we affinity purified RNAP from F . tularensis and identified candidate RNAP-associated proteins by tandem mass spectrometry ( MS/MS ) . To facilitate affinity purification of F . tularensis RNAP , we adapted the tandem affinity purification ( TAP ) strategy [17] for use in F . tularensis . Using the integration vector depicted in Figure 1A , we constructed a strain of F . tularensis LVS in which the native chromosomal copy of the rpoC gene ( encoding the β′ subunit of RNAP ) has been modified to encode a TAP-tagged form of β′ ( β′-TAP ) . The resulting strain ( LVS β′-TAP ) thus synthesizes , at native levels , β′ with a TAP-tag fused to its C-terminus . As a control we constructed , in a similar fashion to the LVS β′-TAP strain , a strain that synthesizes SucD ( a subunit of the succinyl CoA synthetase complex ) with a TAP-tag fused to its C-terminus ( LVS SucD-TAP ) . Lysates were made from cells of the LVS SucD-TAP strain and cells of the LVS β′-TAP strain . Proteins were then purified by TAP essentially as described [18] , separated by SDS-PAGE , and stained with silver . Two proteins with apparent molecular weights of ~20 kDa were found that specifically co-purified with β′ ( Figure 1B ) . Bands corresponding to these proteins were excised from the gel , and , following in-gel digestion with trypsin , nanoelectrospray MS/MS was used to identify the proteins as MglA and SspA ( unpublished data ) . Among the proteins that appear to specifically co-purify with β′ , one had an apparent molecular weight of ~70 kDa , two had apparent molecular weights of ~40 kDa , and one had an apparent molecular weight of ~10 kDa ( Figure 1B ) . Bands corresponding to these proteins were excised from the gel , and , following in-gel digestion with trypsin , nanoelectrospray MS/MS was used to determine the identity of each protein . Accordingly , the ~70- and ~10-kDa proteins were found to be the σ70 and ω subunits of RNAP , respectively , and the two ~40-kDa proteins were found to be α1 and α2 , the products of the rpoA1 and rpoA2 genes , respectively ( unpublished data ) . Therefore , F . tularensis RNAP contains two distinct α subunits and associates with both MglA and SspA . We reasoned that if MglA were associated with RNAP in F . tularensis , then RNAP would be expected to co-purify with TAP-tagged MglA . To test this prediction , we constructed a strain of LVS that synthesized MglA with a TAP-tag fused to its C-terminus ( LVS MglA-TAP ) . Figure 1B shows the results following TAP of MglA from this strain . In support of the hypothesis that MglA associates with RNAP in F . tularensis , subunits of RNAP ( including α1 , α2 , β , and β′ ) co-purified with MglA . Furthermore , SspA also co-purified with MglA . If MglA and SspA were to interact with RNAP independently of one another , then SspA might not be expected to co-purify with MglA-TAP . In particular , complexes containing RNAP and MglA would be expected to be distinct from those containing RNAP and SspA if MglA and SspA competed for the same binding site on RNAP . It is of course possible that there is more than one contact site for MglA and/or SspA on RNAP . However , because SspA from Yersinia pestis is a dimer [19] , one plausible explanation for our results is that MglA and SspA physically interact with one another and associate with RNAP as a heterodimer . An alternative possibility is that MglA might form two distinct complexes , one with SspA , and another with RNAP; in this case , the SspA that co-purifies with β′-TAP ( Figure 1B , lane 2 ) would represent SspA-RNAP complexes that are distinct from MglA-RNAP complexes . In an attempt to distinguish between the possibilities that MglA and SspA form a single complex or separate complexes with RNAP , we constructed a strain that synthesized MglA-TAP and carried an in-frame deletion of sspA ( LVS ΔsspA MglA-TAP ) . Figure 1C shows the results following TAP of MglA-TAP from cells of this strain . In the absence of SspA , MglA fails to co-purify with RNAP . Our findings are therefore consistent with the hypothesis that MglA and SspA are associated with one another in F . tularensis and that the resulting MglA–SspA complex associates with RNAP . In order to test explicitly whether or not MglA and SspA can interact with one another directly , we used a bacterial two-hybrid assay configured to permit the detection of both dimeric and higher order complexes . This two-hybrid assay is based on the finding that any sufficiently strong interaction between two proteins can activate transcription in E . coli , provided that one of the interacting proteins is tethered to the DNA by a DNA-binding protein and the other is tethered to a subunit of E . coli RNAP [20 , 21] . Specifically , contact between a protein ( or protein domain ) fused to the ω subunit of E . coli RNAP and another protein fused to a zinc finger DNA-binding protein ( referred to as Zif ) activates transcription of a lacZ reporter gene ( with an upstream Zif binding site ) in E . coli cells ( Figure 2A ) [22] . Because Zif binds its cognate recognition site as a monomer , and because the ω subunit is monomeric in the RNAP holoenzyme complex , this configuration of the assay is ideally suited to detecting interactions between two protein monomers ( i . e . , dimer formation ) . Our strategy for determining whether MglA can interact with SspA ( thus potentially forming a heterodimer ) involved the use of two fusion proteins , one comprising MglA fused to Zif and the other comprising F . tularensis SspA fused to the ω subunit of E . coli RNAP ( Figure 2A ) . Accordingly , we fused full-length MglA ( residues 1–205 ) to the N-terminus of Zif , and we fused full-length SspA ( residues 1–210 ) to the N-terminus of ω . We then tested whether the MglA-Zif fusion protein could activate transcription from an appropriate test promoter in cells containing the SspA-ω fusion protein . Plasmids expressing the MglA-Zif and SspA-ω fusion proteins were introduced into E . coli strain KDZif1ΔZ , which harbors a test promoter linked to lacZ on an F′ episome ( Figure 2A; [22] ) . In support of the idea that MglA and SspA interact with one another directly , the MglA-Zif fusion protein activated transcription of the reporter gene strongly ( up to ~22-fold ) in cells containing the SspA-ω fusion protein , whereas Zif ( without the fused MglA moiety ) did not ( Figure 2B ) . The observed activation was dependent upon the ability of the ω moiety of the SspA-ω fusion to interact with E . coli RNAP ( unpublished data ) . Another control revealed that MglA-Zif did not activate transcription from the test promoter in the presence of the unrelated Gal11P-ω fusion protein ( Figure 2B ) . Given that MglA and SspA are similar to one another at the primary amino acid level , we sought to determine whether MglA or SspA ( or both ) could also form homomeric complexes . We therefore constructed two additional fusion proteins , one in which full-length MglA was fused to the N-terminus of ω , and another in which full-length F . tularensis SspA was fused to the N-terminus of Zif . Results depicted in Figure 2B show that specific interactions were also detected between the MglA-Zif and MglA-ω fusion proteins , as well as between the SspA-Zif and SspA-ω fusion proteins . Our findings suggest that MglA and SspA can form both heteromeric and homomeric complexes . As an additional test of whether MglA and SspA interact with one another to form a heteromeric complex , we co-expressed genes encoding S-tagged MglA ( MglA-S ) and His-tagged SspA ( SspA-His6 ) in E . coli and asked whether the two proteins co-purified with one another following metal chelate affinity chromatography . Specifically , we made a cell lysate from E . coli , synthesizing both SspA-His6 and MglA-S . We then applied the lysate to a metal chelate affinity resin to capture SspA-His6 and any associated proteins , washed the resin , and eluted proteins that were specifically bound to the resin with imidazole . As a control , we treated an E . coli cell lysate containing both non-tagged SspA and MglA-S in an identical manner . ( It was important to use non-tagged SspA as a control , rather than no SspA , because MglA-S is insoluble when expressed alone , at least under the conditions of our experiment . ) Immunoblot analyses revealed that equal amounts of soluble MglA-S were present in both lysates and that MglA-S co-purified with SspA-His6 ( Figure 3 ) . Furthermore , immunoblotting with an antibody that recognizes the α subunit of E . coli RNAP revealed that the association between MglA-S and SspA-His6 was not mediated indirectly through interactions with the E . coli RNAP complex ( Figure 3 ) . These findings provide additional evidence that MglA and SspA interact with one another directly . Previous studies revealed that MglA positively regulates expression of the pdpD , iglA , iglC , iglD , and pdpA genes in F . tularensis subspecies novicida ( and of these , iglA , iglC , and pdpA have been implicated in intramacrophage survival and virulence ) [4 , 5 , 7] . These genes are located on a pathogenicity island [7] . In subspecies novicida there is only one copy of this island [7] , whereas in the highly virulent Schu S4 strain ( subspecies tularensis ) , there are two identical copies of the pathogenicity island [15] . Similarly , the attenuated strain LVS ( a derivative of one of the subspecies holarctica strains ) , contains two copies of the pathogenicity island . We have shown that in LVS both MglA and SspA associate with RNAP and that the association of MglA with RNAP is dependent upon SspA . We therefore reasoned that if MglA controls virulence gene expression through its association with RNAP , then the absence of SspA in an sspA mutant strain should result in decreased levels of virulence gene expression . That is to say , we would expect that both MglA and SspA are required for virulence gene expression . To compare the effect on virulence gene expression of removing either MglA or SspA in LVS , we constructed mutant strains of LVS carrying in-frame deletions of either the mglA gene ( LVS ΔmglA ) or the sspA gene ( LVS ΔsspA ) . Because both ΔmglA and ΔsspA mutant cells had a growth defect compared to wild-type LVS cells ( culture doubling times of ~171 , ~182 , and ~147 min , respectively , under the conditions of our experiment ) , we also used , as a control , cells of a mutant containing a mariner transposon in the gene FTL_0951 ( LVS::FTL0951 ) ; this strain has a more severe growth defect than either the ΔmglA or ΔsspA mutant strain ( with a corresponding culture doubling time of ~285 min ) . Cells of the wild-type LVS strain , cells of the LVS ΔmglA mutant strain , cells of the LVS ΔsspA mutant strain , and cells of the control strain LVS::FTL0951 were grown to mid-log ( where , in the wild-type strain , the mglA and sspA genes are highly expressed; see Figure S1 ) , total RNA was isolated , and expression of the iglA , iglC , and pdpA virulence genes in each of the four strains was measured using quantitative real-time reverse transcriptase ( RT ) –PCR . The results depicted in Figure 4 show drastically reduced amounts of the iglA , iglC , and pdpA transcripts in both LVS ΔmglA mutant cells and LVS ΔsspA mutant cells when compared to LVS wild-type cells . Complementation of the LVS ΔmglA and LVS ΔsspA mutant cells with plasmids expressing either the mglA or the sspA gene , respectively , restored the amounts of iglA , iglC , and pdpA transcripts to near wild-type levels ( Figure S2 ) . Moreover , the effects of the mglA and sspA deletions on virulence gene expression were not simply due to the effect of these deletions on growth rate; only a modest reduction in the amounts of the iglA , iglC , and pdpA transcripts were evident in cells of the slow growing LVS::FTL0951 mutant when compared to LVS wild-type cells ( Figure 4 ) . Thus , both MglA and SspA appear to positively control expression of the iglA , iglC , and pdpA virulence genes in LVS , though we cannot exclude the possibility of indirect effects . Because we showed that the association of MglA with RNAP is dependent on SspA ( Figure 1C ) , we infer that MglA controls virulence gene expression through its association with RNAP . Taken together , our findings suggest that MglA and SspA interact with one another directly to form a complex that associates with RNAP to positively control expression of the iglA , iglC , and pdpA virulence genes located on the F . tularensis pathogenicity island . Recently , DNA microarray analyses have revealed that in subspecies novicida , MglA controls the expression of ~100 genes , with the vast majority being positively regulated [6] . In particular , the expression of all of the genes on the novicida pathogenicity island was found to be positively regulated by MglA . In addition , MglA was found to regulate the expression of virulence genes located outside of the pathogenicity island , as well as other genes not thought to play a role in virulence [6] . If MglA and SspA primarily function together to regulate gene expression , then we would predict that MglA and SspA should regulate expression of the same set of genes . To test this prediction on a genome-wide scale , we performed DNA microarray analyses to compare the global gene expression profiles of ΔmglA mutant cells , ΔsspA mutant cells , LVS wild-type cells , and cells of the control strain LVS::FTL0951 . The results of these analyses show that in LVS , MglA and SspA regulate expression of essentially the same set of genes ( Figure 5; Table 1 ) . Specifically , we found that the expression of 30 genes changed by a factor of 3 or more when mglA was deleted , and the expression of 28 genes changed by a factor of 3 or more when sspA was deleted ( Table 1 ) . Furthermore , of the 33 genes whose expression was altered by a factor of 3 or more in either the ΔmglA or the ΔsspA mutant cells compared to wild-type cells , 25 showed at least a 3-fold change in expression in the other mutant background ( Figure 5; Table 1 ) , and all showed at least a 2-fold change in expression in the other mutant background . Because expression of five of the MglA- or SspA-regulated genes was also altered in cells of our slow growing control strain ( LVS::FTL0951 ) when compared to wild-type cells , we do not necessarily know whether these particular genes are truly regulated by MglA or SspA; their expression might alter in the ΔmglA and ΔsspA mutants simply because these five genes are growth-rate regulated . The results of our microarray analyses thus show not only that MglA and SspA co-regulate expression of essentially the same set of genes in F . tularensis , but also that MglA and SspA exert similar effects on these target genes , as might be expected if MglA and SspA were to function together primarily as a heteromeric complex .
We have found that virulence gene expression is controlled by both MglA and SspA in F . tularensis . Specifically , we have shown that both MglA and SspA are RNAP-associated proteins in F . tularensis , and that MglA and SspA can interact with one another directly to form a heteromeric complex . Although MglA and SspA may also form homomeric complexes , we present evidence that the ability of MglA to associate with RNAP is dependent upon SspA . One implication of this finding is that members of the SspA protein family likely associate with RNAP as an oligomer , and presumably as a dimer [8 , 19] . Our findings suggest that although MglA homomers might form in the cell , they are unlikely to associate with RNAP . Unfortunately , we were unable to determine whether the association of SspA with RNAP is dependent upon MglA , and so we do not currently know whether SspA homomers can associate with RNAP; we were unable to construct a strain of F . tularensis containing a chromosomally TAP-tagged SspA analogous to the LVS MglA-TAP strain , presumably because integration of the sspA-TAP vector into the F . tularensis chromosome results in polar effects on the expression of genes downstream of sspA that are important for growth . In support of the idea that MglA and SspA function together as a heteromeric complex , the results of our microarray analyses show that the MglA and SspA regulons are almost identical . In particular , we show that 30 genes belong to the MglA regulon and 28 belong to the SspA regulon , and that essentially the same set of genes are found in both ( Figure 5; Table 1 ) . In general , the results of our microarray analyses with the ΔmglA mutant strain are in good agreement with those of Brotcke and colleagues , who analyzed the effects of an mglA point mutation on gene expression in F . novicida [6] ( see Table 1 ) . Note that when defining the MglA regulon , we listed those genes whose expression changed by a factor of 3 or more when mglA was deleted . If we had defined the regulon by listing those genes whose expression changed by a factor of 2 or more , then the MglA regulon would be composed of more than 100 genes ( Table S1 ) . Although the molecular details of how MglA exerts its effects on virulence gene expression ( or on the expression of any other target gene ) are not yet known , we have presented evidence that its association with RNAP plays an important , if not essential , role . For example , because expression of the iglA , iglC , and pdpA virulence genes is downregulated in a ΔsspA mutant , and because MglA is present but is not associated with RNAP in a ΔsspA mutant , we disfavor any model whereby MglA upregulates these genes without being associated with RNAP . Accordingly , we suggest two possible models for how MglA , together with SspA , positively controls virulence gene expression in F . tularensis based on association of the MglA–SspA complex with RNAP . The first model specifies that a DNA-bound transcription activator ( s ) contacts the RNAP-associated MglA–SspA complex directly to stimulate transcription initiation at target promoters . Put another way , we hypothesize that the MglA–SspA complex effectively becomes a “subunit” of RNAP that serves as a target for DNA-bound transcription activators that regulate the expression of virulence genes in F . tularensis ( see Figure 6A ) . Precedent for such a model comes from studies of bacteriophage P1 late gene expression . In particular , the P1 late promoter activator ( Lpa or gp10 ) is a phage-encoded , sequence-specific DNA binding protein that is thought to activate P1 late gene expression through a direct interaction with E . coli SspA [11] . In this particular situation , E . coli SspA evidently functions as a co-activator of P1 late gene expression by making simultaneous contact with RNAP and DNA-bound Lpa [11] . In F . tularensis there may be Lpa-like proteins ( at least functionally ) that bind the MglA–SspA complex and are required for the co-activation of MglA- and SspA-dependent virulence gene expression ( see Figure 6A ) . Such MglA- and/or SspA-binding proteins , if they do indeed exist , would be predicted to be DNA-binding proteins that bind the promoter regions of MglA- and SspA-dependent virulence genes . The second model specifies that the RNAP-associated MglA–SspA complex interacts directly with the promoter DNA of MglA- and SspA-dependent virulence genes to activate transcription initiation from the cognate promoters ( Figure 6B ) . However , there is no evidence that any SspA homolog is capable of binding directly to DNA; in particular , E . coli SspA does not appear to bind P1 late promoter DNA [11] . Other models can also be envisioned . These include the MglA–SspA complex modifying the promoter-recognition properties of the associated RNAP holoenzyme without interacting directly with the DNA , or modifying the elongation properties of the associated RNAP . We note that the latter model would be unlikely if the MglA–SspA complex , like SspA in E . coli [8] , could interact only with the RNAP holoenzyme and not with the core enzyme and therefore was not a component of mature elongation complexes . The models discussed above and depicted in Figure 6 are meant only to explain how MglA and SspA positively regulate the expression of target genes . However , MglA and SspA also appear to negatively regulate the expression of at least some target genes ( Table 1; [6] ) . Whether those genes whose expression is negatively controlled by MglA and SspA are controlled directly , or indirectly , remains to be determined . F . tularensis is unique in that it contains two genes , designated rpoA1 and rpoA2 , encoding different α subunits of RNAP . This situation is without precedent; in all other bacterial genomes sequenced to date there is only one rpoA gene . The two α subunits are 32% identical and 55% similar to one another at the primary amino acid level and differ from one another in regions that may be important for dimer formation , promoter recognition , and interaction with DNA-bound transcription activators . We have shown that both α1 and α2 are components of RNAP in F . tularensis . Because α is a dimer in the RNAP complex , there could be as many as four different species of RNAP core enzyme in F . tularensis; one composed of α1 homodimers , one composed of α2 homodimers , and two composed of α1α2 heterodimers ( that differ from one another with respect to which α protomer interacts with the β subunit of RNAP and which α protomer interacts with β′ ) . However , a subset of the predicted dimerization determinants [23] present in the N-terminal domains of α1 and α2 differ from one another , raising the possibility that α1 and α2 might exclusively form either homodimers or heterodimers . The experiments we have performed do not allow us to say anything more than that both α1 and α2 are subunits of RNAP in F . tularensis . Because α participates in promoter recognition through direct sequence-specific protein–DNA interactions [16 , 24] , and because α is a common target for transcription activators [25] , the presence of RNAP species containing distinct α subunits might significantly affect the control of gene expression in F . tularensis . The F . tularensis RNAP complex is also unusual in that it is associated with two different members of the SspA protein family ( MglA and SspA ) . Other bacterial pathogens where SspA homologs play roles in virulence gene control , such as Neisseria gonorrhoeae [12] , Yersinia enterocolitica [13] , and Vibrio cholerae [14] , appear to encode only one SspA homolog . Whether the use of two SspA homologs allows for the integration of multiple environmental signals , or is related to the fact that the F . tularensis RNAP contains two distinct α subunits , or both , remains to be determined .
F . tularensis subspecies holarctica strain LVS was provided by Karen Elkins ( Food and Drug Administration , Rockville , Maryland , United States ) . LVS was grown at 37 °C in modified Mueller Hinton ( MH ) broth ( Difco , http://www . bd . com ) supplemented with glucose ( 0 . 1% ) , ferric pyrophosphate ( 0 . 025% ) , and Isovitalex ( 2% ) , or on cysteine heart agar ( Difco ) medium supplemented with 1% hemoglobin solution ( VWR , http://www . vwrsp . com ) ; when appropriate , kanamycin was used for selection at 5 μg/ml . E . coli strains XL1-blue ( Stratagene , http://www . stratagene . com ) and DH5αF′IQ ( Invitrogen , http://www . invitrogen . com ) were used as recipients for all plasmid constructions . E . coli strain KDZif1ΔZ was used as the reporter strain for the bacterial two-hybrid experiments . KDZif1ΔZ harbors an F′ episome containing the lac promoter derivative placZif1–61 driving expression of a linked lacZ reporter gene and has been described previously [22] . E . coli strain BL21-CodonPlus ( DE3 ) -RP ( Stratagene ) was used for the overproduction of recombinant proteins . The deletion constructs for mglA and sspA were generated by amplifying flanking regions by the PCR and then splicing the flanking regions together by overlap-extension PCR; deletions were in-frame and contained the first three codons and last three codons of the open reading frame ( ORF ) separated by the 9-bp linking sequence 5′–GCGGCCGCC–3′ . The resulting PCR products were cloned into plasmid pEX18km ( provided by Shite Sebastian and Simon Dillon , Harvard Medical School , Boston , Massachusetts , United States ) , yielding pEX-ΔmglA and pEX-ΔsspA . Plasmid pEX18km contains a ColE1 origin of replication ( which is not functional in LVS ) , the Tn903 kanamycin resistance gene ( Epicentre , http://www . epibio . com ) expressed from the LVS groEL promoter , and a sacB gene . Because the sacB gene on pEX18km is insufficient to mediate efficient sucrose counterselection in LVS , a second copy of the sacB gene was subcloned from plasmid pPV [26] on a PstI fragment into plasmids pEX-ΔmglA and pEX-ΔsspA that had been digested with PstI , creating plasmids pEX2-ΔmglA and pEX2-ΔsspA , respectively . Plasmids pEX2-ΔmglA and pEX2-ΔsspA were then used to create strains LVS ΔmglA and LVS ΔsspA by allelic exchange ( see [26] ) . Deletions were confirmed by the PCR and by Southern blotting . Strain LVS::FTL0951 contains a mariner transposon conferring resistance to kanamycin integrated into the LVS chromosome at the FTL_0951 locus; the presence of a single transposon in this strain was confirmed by Southern blotting , and the location of the transposon was determined by sequencing of the chromosomal DNA . Complementation vectors: Plasmid pF-MglA was used for complementation of the LVS ΔmglA mutant strain and was constructed by inserting full-length mglA in place of gfp in the plasmid pFNLTP6 gro-gfp [27] . Plasmid pF was used as a control for complementation experiments and was constructed by deleting the gfp gene from plasmid pFNLTP6 gro-gfp [27] . Plasmid pF2-SspA was used for complementation of the LVS ΔsspA mutant strain and was constructed in two steps . First , full-length sspA was inserted into pFNLTP6 gro-gfp [27] in place of the gfp gene creating plasmid pF-SspA . Second , to compensate for the apparent deleterious effect of overexpressing sspA in F . tularensis , the full-length LVS groEL promoter in pF-SspA was replaced with a more minimal groEL promoter lacking the putative UP-element . As a control , plasmid pF2 was constructed by replacing the full-length LVS groEL promoter in pF with the same minimal groEL promoter used in plasmid pF2-SspA . Plasmids and strains for TAP-tag experiments: Plasmids for generating TAP-tag strains were constructed by inserting the last ~400 bp of the ORF into the vector pEXTAP ( this work; Figure 1A ) , which contains the TAP-tag sequence from plasmid pP30ΔYTAP [22] cloned into the multiple cloning site of pEX18km . Specifically , plasmid pEX-MglA-TAP was made by cloning an ~400-bp fragment of DNA corresponding to a 3′ portion of the mglA gene into pEXTAP such that the mglA portion was in-frame with the DNA specifying the TAP-tag . Strain LVS MglA-TAP was constructed by electroporating pEX-MglA-TAP into LVS ( essentially as described in [27] ) and selected on cysteine heart agar medium containing hemoglobin ( 1% ) and kanamycin ( 5 μg/ml ) . Because pEX-MglA-TAP cannot replicate in F . tularensis , only those cells in which the plasmid has integrated into the chromosome can grow on media containing kanamycin . Strains LVS SucD-TAP and LVS β′-TAP were made in a similar fashion using plasmids pEX-SucD-TAP and pEX-RpoC-TAP , respectively . Strain LVS ΔsspA MglA-TAP was made using plasmid pEX-MglA-TAP together with strain LVS ΔsspA . For all strains , insertion of the TAP integration vector at the correct chromosomal location was confirmed by the PCR , and production of the corresponding TAP-tagged protein was confirmed by Western blotting with PAP [28] , which binds the ProtA moieties of the TAP-tag . Plasmids for bacterial two-hybrid assays: Plasmid pBRGPω directs the synthesis of the Gal11P-ω fusion protein and has been described before [22] . Plasmid pBRGPω can be used to create fusions to the N-terminus of the ω subunit of E . coli RNAP; proteins are fused to ω via a small linker composed of three alanine residues specified in part by a NotI restriction site . Plasmid pBRGPω confers resistance to carbenicillin , harbors a ColE1 origin of replication , and carries an IPTG-inducible lacUV5 promoter that drives expression of the ω fusion protein [22] . Plasmids pBRMglA-ω and pBRSspA-ω direct the synthesis of either full-length MglA or full-length LVS SspA fused to residues 2–90 of E . coli ω . The pBR-ω fusion plasmids were made by cloning the appropriate NdeI-NotI–digested PCR products into NdeI-NotI–digested pBRGPω . Each of the respective ω fusion proteins is therefore under the control of an IPTG-inducible lacUV5 promoter . Plasmid pACTR-AP-Zif directs the synthesis of Zif , the zinc finger DNA-binding domain of the murine Zif268 protein , and has been described previously [22] . This plasmid can be used to create fusions to the N-terminus of Zif; proteins are fused to Zif via a nine–amino acid linker ( Ala-Ala-Ala-Pro-Arg-Val-Arg-Thr-Gly ) specified in part by a NotI restriction site . Plasmid pACTR-AP-Zif confers resistance to tetracycline and harbors a p15A origin of replication . Plasmids pACTR-MglA-Zif and pACTR-SspA-Zif direct the synthesis of either full-length MglA ( residues 1–205 ) or full-length LVS SspA ( residues 1–210 ) fused to Zif . The pACTR-Zif fusion plasmids were made by cloning the appropriate NdeI-NotI–digested PCR products into NdeI-NotI–digested pACTR-AP-Zif . Each of the respective Zif fusion proteins is therefore under the control of an IPTG-inducible lacUV5 promoter . Cells were grown at 37 °C with aeration in 200 ml of MH broth supplemented with glucose ( 0 . 1% ) , ferric pyrophosphate ( 0 . 025% ) , and Isovitalex ( 2% ) . Cells were grown to an OD600 of ~0 . 4 and harvested by centrifugation at 4 °C . TAP was then performed essentially as described earlier [18] . Cells were grown with aeration at 37 °C in LB supplemented with kanamycin ( 50 μg/ml ) , carbenicillin ( 100 μg/ml ) , tetracycline ( 10 μg/ml ) , and IPTG at the concentration indicated . Cells were permeabilized with SDS-CHCl3 and assayed for β-galactosidase activity as described previously [29] . Assays were performed at least three times in duplicate on separate occasions . Representative data sets are shown . Values are averages based on one experiment; duplicate measurements differed by less than 10% . Plasmid pETDuetMglA-S directs the synthesis of MglA with a C-terminal S-tag ( MglA-S ) and confers resistance to carbenicillin , while plasmid pRSFDuetSspA-His6 directs the synthesis of SspA with a C-terminal hexa-histidine tag ( SspA-His6 ) and confers resistance to kanamycin . Plasmid pRSFDuetSspA directs the synthesis of wild-type SspA and confers resistance to kanamycin . Plasmid pETDuetMglA-S was introduced into E . coli strain BL21-CodonPlus ( DE3 ) -RP ( Stratagene ) along with pRSFDuetSspA-His6 or pRSFDuetSspA . Single colonies were used to inoculate Overnight Express Instant TB medium ( Novagen , http://www . emdbiosciences . com/html/NVG/home . html ) supplemented with kanamycin ( 50 μg/ml ) , carbenicillin ( 100 μg/ml ) , and chloramphenicol ( 25 μg/ml ) , and cultures were grown overnight at 30 °C with aeration to allow for induction . Cell pellets from 50 ml of overnight culture ( OD600 of ~3 . 0 ) were resuspended in 1 ml of binding buffer ( 10 mM imidazole; 10 mM Tris-HCl [pH 8 . 0]; 300 mM NaCl; 10% glycerol; and 1X EDTA-free protease inhibitor cocktail [Roche , http://www . roche . com] ) . Each 1-ml suspension was separated into 500 μl aliquots , and 100 μl of lysozyme ( 10 mg/ml ) was added to each aliquot . Cells were then incubated on ice for 1 h and lysed by sonication , and cell debris was removed from each lysate by centrifugation . Lysates were then added to TALON polyhistidine-tag purification resin ( Clontech , http://www . clontech . com ) , and following washing with binding buffer , proteins specifically bound to the resin were eluted in buffer containing 250 mM imidazole , 10 mM Tris-HCl ( pH 8 . 0 ) , 300 mM NaCl , 10% glycerol , and 1X EDTA-free protease inhibitor cocktail . Cell lysates and eluted proteins were separated by SDS-PAGE on 4%–12% Bis-Tris NuPAGE gels in MOPS running buffer ( Invitrogen ) and transferred to nitrocellulose using the iBlot dry blotting system ( Invitrogen ) . Membranes were blocked with 25 ml of SuperBlock Blocking Buffer ( Pierce , http://www . piercenet . com ) in TBS with 250 μl Surfact-Amps 20 ( Pierce ) . Membranes were then probed with monoclonal antibodies against the S-Tag ( Novagen ) , the His-Tag ( Novagen ) , or the α subunit of E . coli RNAP ( NeoClone , http://www . neoclone . com ) . Proteins were detected using goat polyclonal anti-mouse IgG conjugated with horseradish peroxidase ( Pierce ) , and visualized using SuperSignal West Pico Chemiluminescent Substrate ( Pierce ) . Cells were grown with aeration at 37 °C in MH broth supplemented with glucose ( 0 . 1% ) , ferric pyrophosphate ( 0 . 025% ) , and Isovitalex ( 2% ) . RNA isolation and cDNA synthesis were essentially as described previously for Pseudomonas aeruginosa [30] . Transcript quantities for iglA , iglC , and pdpA were determined relative to the amount of tul4 transcript by RT-PCR with the iTaq SYBR Green kit ( Bio-Rad , http://www . bio-rad . com ) and an ABI Prism 7000 Sequence Detection System ( Applied Biosystems , http://www . appliedbiosystems . com ) ; the REST 2005 software package was used for statistical analysis of the data [31] . ORFs for LVS were predicted using GLIMMER [32] and annotated by comparison with the annotated Schu S4 genome sequence using BLAST [15] . ORFs were manually annotated guided by BLAST results but with appropriate alterations . For example , ORFs containing sequencing errors were identified as being those where adjacent genes were fully contained within predicted Schu S4 ORFs . The DNA microarray consisted of sequence fragments constructed by PCR . Primers for PCR were designed using PRIMER3 ( [33]; http://primer3 . sourceforge . net ) software using the genome annotation . The 1 , 932 primer sets produced amplicons of 42 bp to 400 bp in length centrally located within each gene . They correspond to sequences for 1 , 678 LVS ORFs with Schu S4 homologs ( including the 15 duplicated ORFs in the pathogenicity island ) , 24 GLIMMER-predicted ORFs with no homology to Schu S4 , and 230 intergenic regions . Greater than 95% of LVS non-transposase genes were represented on the microarray . The primer sets contained 5′ extensions to each forward and reverse primer , GGCATCTAGAGCAC and CCGCACTAGTCCTC , respectively . All amplifications were performed using a standard protocol . In the first round of amplification we used 20-μl PCR reactions containing 1 ng of LVS genomic DNA as template , 20 pmol of unique primers , and 5 units of Taq polymerase ( Takara , http://www . takarabiousa . com ) . These reactions were then diluted 1:150 . A second round contained 1 μl of the diluted first reaction as template , 20 pmol of universal primers containing 5′ amino modification including a 3-carbon linker ( GAACCGATAGGCATCTAGAGCAC and GAAATCCACCGCACTAGTCCTC ) , and 5 units of Taq polymerase in a total volume of 100 μl . Aliquots ( 5 μl ) of each first and second round reaction were run on 2% agarose gels to confirm the presence of appropriately sized products . The PCR products were purified using 96-well QiaQuick Multiwell PCR Purification plates ( Qiagen , http://www . qiagen . com ) , eluted in deionized water , and transferred to 384-well plates . Purified products were air-dried in a biosafety cabinet at room temperature and resuspended in 50 mM sodium phosphate ( pH 8 . 5 ) . Each PCR amplicon was printed in duplicate onto CodeLink Activated Slides ( Amersham , http://www . amersham . com ) using a Qarrayer ( Genetix , http://www . genetix . com ) . The microarrays were then post-processed according to the slide manufacturer's instructions . Strains were grown to mid-log in supplemented MH medium and RNA was isolated from 15 ml of culture using Qiagen RNeasy Midi columns . Samples were DNase-treated twice: on the columns using a Qiagen RNase-Free DNase Set , and after purification using RQ1 DNase ( Promega , http://www . promega . com ) . RNA was purified from three separate cultures for each strain . cDNA was synthesized using ~10 μg RNA , 750 ng NSNSNSNSNS random primers ( where N refers to any base and S refers to G or C ) , and Superscript II Reverse Transcriptase ( Invitrogen ) . Amino acyl-dUTP ( aa-dUTP ) was incorporated into the cDNA by including a 2:1 ratio of aa-dUTP to dTTP in the reverse transcription reaction . RNA was removed by increasing the pH with NaOH , and then unincorporated aa-dUTP and free amines were removed by purifying the cDNA with Qiagen MinElute PCR Purification columns . Eluted cDNA was dried to completion in a speed-vac and then labeled with dye; in a 9-μl volume , wild-type LVS cDNA was labeled with Cy5 dye ( Amersham; Cy5 mono-Reactive Dye Pack ) in 0 . 1 M sodium carbonate ( pH 9 . 3 ) , and cDNA from LVS ΔmglA , LVS ΔsspA , and LVS::FTL0951 were similarly labeled with Cy3 dye ( Amersham; Cy3 mono-Reactive Dye Pack ) . After coupling , 35 μl of 100 mM NaOAc ( pH 5 . 2 ) was added to each reaction and free dye was removed using Qiagen MinElute PCR purification columns . cDNA equivalent to ~600 pmol Cy5 dye was mixed with cDNA equivalent to ~600 pmol Cy3 dye from either LVS ΔmglA , LVS ΔsspA , or LVS::FTL0951 and then dried to completion in a speed-vac . The labeled probes were resuspended in hybridization buffer ( 5X SSC , 50% formamide , 0 . 1% SDS , 0 . 2 mg/ml tRNA ) , and applied to arrays using a Tecan HS400 Hybridization Station . Arrays were scanned using a GenePix 4000B microarray scanner ( Molecular Devices , http://www . moleculardevices . com ) and data were analyzed with GeneSpring GX .
Proteins discussed in this study , followed by their corresponding locus tag and Entrez Protein ( http://www . ncbi . nlm . nih . gov/entrez/query . fcgi ? db=Protein ) accession numbers are as follows: α1 ( FTL_0261; YP_513056 . 1 ) , α2 ( FTL_0616; YP_513375 . 1 ) , MglA ( FTL_1185; YP_513868 ) , and SspA ( FTL_1606; YP_514245 . 1 ) . The Entrez Genome ( http://www . ncbi . nlm . nih . gov/entrez/query . fcgi ? db=Genome ) accession number for the nucleotide sequence of the F . tularensis LVS genome is NC_007880 . 1 .
|
The Gram-negative bacterium Francisella tularensis is an intracellular pathogen and the causative agent of tularemia . In F . tularensis , the MglA protein is the only known regulator of virulence genes that are important for intracellular survival , yet it is not known how MglA functions . F . tularensis also contains an MglA-like protein called SspA whose function is not known . In this study , we show that both MglA and SspA associate with RNA polymerase ( RNAP ) and positively regulate virulence gene expression in F . tularensis . Our study provides evidence that MglA and SspA interact with one another directly and that the association of MglA with RNAP is dependent on the presence of SspA . We also show that , unlike RNAP from any other bacterium , RNAP from F . tularensis contains two distinct α subunits . Given the fundamental roles the α subunit plays in transcription regulation , this may have far-reaching implications for how gene expression is controlled in F . tularensis . Our study therefore uncovers a new critical regulator of virulence gene expression in F . tularensis ( SspA ) , provides mechanistic insight into how MglA and SspA cooperate to control virulence gene expression , and reveals that the F . tularensis transcription machinery has an unusual composition .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"biochemistry",
"infectious",
"diseases",
"microbiology",
"molecular",
"biology",
"eubacteria"
] |
2007
|
Twin RNA Polymerase–Associated Proteins Control Virulence Gene Expression in Francisella tularensis
|
There is a revolution in the ability to analyze gene expression of single cells in a tissue . To understand this data we must comprehend how cells are distributed in a high-dimensional gene expression space . One open question is whether cell types form discrete clusters or whether gene expression forms a continuum of states . If such a continuum exists , what is its geometry ? Recent theory on evolutionary trade-offs suggests that cells that need to perform multiple tasks are arranged in a polygon or polyhedron ( line , triangle , tetrahedron and so on , generally called polytopes ) in gene expression space , whose vertices are the expression profiles optimal for each task . Here , we analyze single-cell data from human and mouse tissues profiled using a variety of single-cell technologies . We fit the data to shapes with different numbers of vertices , compute their statistical significance , and infer their tasks . We find cases in which single cells fill out a continuum of expression states within a polyhedron . This occurs in intestinal progenitor cells , which fill out a tetrahedron in gene expression space . The four vertices of this tetrahedron are each enriched with genes for a specific task related to stemness and early differentiation . A polyhedral continuum of states is also found in spleen dendritic cells , known to perform multiple immune tasks: cells fill out a tetrahedron whose vertices correspond to key tasks related to maturation , pathogen sensing and communication with lymphocytes . A mixture of continuum-like distributions and discrete clusters is found in other cell types , including bone marrow and differentiated intestinal crypt cells . This approach can be used to understand the geometry and biological tasks of a wide range of single-cell datasets . The present results suggest that the concept of cell type may be expanded . In addition to discreet clusters in gene-expression space , we suggest a new possibility: a continuum of states within a polyhedron , in which the vertices represent specialists at key tasks .
Recent advances allow high-throughput measurement of biological information from individual cells [1–12] . This is an improvement over standard experiments which mask the range of states in the population because they average over millions of cells . Therefore , it is expected that single-cell technologies can reveal new biology , such as the diversity of states of cells in a tissue [13–21] . These experiments portray each cell as a point in a high-dimensional space whose axes are the expression level of each gene , or other molecular parameters . The geometry of how cells are distributed in gene expression space is an open question . One possibility is that each cell type forms a tight cluster , and that these clusters are well separated from each other . This assumption is at the heart of clustering analyses of gene expression data [22 , 23] . The tight-cluster picture relates to the idea of discrete cell types , which is supported by the existence of marker genes that are mutually exclusive between cells . When considering a set of many genes , in contrast to only marker genes , it is possible that cells also form more continuous distributions in gene expression space , and that clusters are more difficult to define . Such distributed states have been suggested in studies on T-cells [24–26] and embryonic stem cells [27] . In such cases , an open question is whether there is meaningful geometry to this continuum of cell states . The question of geometry in gene expression space was recently addressed in the context of a theory on evolutionary tradeoffs [28] . The theory suggests that cells that need to perform multiple tasks are arranged in a simple , low dimensional polygons or polyhedra in gene expression space . The vertices of these shapes , called archetypes , are the expression profiles optimal for each task . Thus , two tasks correspond to data on a line , three tasks to data on a triangle , four tasks to a tetrahedron , and so on . These shapes are generally called polytopes: the generalization of polygons and polyhedral to any number of dimensions . These polytopes represent the optimal tradeoffs between the tasks , in the sense that for any point outside the polytope there is a point inside it that performs equally or better at all tasks . This corresponds to the concept of Pareto optimality , where the polytopes are Pareto fronts for the system [28–32] . Pareto optimality has been applied to a variety of biological datasets [30 , 33–36] , but not to single-cell data . Here , we study single-cell expression from several tissues , collected with different single-cell technologies ( Fig 1 ) . We analyze single-cell qPCR data on human and mouse colonic crypts from Dalerba et al . and Rothenberg et al . [2 , 37] , single-cell mass cytometry data on individual human bone marrow cells from Bendall et al [13 , 25] , and single-cell RNA-Seq on mouse dendritic cells from Jaitin et al [3] . We test whether each dataset is well-described by a low dimensional polytope . We fit the data to a series of polytopes ( line , triangle , tetrahedron , etc . ) , finding the best fit polytope , and assess its statistical significance . We then analyze the gene expression profiles at the cells closest to the vertices ( archetypes ) , to test if they correspond to specific biological tasks . This offers a way to discover potential tasks of cells in a tissue from single-cell data . We find evidence for polytopes and enriched tasks in all of the datasets we analyzed . The well-studied human and mouse colonic crypt system [38 , 39] includes stem cells at the bottom of the crypt , which differentiate into enterocytes that absorb nutrients , and secretory cells , mainly goblet cells , which secrete mucus . We find that single human intestinal cells are arranged in gene expression space , to a good approximation , in a tetrahedron . At its vertices are expression profiles of pure cell types—enterocytes , goblet cells , putative stem cells and a new category of nodal-expressing cells . We further find that when analyzing the intestinal progenitor cells alone , they fill out a distinct tetrahedron in gene expression space , in contrast to other cell types . The vertices of this tetrahedron correspond to four progenitor cell tasks . In addition , we find that the polytopes found in crypt data from human and mice are strikingly similar . Using the same approach , we find that human bone marrow cell data is arranged in a five-vertex simplex ( four-dimensional simplex ) whose vertices correspond to five major cell types , and that mouse dendritic cells uniformly fill a tetrahedron suggesting four immune tasks including maturation , pathogen sensing and communication with lymphocytes . Pareto analysis can thus be useful to understand the geometry of single-cell gene expression , and to infer the tasks of single cells in a tissue .
We begin with analyzing the single-cell gene expression dataset of human colon crypt cells obtained by Dalerba et al [2] . The dataset included 407 individual cells , each analyzed by single-cell qPCR for 83 selected genes in a Fluidigm microfluidic system [5] . We normalized the data by subtracting the mean of each gene as described in Methods . To address the effects of outliers and all/none bimodality in expression data ( [11 , 12] ) , we removed 10% of the cells which had the lowest expression levels across genes , and 10% of the genes which had the lowest expression across cells , resulting in a dataset of 368 cells and 76 genes ( S1 Dataset ) . Removing more cells or genes , up to 25% of the lowest expressing cells or lowest expressed genes , leaves the results essentially unchanged ( S1 Fig and S1A Text ) . Based on theory on evolutionary tradeoffs between tasks [28] , we expect that cells should fall in a low-dimensional polytope whose vertices are the points optimal in each task alone . We therefore asked whether gene expression data is enclosed within a low dimensional polytope ( e . g . line , triangle , tetrahedron and so on ) . We calculated the best fit polytopes that enclose the data , and asked how well they describe the data compared to randomized datasets . We considered polytopes with k vertices: we tested k = 2 ( line ) , k = 3 ( triangle ) , k = 4 ( tetrahedron ) and so on up to 11 vertices . The polytopes were found using the PCHA algorithm [40] . This algorithm seeks k points on the convex hull of the data that define a polytope that encloses as much of the data as possible ( see Methods: Archetype detection using the PCHA algorithm [40] ) . For each polytope , we calculated the deviation of the data from the polytope ( explained variance , Methods: Determining the number of archetypes ) . k = 4 archetypes explain 45% of the variance , whereas adding a fifth archetype added only 4% additional explained variance ( Fig 2a ) . A tetrahedron explained the data variance better than when applying the same algorithm to shuffled data ( p = 0 . 01 , Methods: Statistical significance of best fit polytopes ) . This suggests that a 4-vertex polyhedron , namely a tetrahedron , is a reasonable description of the data . The tetrahedron was found in the full 76-dimensional space . To display the data , it is helpful to use principal component analysis ( PCA ) . PCA indicated that 3 principal components ( 3 PCs ) explain most of the variance in the data ( 47% ) , and that 4 PCs ( 52% ) are significant compared to shuffled data ( Fig 2b , Methods: Determining the number of archetypes ) , indicating that the data is indeed low-dimensional . The tetrahedrality of the data is highlighted by the fact that 96% of the variance explained by the first 3PCs is explained also by the much more stringent description of a tetrahedron whose vertices are on the convex hull of the data . Plotting the data , with each cell represented by a point in the space spanned by the first 3PCs of gene expression space , suggests a tetrahedron-like shape ( Fig 2c ) . The projections of the data on the three principal planes are roughly triangular ( Fig 2d–2f ) , and show well-defined linear edges which meet at pointy vertices . We found that the archetype positions were robust to data sampling , with errors on the order of a few percent in bootstrapping tests in which data is resampled with replacement ( S1B Text and S1a Table and Fig 2c ) . We note that clustering methods , such as k-means or hierarchical clustering , are much more sensitive to sampling errors in this dataset: the continuous distribution of the data makes the cluster boundaries somewhat arbitrary and thus on the order of 20% of the data points are classified to different clusters upon bootstrapping ( S1C Text and S2–S4 Figs ) . Each of the four vertices of the best-fit tetrahedron is a point in the 76-dimensional gene expression space . Within Pareto theory , each vertex is an archetype that can be thought of as an optimal gene expression profile , extrapolated from the data , which best performs the archetype's task . The gene profiles for the archetypes are shown in Fig 3a . We find that each of the four archetypes shows high expression of a set of markers for a specific crypt cell type ( Fig 3b and 3c and S5 Fig ) . Archetype 1 shows high expression of enterocyte markers ( AQP8 , SLC26A3 , MS4A12 , KRT20 [41–44] ) . The cells closest to this archetype show the maximal expression of these markers in the entire dataset ( Methods: 1D Gene enrichment at archetypes ) . In cells closest to archetype 2 , putative stem cells markers are maximally expressed ( including LGR5 , ASCL2 , AXIN2 , c-MYC , CDK6 , OLFM4 [45–49] ) . In cells closest to archetype 3 , markers for goblet cells are maximally expressed ( MUC2 , TFF3 , SPDEF and others [38 , 50 , 51] ) . The complete list of enriched genes is shown in S2 Table . Thus , the first three archetypes correspond each to one of the three main crypt cell types . The fourth archetype is discussed below . We also evaluated the physical position of each cell in the crypt using a proxy for height in the crypt , Axin2 expression [52] . We find that the stem cell archetype has highest Axin2 , followed by the goblet archetype . The enterocyte archetype has lowest Axin2 . This matches the known arrangement of cells in the crypt , with stem cells at the bottom of the crypt , and enterocytes at its top ( S6a Fig ) [37] . The fourth archetype is enriched with a specific set of genes related to development and embryonic patterning ( NODAL , CFC1 , TDGF1 [53–56] ) , a transcriptional repressor that has a role in development ( PCGF6 [57] ) , and an enzyme involved in hormone secretion ( UGT1A1 [58] ) , and a member of the claudin family CLDN8 [59] . We call these cells Nodal cells . Their position in the crypt , based on Axin2 levels , is intermediate between the bottom and top ( S6a Fig ) . To better understand Nodal cells , we ordered the cells according to a pseudo-temporal order inferred from their gene expression using the Wanderlust algorithm [60] . NODAL cells seem to lie on the developmental axis between stem cells and enterocytes ( S7a Fig ) . This tentatively suggests NODAL cells as a differentiation step between precursors and enterocytes . In summary , Pareto analysis finds four archetypes which correspond to gene expression profiles . Three of these define the three main cell types in the crypt . The fourth may indicate a step between stem cells and enterocytes . The archetypes can be interpreted as an idealized gene profile for each of the cell types , and cells of a given type are arranged in proximity to the corresponding archetype in gene expression space . We next zoom in on each inferred cell type to examine the variation between cells within a type . We repeated the Pareto analysis on each class of cells separately ( S1D Text ) . We find that the three non-progenitor cell types ( enterocytes , goblet cells , nodal cells ) cannot be explained by a statistically significant polytope with 5 or less vertices ( see S3 Table , all p-values >0 . 15 ) . Thus , the expression of these cells seems to form a cloud in gene expression space with no easily discernible vertices ( Fig 4a ) . This may hint that other effects dominate the structure of the data , or that not enough cells or not enough relevant genes were measured , see S1E Text and S8 Fig . In contrast , the progenitor cells were well-described by a tetrahedron ( p = 0 . 01 , Methods: Statistical significance of best fit polytopes , Fig 4b showing tetrahedron and projections , Fig 4c shows explained variance curves , S2 Dataset ) . The progenitor cells seem to uniformly fill out this tetrahedron , suggesting that precursors span a continuum of gene expression states , with some cells coming close to one of four archetypal precursor profiles , and others showing a more even mixture of the archetypes . We examined the expression profiles of the cells closest to each archetype ( Methods: 1D Gene enrichment at archetypes ) . Archetype 1 is enriched with stem markers LGR5 and ASCL2 , and lies physically at the lowest point in the crypt according to Axin2 levels ( S7b Fig inset ) . It has proliferation markers ( MYC , Cdk6 ) , and telomerase ( TERT ) . These genes are characteristic of dividing stem cells [37 , 46 , 48 , 49] . The other three archetypes all display the progenitor marker OLFM4 [61] , but also have characteristics more similar to the differentiated cells . Archetype 2 includes enrichment in enterocyte and goblet markers . Archetype 3 is enriched in division inhibitor CDKN1A [62] . Archetype 4 has low expression of all genes . We hypothesize that these three archetypes represent three tasks needed for differentiation: ( i ) expression of effector cell-specific genes ( ii ) inhibition of cell division ( iii ) reduction in global gene expression . For a complete list of enriched genes see S4 Table , archetype gene expression profiles are shown in S9 Fig . The progenitor cells fill out the tetrahedron quite uniformly . According to Axin2 expression , as they move up the crypt they move away from the stem archetype and parallel to the plane defined by the other three archetypes . Pseudo-temporal order derived by the Wanderlust algorithm suggests similar conclusion ( S7b Fig ) . This may suggest multiple temporal paths between the three tasks , such that each progenitor is a different weighted average of the archetypes . We also asked how the progenitor tetrahedron relates to the rest of the cells in the crypt . We find that the fully differentiated cells types ( enterocytes , goblet cells ) are closest to archetype 4 , because they all have lower overall gene expression than the progenitors . Up to now we analyzed a human crypt dataset . We now compare it to a mouse crypt dataset , presented by Rothenberg et al . [37] using qPCR ( S3 Dataset ) . We used the same Pareto analysis approach . We removed the lowest 10% of cells and genes , remaining with 161 cells and 41 genes . The two datasets overlap in 24 genes . The mouse cells in the dataset were only from the bottom of the crypt: they were harvested by cell sorting ( FACS ) using the markers CD66 and CD44 . We find that the mouse single-cell data is well-described by a triangle ( p = 0 . 003 , 2PCs explain 43% of the variance ) . To compare the mouse data to human data described in the previous sections , we analyzed human cells in the lower part of the crypt ( defined in [2] , by FACS sort for high and low cells using CD66 and CD44 markers ) . This dataset ( 213 cells ) is also well described by a triangle ( p = 0 . 001 , 2PCs describe 35% variance ) . The archetypes in both the datasets correspond to stem cells , goblet cells and enterocytes . The nodal cells are on the edge connecting progenitors and enterocytes ( Fig 5b ) , representing the projection of the nodal tetrahedral archetype on the triangle that describes the crypt-bottom cells . In the mouse dataset 5 out of the 6 genes that are enriched in the Nodal archetype were not measured; however the Nodal co-receptor TDGF1 [27] is highly expressed on the edge that connects stem cells and enterocytes . The geometry of mouse and human datasets are strikingly similar ( Fig 5 ) . The three mouse archetypes are very similar to the three corresponding human archetypes in their overlapping genes ( R2 = 0 . 65 , p<10–9 , S10 Fig , enriched genes are shown in Fig 5 ) , and signify stem cells , goblet cells and enterocytes . Moreover , the distribution of points on the triangle suggests a gradual differentiation process from stem cells to enterocytes , compared to more abrupt temporal switch in the case of differentiation to goblets , a difference shared by both species . We also analyzed single-cell data by Dalerba et al [2] for human colon cancer xenograft in mouse , derived from a single cancer cell ( S11 Fig and S1F Text and S4 Dataset ) . We compared this data to the triangle found when analyzing human crypt-bottom normal tissue cells ( Fig 5b ) . We find that the cancer cells lie in a similar triangle , with a density distribution that peaks near the stem cells , enterocyte and goblet archetypes . This hints that the human cancer cells undergo differentiation similar to the normal mouse and human tissues [2] . Finally , we asked whether this approach can be used with other experimental techniques and other tissue types . We studied a single-cell mass cytometry ( Cytof ) dataset on human bone marrow by Bendall et al [13 , 25] ( S5 Dataset ) . Here , 10 , 000 cells were each characterized by the expression of 31 proteins detected using antibodies . We find that this dataset is well-described by a four-dimensional simplex ( a polytope with five vertices , p = 0 . 005 , Fig 6 top row and S12 Fig ) . The five archetypes are each enriched with genes that clearly define a specific cell type ( CD4 T cells , CD8 T cells , monocytes/ macrophages , B cells and non-leukocytes ) ( see S1G Text and S5 Table and S13 and S14 Figs ) . Cells density peaks near each vertex . A sizable set of cells , which formed an unidentified cluster in the viSNE analysis of [13] , lies in the middle of the tetrahedron ( S15 Fig ) , possibly indicating cells whose protein expression is intermediate between the classical cell type profiles . We also analyzed a single-cell RNA-Seq dataset from mouse spleen by Jaitin et al [3] ( S6 Dataset ) . Each cell is characterized by 20 , 091 gene expression counts , based on sampling a fraction of the cell’s mRNA pool . This data was classified by Jaitin et al into seven groups of cells using a probabilistic mixture model . One group of cells , however , seemed to defy clear clustering ( class VII in [3] ) . These are the dendritic cells in the spleen , known to carry out a wide range of immune functions including detection of pathogens and activation of lymphocytes [63–65] . We find that LPS treated dendritic cells in this dataset are well-described by a tetrahedron ( 312 cells , p<0 . 001 , Fig 6 bottom row ) . Several functional gene groups [66] ( see S1H Text ) are highly and maximally enriched in the cells closest to each of the four archetypes ( S6 Table ) . This enrichment indicates four key immune tasks: Archetype 1- response to virus ( cytoplasmic DNA response and interferon pathways ) [7 , 67]; Archetype 2- Dendritic cell maturation and formation of cytoskeletal features [64 , 68 , 69]; Archetype 3- Stimulation of lymphocytes ( cytokine secretion , antigen presentation ) [7 , 70 , 71]; Archetype 4- putative cell-death pathways [72 , 73] . We compared the tissues in terms of how uniformly they fill out their polyhedra or polytopes . We therefore computed the ratio ρ between the mean local density of the data and the local density expected in a uniform distribution [74] ( S1I Text and S16 Fig ) . The closer this ratio is to one , the more uniform the distribution of points; a high ratio means the data is clumped into clusters . We find that bone marrow cells are the most clustered or clumped among the datasets ( ρ = 4 . 67 ) , in line with the classic view of hierarchical hematopoietic lineages [75 , 76] . Intestinal cells are less clustered ( ρ = 2 . 93 ) . The closest to uniform distribution inside their respective polyhedra are dendritic cells ( ρ = 2 . 42 ) and intestinal progenitors ( ρ = 1 . 06 ) . We conclude that the present approach can describe the geometry and potential tasks of single-cell data from diverse tissues and different technologies .
We studied the geometry of single cells in gene expression space using Pareto archetype analysis . We used data from different single-cell technologies employing qPCR , RNA-Seq or mass cytometry , and different tissues including intestinal crypt , bone marrow and spleen . We find that single-cell data fall in low dimensional shapes with well-defined vertices , such as tetrahedrons . The cells closest to each vertex are enriched with genes that reveal key biological functions relevant to each tissue . Some datasets fall into distinct clusters , with one cluster near each vertex , and thus support a picture of distinct and well-separated cell types . Other contexts , such as intestinal progenitor cells and spleen dendritic cells , show a continuum of gene expression states which uniformly fills the tetrahedron , supporting a picture of a continuous range of cell states that carry out mixtures of the biological functions defined by the vertices . These findings expand the concept of cell type , by demonstrating the possibility of a polyhedral continuum of expression states: cells can range between being task specialists near the vertices of the polyhedron , and generalists suitable for multiple tasks near the center . It is interesting to ask when is it better to design distinct cell types with separated biological functions , and when to design a continuum of cell expression states . Distinct cell types have the advantage of being specialists at a given task , with optimal function . However , if the proportions of the tasks needed in the tissue changes more rapidly than the ability to make new cells or to adjust their protein composition , a continuum of states may have an advantage . It allows a reserve of cells ( cells in the middle of the polyhedron ) that can perform multiple functions , albeit less optimally than specialists , and can therefore be recruited to each task in times of need . This type of reasoning was used to explain why ant morphologies in a nest tend to show a continuous distribution rather than distinct clusters: Intermediate morphology ants can be recruited to defense , foraging or nursing tasks according to changing needs [77] . Other factors that may influence clustering versus continuum include the biochemical feasibility of multiple functions to co-exist in the same cell [78 , 79] and the existence of a continuous range of spatial and temporal niches in a tissue related for example to migration , differentiation processes [17 , 80] , or to distances from blood vessels or tissue boundaries [81] . We find a continuous filling of a tetrahedron in the context of progenitor cells in the intestine . The progenitor tetrahedron suggests four tasks—one related to stemness and stem cell renewal , and the other three archetypes related to potential tasks required for early stages of differentiation: stopping division , expressing effector genes , and down regulating global expression [82] . The uniform filling of the progenitor tetrahedron suggests that there is not one temporal path to differentiation , but rather many paths with different ordering of the tasks , each taken with more or less equal probability . This relates to the idea that stem cells show heterogeneity [15 , 27 , 83] , in which different molecular states confer functional biases to individual cells , contributing to their overall regulation [84] , and aligns with recent findings about their plasticity [85 , 86] . With more data , one may be able to infer temporal paths from static data , as was done in the context of the cell cycle [87] and cell differentiation [17 , 60] . Similarly , the continuum observed between intestinal progenitors , nodal cells and enterocytes ( Figs 2c and 5b ) suggests a gradual differentiation process , with the nodal cells—a new class of cells defined in the present study—possibly an intermediate station between progenitors and enterocytes . The polyhedral continuum of states we find in stimulated dendritic cells ( DCs ) may likewise suggest spatiotemporal trajectories in the spleen , with external DCs active in pathogen detection , followed by migration into the central spleen for lymphocyte activation [64 , 65] . The present Pareto analysis is a new way of looking at single-cell data that emphasizes the geometric contours that enclose the data . This approach was used recently also in other biological contexts , to understand C . elegans foraging behavior [34] , bird toe-bone proportions [35] , and bacterial [88] and tumor [33] population-level gene expression . It is useful to compare the present approach to other methods of analyzing high-dimensional data , such as clustering or self-organizing maps [88–92] . If the data is arranged in separated non-overlapping clouds , all methods can reveal its structure . If the data , in contrast , is spread more along a continuum , clustering methods can lead to arbitrary classification , because it is not possible to tell where one cluster ends and another begins . For example , in the current crypt dataset , cluster assignment for 20% of the crypt cells changed upon data resampling with returns ( bootstrapping , see S1C Text ) . In contrast , archetypes varied by only a few percent upon bootstrapping , and data description as a weighted average of archetypes was likewise much less sensitive to data resampling than clustering analysis ( S2–S4 Figs and S1 Table ) . The present approach also differs from principal component analysis ( PCA , see S1J Text ) : whereas PCA finds the axes of the space which describes most of the data variance , our approach finds a polytope within which the data resides and is thus much more restrictive ( S17 Fig ) ; moreover the vertices of the polytope are different from the principal components ( e . g . a tetrahedron resides in a 3D space but has four vertices , S18 and S19 Figs ) , and our findings support their interpretation as archetypal profiles that reveal clear biological functions . PCA and other dimensionality reduction techniques are not needed to find the polytopes , but can help to visualize them . Major challenges remain , related to the high dimensionality of gene expression data , which can be in the many thousands . Fitting such data to a low-dimensional polytope can capture major trends . However , if there are small sets of genes that vary independently of the others , in a biologically important way , the current implementation of Pareto analysis will miss them because they make a small contribution to the overall shape of the data . Future work can develop ways to detect such groups of genes , separate the data and analyze each subgroup independently . This may lead to a presentation of the data as a collection ( outer product ) of Pareto fronts , each with a subset of genes related to a distinct set of tasks . Similarly , Pareto analysis can be used as a microscope to zoom in on a subset of cells or genes—in the present study , progenitor cells when considered separately without the rest of the crypt cells , revealed an informative tetrahedron of their own . Finally , one can analyze polytopes with increasing number of vertices and in this way observe the splitting of archetypes into finer distinctions ( S1K Text and S20 Fig ) . The polyhedra found here seem to be a distinct feature of the dataset , especially the straight edges and faces apparent in plots of the data distribution . However , it is possible that the observed polytopic-like structure results from other ( unknown ) reasons , such as systematic experimental errors or biological phenomena . Further application of Pareto analysis , with different tissues and different single-cell technologies that have different noise and biases , is needed to test the generality of the present conclusions . With these caveats in mind , the present approach of using Pareto analysis for single-cell data can be generally used to understand the geometry of single-cell data and to infer the tasks of individual cells in a tissue . More generally , this study indicates that the concept of cell type may be expanded . In addition to separated clusters in gene-expression space , we suggest a new possibility: a continuum of states within a polyhedron , in which the vertices represent specialists at key tasks , with generalist cells lying in the middle .
Single-cell gene expression of primary human colon crypt cells obtained by Dalerba et al [2] included 407 individual cells , each analyzed by single-cell qPCR for 83 selected genes in a Fluidigm microfluidic system [5] ( S7 Dataset ) . Note that the present dataset contains 34 genes not included in the original publication , kindly provided by Dalerba et al . These genes , including CFC1 , UGT1A1 , CLDN8 , NODAL , TDGF1 , and PCGF6 , allow identification of the nodal cell type . Some genes were measured by multiple primers , see S1A Text and S21 Fig . Cells were sorted by FACS so that they belong to one of two populations: EpCAMhigh/CD44+ , which corresponds to the bottom of the crypt , and EpCAM+/CD44-/CD66high , which corresponds the top of the crypt . The measured genes were selected using publicly available gene-expression array data sets from human colon epithelia , using a bioinformatics approach designed to identify developmentally regulated genes [2] . Gene expression was measured in Cycle threshold units ( Ct ) —the number of amplification cycles it takes to detect the gene—which correspond to the logarithm of the fold change in expression . The data was bi-modal , as seen in other applications of this technology [11 , 12] , with some cells ranging from 2 . 5Ct to 30Ct , and another set of cells with Ct>40 . Since there were no values larger than 30Ct , we treated this as the minimal threshold level of detection and set these values to 30 , and then linearly transformed the data so that minimal CT is zero ( 30-data ) . Finally , we subtracted the mean of each gene . Cells and genes with low expression were removed as described in Results . The processed data can be found in S1 Dataset . The same normalization was used for the cancer xenograft data [2] ( S4 Dataset ) and for the mouse intestinal cells data ( S3 Dataset ) [37] . We used data by Bendall et al [25] and Amir et al [13] , of human bone marrow samples from healthy donors . The cells were analyzed using single-cell mass cytometry as described in [25] , resulting in the antibody-detected expression of 31 proteins in 10 , 000 cells . Multiple detections of one protein—CD3—were united by taking their mean for each cell . Following the original publication , we transformed the data by applying a hyperbolic arcsin transformation with a cofactor of five . Processed data can be found in S5 Dataset . We used data from [3] , of mice spleen CD11c+ cells that were exposed to lipopolysaccharide ( LPS ) for 2 hours . We focused on cells that were classified in the original paper as dendritic cells ( group VII ) . We used the likelihood model of [3] in order to choose cells which have the highest likelihood to belong to this group ( rather than to other groups ) . We didn’t consider 16 genes with high batch-to-batch variability ( as in [3] ) , and removed measures of control ERCC molecules . Since the data is very sparse ( 97% of the matrix entries are 0 ) , we considered the 500 genes with highest standard deviation across samples . Results are similar for choosing genes by highest mean or median . Following Jaitin et al , we normalized the data by down-sampling: Defining a target number of molecules N , and then sampling from each cell having m> = N molecules precisely N molecules without replacement . Cells with m<N are not used for analysis . We used N = 400 , and consequently considered 312 cells in our analysis . We then transformed the data by adding 1 and applying log2 , in order to examine fold changes in expression , and reduced the average expression of each gene in all 312 cells . Processed data can be found in S6 Dataset . We found the archetypes of the best fit polytope using the PCHA algorithm [40] . This algorithm finds the best fit polytope whose vertices are on the convex hull of the data . It does so by constraining the vertices to be a weighted average of the data points , where the weights are given by a matrix C , and then approximating the data points by a weighted average of the archetypes , where the weights are given by a matrix S . The algorithm then solves the following optimization problem using a projected gradient procedure: argminC , SX-XCSF2 s . t . Cd = 1 , Sn = 1 C≥0 , S≥0 where X is the data matrix . The algorithm allows a relaxation of the optimization problem by introducing a parameter δ and relaxing the constraint on C to be 1-δ≤Cd≤1+δ This relaxation allows archetypes to be found within a certain volume around the data convex hull . Unless stated otherwise , we used δ = 0 . 5 for archetypes characterization , polytope visualization and enrichment analysis , and δ = 0 for explained variance calculation . To determine the number of archetypes that describes the data , we computed the explained variance of the data ( EV ) for each number of archetypes ( k = 2 , 3 , 4…kmax ) . The explained variance is computed by PCHA as [88 , 93]: EV = 1N∑n = 1N1-pn-snpn Where pn is the nth data point out of N points and Sn is the closest point to pn in the polytope , and | denotes Euclidean distance . Then , we identify a number of archetypes k* for which adding an additional archetype does not increase EV by much ( see Fig 2a ) . Operationally , k* was determined by the bend ( also called elbow ) in the EV versus k curve , defined by taking the most distant point from the line that passes through the first ( k = 2 ) and the last ( k = kmax = 11 ) points in graph . Changing kmax in the range 6–16 did not change the results for k* on the present datasets . We find for the present datasets that most polytopes beyond k* show low p-values ( p<0 . 01 ) , precluding the concern of type II errors ( multiple hypothesis testing ) . As a second indication for the order of the best fit polytope is the effective dimensionality of the data ( e . g . a tetrahedron suggests that the data is effectively 3D ) . We therefore tested data dimensionality using Principal Component Analysis ( PCA ) [94] . We compared the variance explained by each PC to that of randomly shuffled data . Randomized data was created by shuffling the values of each coordinate , to preserve the density distribution of each coordinate while breaking correlations between coordinates . PCA on random data yields non-equal eigenvalues , due to stochastic correlations . Data dimensionality was found by detecting the point where the explained variance of the real data comes within a standard deviation of the shuffled data , ( see Fig 2b ) . Data that is explained well by a k-vertex polytope is expected to have dimensionality of k-1 ( e . g . data explained by a tetrahedron with k = 4 is essentially 3D ) . Statistical significance of polytopes was tested by computing the EV of the real data compared to 1 , 000 sets of shuffled data , produced by randomly permuting each coordinate of the data separately . The explained variance ( EV ) by the PCHA algorithm as described above . The p-value was defined as the fraction of shuffled data sets for which the EV was larger or equal to the EV of the real dataset . To make sure the low p-value stems from the similarity of the data to a polytope and not merely from its low dimensionality , we also performed PCA on the data and checked how similar it is to a polytope when projected to the space spanned by the first n-1 PCs ( where n is the number of archetypes ) . A measure for the similarity of the data to a polytope is its t-ratio [28] , defined as the ratio of the volume of convex hull of the data to the volume of the best fit polytope ( we use PCHA with δ = 0 to find the polytope with vertices on the convex hull ) . The bigger the t-ratio , the more similar the data to the polytope . The t-ratio of the data was compared to that of 1 , 000 sets of shuffled data , produced by randomly permuting each coordinate of the data separately . p-value was defined as the fraction of sets for which the t-ratio is equal to or bigger than the t-ratio of the data . To be stringent , the p-values reported in the paper are from the second method , as they were always larger than the p-values found by the first method . To infer potential tasks of archetypes , we tested , for each gene , whether it is expressed maximally in the cells closest to one of the archetypes . To avoid circularity concerns , we removed that gene from the dataset , and recalculated the archetypes ( in the D-1 dimensional space , where D is the number of genes ) . The changes in the archetype coordinates upon removal of the gene were minor ( mean relative change 0 . 02% ) . We divided cells into bins according to their Euclidean distance from the archetype ( computed in the D-1 gene expression space ) , and asked whether the expression of the removed gene is maximal in the bin closest to the archetype , and at what statistical significance . Choosing a number of bins between 5–25 left the results essentially unchanged ( see S22 Fig ) . The p-value was computed using Wilcoxon rank-sum test [95] , comparing the distribution of expression in the first bin ( closest to archetype ) to the distribution in other bins . We set a p-value threshold of p = 0 . 001 , to avoid type II error concerns ( multi-hypothesis testing ) . Shuffled data using the same tests resulted in 0 . 1 enriched genes on average , whereas the real data showed 70 enriched genes for the intestinal cells data . In addition to the 1-dimensional leave-1-out enrichment check described above , we also performed a 2-dimenisonal enrichment check for selected marker genes . After taking out the selected gene , we projected the data on its 2 first PCA axes , and smoothed the marker values to obtain a 2D density function M ( PC1 , PC2 ) . To smooth the data we used a Gaussian kernel function with variance computed by Silverman's Rule of Thumb [96] , multiplied by a hill function of the density to avoid effects of low density of points ( in low density regions the denominator is very small , resulting in artificial high marker values ) . Where x⃑i is the ith data point , mx⃑i is the value of the marker at point x⃑i , H is a diagonal matrix H = hx00hy , hj = 4σj^53N15 , and k , n are the hill function coefficients , set to 8 , 2 respectively . We checked whether the location of the maxima of Mx⃑ is close to the computed position of the archetypes . This served as a qualitative check for enrichment . A software package that fits polytopes to biological datasets , finds their significance , and computes features such as gene categories enriched near each archetype [33] can be found at http://wws . weizmann . ac . il/mcb/UriAlon/download/ParTI .
|
In the past , biological experiments usually pooled together millions of cells , masking the differences between individual cells . Current technology takes a big step forward by measuring gene expression from individual cells . Interpreting this data is challenging because we need to understand how cells are arranged in a high dimensional gene expression space . Here we test recent theory that suggests that cells facing multiple tasks should be arranged in simple low dimensional polygons or polyhedra ( generally called polytopes ) . The vertices of the polytopes are gene expression profiles optimal for each of the tasks . We find evidence for such simplicity in a variety of tissues—spleen , bone marrow , intestine—analyzed by different single-cell technologies . We find that cells are distributed inside polytopes , such as tetrahedrons or four-dimensional simplexes , with cells closest to each vertex responsible for a different key task . For example , intestinal progenitor cells that give rise to the other cell types show a continuous distribution in a tetrahedron whose vertices correspond to several key sub-tasks . Immune dendritic cells likewise are continuously distributed between key immune tasks . This approach of testing whether data falls in polytopes may be useful for interpreting a variety of single-cell datasets in terms of biological tasks .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Geometry of the Gene Expression Space of Individual Cells
|
Follicular T helper cells ( TFH ) are fundamental in orchestrating effective antibody-mediated responses critical for immunity against viral infections and effective vaccines . However , it is unclear how virus infection leads to TFH induction . We here show that dengue virus ( DENV ) infection of human dendritic cells ( DCs ) drives TFH formation via crosstalk of RIG-I-like receptor ( RLR ) RIG-I and MDA5 with type I Interferon ( IFN ) signaling . DENV infection leads to RLR-dependent IKKε activation , which phosphorylates IFNα/β receptor-induced STAT1 to drive IL-27 production via the transcriptional complex ISGF3 . Inhibiting RLR activation as well as neutralizing antibodies against IL-27 prevented TFH formation . DENV-induced CXCR5+PD-1+Bcl-6+ TFH cells secreted IL-21 and activated B cells to produce IgM and IgG . Notably , RLR activation by synthetic ligands also induced IL-27 secretion and TFH polarization . These results identify an innate mechanism by which antibodies develop during viral disease and identify RLR ligands as potent adjuvants for TFH-promoting vaccination strategies .
Dengue virus is a global mosquito-transmitted pathogen that infects 400 mln people annually [1] . The majority of patients experience only mild fever , but the disease can progress to life-threatening dengue shock syndrome and dengue hemorrhagic fever . With no specific antivirals or effective vaccine available there is urgent need to advance our understanding of the human immune response against DENV to improve vaccine development and identify molecular targets for drug development . Antibodies are critical for the host immune response to control , eradicate and prevent ( future ) viral infections . Antibodies play a dual role in DENV pathology as neutralizing high-affinity antibodies are protective while cross-reacting antibodies can possibly enhance disease of heterologous DENV strains via antibody-dependent enhancement of infection [2] . However , it is unclear how antibodies are induced upon DENV infection . High-affinity antibodies are formed by B cells in germinal centers ( GC ) during somatic hypermutation [3] . TFH cells are critical for the formation and maintenance of GCs , and induce B cell proliferation and Ig isotype class switching by producing IL-21 . TFH cells selectively stimulate high-affinity B cell entry into GCs to promote effective antibody-mediated responses [3–5] . Formation of TFH cells is driven by transcription factor Bcl-6 , which induces IL-21 production and expression of chemokine receptor CXCR5 , which is pivotal for TFH migration into GCs [6] . DCs are essential for TFH differentiation from naïve CD4+ T cells and previously we have shown that in humans IL-27 is pivotal for TFH formation while IL-6 enhances TFH formation in response to fucosylated parasitic/bacterial ligands [7] . DCs are equipped with numerous sensors including Toll-like receptors ( TLRs ) and RLRs that sense viral particles or viral replication products to induce innate signaling and drive TH polarization for tailored adaptive immune responses . DENV can both activate TLRs as well as RLRs , depending on the cell type , leading to cytokine secretion and type I IFN production [8–10] . TLR3 resides in endosomes while RLRs are cytoplasmic receptors . Both their activation by viral RNA leads to IFN-β induction via Tank binding protein 1 ( TBK1 ) and transcription factor IRF3 . However , RLR signaling also involves IkB kinase ( IKK ) -related kinase IKKε , which functions in concert with TBK1 to activate IRF3 [11 , 12] . In parallel with IRF3-dependent IFN-β transcription , TLR3 and RLRs activate NFκB signaling to induce cytokine expression and the combined effect of type I IFN and cytokines determines which differentiation program is initiated in naïve CD4+ T cells . However , it remains unclear how viral sensing in DCs leads to TFH development and subsequent B cell activation . Here we show that DENV-infected DCs instruct naïve CD4+ T cells to differentiate into Bcl-6+CXCR5+PD-1+ IL-21-secreting TFH cells . DENV RNA replication in both monocyte-derived and primary skin DCs triggers RLR RIG-I and MDA5 leading to IFN-β transcription and IFN-α/βR activation . Notably , RLR-induced IKKε activation modulates IFNα/βR signaling by phosphorylating STAT1 . This results in the formation of transcriptional complex ISGF3 instead of STAT1 homodimers , which is pivotal for IL-27 production by DCs and TFH formation . Inhibiting RLR signaling by silencing adapter protein MAVS abrogates IL-27 production and TFH polarization by DENV-infected DCs . In addition , direct RLR activation by synthetic ligands is sufficient to induce IL-27 transcription and TFH formation .
DCs drive TH differentiation and therefore we examined DC-induced immune responses upon DENV infection . DENV efficiently infected human DCs and induced DC maturation as indicated by increased surface expression of CD83 and CD86 ( S1 Fig ) . To investigate TFH differentiation , DENV-infected DCs were co-cultured with naïve CD4+ T cells and TFH induction was determined by measuring expression of CXCR5 and PD-1 , which are both expressed by lymph node TFH cells in vivo [13 , 14] . Strikingly , DENV-infection of DCs induced a robust CXCR5+PD-1+ subset of differentiated TH cells ( Fig 1A and 1C ) , which expressed high levels of TFH-specific transcription factor Bcl-6 ( Fig 1B and 1D ) . T cell differentiation induced by DENV-infected DCs also resulted in strong secretion of IL-21 , which is the main effector cytokine of TFH cells ( Fig 1E ) . To investigate whether DENV-induced TFH cells have the capacity to activate B cells , we co-cultured DENV-differentiated TH cells with CD19+ B cells and measured antibody production . Remarkably , differentiated TH cells from DENV-infected , but not mock-treated DCs , induced secretion of both IgM and IgG by B cells ( Fig 1F ) . Blocking DENV RNA replication and infection of DCs ( S2 Fig ) with DENV RNA replication inhibitor SDM25N [15] abolished the formation of IL-21-secreting CXCR5+PD-1+Bcl-6+ TFH cells ( Fig 1A and 1C–1E ) . These data strongly indicate that DENV replication in DCs induces a TH differentiation program leading to TFH induction and B cell activation . We set out to identify the molecular mechanism in DCs essential for TFH differentiation . We investigated induction of type I IFN responses upon DENV infection . DENV induced IFN-β transcription in DCs at 18 hours post infection ( h . p . i . ) , which increased over time and correlated with DENV RNA replication ( Fig 2A ) . Next we measured induction of antiviral IFN stimulated genes ( ISGs ) , which are induced by IFNα/βR signaling and indicative of functional type I IFN responses [16] . DENV-infection of DCs induced expression of ISGs MxA , APOBEC3G , ADAR1 and TRIM5α 24 h . p . i . ( Fig 2B ) . These responses depended on DENV RNA replication as the inhibitor SDM25N abrogated the induction of IFN-β and ISGs ( Fig 2C ) . IRF7 is crucial for the induction of IFN-α , which is required for enhancing type I IFN responses [17] . Interestingly , DENV infection induced IRF7 and IFN-α expression after IFN-β induction ( Fig 2D and 2E ) . Both IRF7 and IFN-α expression depended on IFNα/βR signaling , while IFN-β expression was not decreased by blocking IFNα/βR antibodies at early time points ( Fig 2D and 2E ) . IFN-β increased at 32 h . p . i . upon blocking IFNα/βR , probably because of increased DENV replication due to absence of antiviral ISGs ( Fig 2E ) . Notably , silencing IRF7 using RNA interference abrogated IFN-α expression ( Fig 2F ) . Thus , DENV RNA replication induces functional type I IFN responses that are initiated by IFN-β . The induction of IFNα/βR-dependent IRF7 expression subsequently drives IFN-α expression to increase and prolong type I IFN responses against DENV . Both TLRs and RLRs have been implicated in DENV sensing [8–10] . To elucidate the PRR involved in DENV-induced IFN responses in DCs , we silenced adapter molecules TRIF/MYD88 and MAVS , which are essential for TLR and RLR signaling , respectively ( S3 Fig ) . Silencing MAVS strongly decreased IFN-β and ISG expression in DENV-infected DCs , in contrast , neither silencing of TRIF nor MYD88 affected type I IFN responses ( Fig 3A , S2 Fig ) even though their silencing abrogated responses to known TLR ligands ( S4 Fig ) . Notably , silencing RIG-I and MDA5 alone or together decreased DENV-induced IFN-β as well as ISG expression ( Fig 3B , S2 Fig ) . RLR triggering leads to activation and phosphorylation of IKKε and TBK1 , which target transcription factor IRF3 for nuclear translocation to drive IFN-β expression . DENV infection induced phosphorylation of TBK1 and IKKε , which was dependent on DENV RNA replication ( Fig 3C , 3D and 3E ) . Silencing of TBK1 and IKKε or treatment with the TBK1/IKKε inhibitor BX795 strongly decreased DENV-induced IFN-β expression and supports an important role for these kinases in the type I IFN response against DENV ( Fig 3F and 3G ) . Moreover , DENV infection resulted in nuclear translocation of IRF3 ( Fig 3H , 3I and 3J ) and silencing IRF3 abrogated IFN-β expression ( Fig 3K ) . These data show that RIG-I and MDA5 sense DENV RNA replication and induce type I IFN responses via TBK1 and IKKε-mediated IRF3 signaling . Parallel to type I IFN responses , RIG-I and MDA5 induce type III IFN responses which have been implicated in DC migration , viral suppression and modulation of T and B cell responses [18–21] . Therefore , we investigated type III IFN responses by analyzing the expression of IFN-λ genes ( IFNL1-4 ) . DENV induced the expression of IFNL1 and IFNL2 and these responses were not affected by type I IFN as blocking IFNα/βR signaling did not affect IFNL1 or IFNL2 expression ( Fig 4A ) . Blocking IFNLR did also not impact DENV-induced IFN-α or IFN-β responses indicating that type I and type III responses operate independently ( Fig 4B ) . IFN-λ is known to suppress viral replication and therefore we examined the effect of blocking IFNLR antibodies on DENV RNA replication . Blocking IFNLR increased DENV RNA replication but not to a similar extent as blocking IFNα/βR ( Fig 4C ) . These data suggest that although type I and type III IFN suppress DENV replication , type I IFN is more effective than type III IFN . Next , we investigated how IFN-λ is induced by DENV . Interestingly , the replication inhibitor SDM25N abrogated DENV-induced IFNL1 and IFNL2 expression ( Fig 4D ) . Moreover , IFNL1 and IFNL2 were induced by MAVS , RIG-I and MDA5 as silencing MAVS or RIG-I and MDA5 together strongly decreased both IFNL1 and IFNL2 expression by DENV infection ( Fig 4E ) . These data strongly suggest that RIG-I and MDA5 triggering by DENV replication induces type III IFN responses that operate independently of type I IFN to suppress viral replication . Next , we investigated induction of cytokines involved in TFH differentiation . Mice lacking IL-27R have impaired TFH formation [22] and we have recently shown that IL-27 is crucial for TFH polarization by human DCs in response to fucosylated parasitic/bacterial ligands [23] . In addition , Activin A and IL-12 are known to be important factors to drive human TFH formation [24 , 25] . Therefore , we examined whether DENV infection leads to Activin A , IL-12 or IL-27 expression . IL-12 is heterodimeric protein consisting of subunit p35 and p40 . Although DENV induced low levels of IL-12p35 , we were unable to detect IL-12p40 expression and this resulted in a lack of IL-12p70 protein ( S5A and S5B Fig ) . We were also unable to detect increased Activin A expression in DENV-infected cells , while LPS strongly induced both IL-12p70 and Activin A ( S5A and S5B Fig ) . Direct stimulation of RIG-I and MDA5 with RLR ligand poly ( I:C ) Lyovec also did not induce IL-12p70 production or Activin A expression ( S5A and S5B Fig ) , indicating that these factors are more associated with TLR-induced TFH formation and that RLR-mediated TFH differentiation depends on other factors . Interestingly , DENV infection of DCs induced IL-27 production , which decreased by silencing either MAVS or both RIG-I and MDA5 ( Fig 5A ) . IL-27 is a heterodimeric cytokine consisting of subunit p28 and Epstein-Barr virus-induced gene 3 ( EBI3 ) [26] . DENV infection induced expression of both IL-27p28 and EBI3 , which was dependent on RLR signaling as well as viral replication ( Fig 5B; S5C Fig ) . Furthermore , blocking IFNα/βR antibodies abrogated IL-27p28 expression without affecting EBI3 expression ( Fig 5C ) , supporting differential regulation of IL-27p28 and EBI3 [27 , 28] and an important role for crosstalk between RLR and IFNα/βR signaling . Interestingly , IL-27p28 expression was not affected by blocking IFNLR antibodies indicating that specific IFN signaling is necessary for IL-27p28 expression ( Fig 5D ) . Il-27p28 contains an IFN-stimulated response element ( ISRE ) , which is induced by IFN-stimulating gene factor 3 ( ISGF3 ) , a complex of STAT1 , STAT2 and IRF9 [27] . Differential signaling by IFNα/βR triggering leads to induction of either STAT1 homodimers or ISGF3 induction [29] . As ISGF3 formation is controlled by IKKε-dependent phosphorylation of STAT1 at Ser708 , which prevents STAT1 homodimer formation [30] and IKKε also enhances the transactivation capacity of ISGF3 by phosphorylating STAT1 at Ser727 [23] , we investigated whether ISGF3 was involved in IL-27 induction upon DENV infection . We observed STAT1 phosphorylation at Ser708 and Ser727 which was abrogated after treatment with TBK1/IKKε inhibitor BX795 ( Fig 5E , 5F and 5G ) . We also observed nuclear translocation of IRF9 in DENV infected DCs ( Fig 5H , 5I and 5J ) . Moreover , both IKKε and IRF9 were crucial for IL-27p28 expression since silencing of IKKε or IRF9 strongly decreased IL-27p28 expression ( Fig 5K ) . These data show that RLR activation by DENV leads to IFN-β production and subsequent IFNα/βR triggering , and that IKKε activation modulates IFNα/βR signaling to drive IL-27 production . We next set out to investigate the importance of RLR sensing in TFH formation by DENV-infected DCs . Silencing of MAVS abrogated the formation of CXCR5+PD-1+ TFH cells by DENV-infected DCs ( Fig 6A ) . Moreover , silencing of MAVS diminished Bcl-6 expression in T cells polarized by DENV-infected DCs ( Fig 6B ) . These data strongly suggest that MAVS activation by DENV is pivotal for the formation of TFH cells and identifies an important role for RLR activation in TFH cells induction . To examine this , we transfected DCs with poly ( I:C ) or 5’pppRNA to activate RIG-I and MDA5 or RIG-I alone , respectively , and investigated TFH differentiation . Strikingly , stimulation of DCs with either poly ( I:C ) Lyovec or 5’pppRNA-Lyovec induced CXCR5+PD-1+ TFH formation and increased Bcl-6 expression ( Fig 6E and 6F ) . Notably , TH cells polarized by poly ( I:C ) or 5’pppRNA transfected DCs activated B cells to produce IgM and IgG ( Fig 6G ) . In contrast , LPS-stimulated DCs did neither induce TFH formation nor B cell activation ( Fig 6E–6G ) . We next investigated the importance of IL-27 secretion by DCs for TFH formation . Interestingly , both poly ( I:C ) and 5’pppRNA transfected DCs produced IL-27 ( Fig 6C and 6D ) and neutralizing antibodies against IL-27 abrogated CXCR5+PD-1+ TFH formation ( Fig 6F ) . Neutralizing IL-27 also diminished the capacity of TH cells polarized by poly ( I:C ) or 5’pppRNA transfected DCs to activate B cells to produce IgM and IgG ( Fig 6G ) . These data strongly indicate that RLR activation , either by DENV or synthetic ligands , drives IL-27-dependent TFH polarization . DENV infection is initiated after a mosquito bite in the skin and swift immune responses against DENV depends on activation of skin DCs [31] . Human skin harbors several DC subsets including CD14+ and CD1c+ dermal DCs of which CD14+ dermal DCs are specialized in the induction of TFH cells [32] . Therefore , we isolated CD14+ and CD1c+ dermal DCs from human skin and investigated if dermal DCs mount immune responses against DENV . Interestingly , DENV induced type I IFN responses in CD1c+ as well as CD14+ dermal DCs , although the induction of type I IFN was more robust in CD14+ dermal DCs than in CD1c+ dermal DCs ( Fig 7A ) . Notably , CD14+ dermal DCs specifically expressed IL-27p28 in response to DENV infection ( Fig 7A ) . We next set out to investigate the importance of DENV replication in IL-27p28 expression by CD14+ dermal DCs . Remarkably , DENV RNA replication inhibitor SDM25N abrogated the induction of IL-27p28 as well as IL-27EBI3 in DENV infected CD14+ dermal DCs ( Fig 7B ) . These data indicate that human dermal DCs mount type I IFN responses against DENV and that DENV replication is essential to trigger IL-27 expression in CD14+ dermal DCs .
TFH cells play a key role in antibody-mediated responses during viral infections or vaccination . Although numerous studies have demonstrated the importance of TFH cells in GC reactions [3–5] , little is known about the factors that drive TFH differentiation from naïve CD4+ T cells upon viral infection . Here we show that DENV induces RLR-crosstalk with IFNα/βR signaling leading to IL-27 secretion , which is pivotal for the formation of IL-21-producing CXCR5+PD-1+Bcl-6+ TFH . DENV replication triggered RLR signaling leading to IKKε activation , which is essential for RLR-IFNα/βR crosstalk by phosphorylating STAT1 and inducing ISGF3 formation . Notably , direct activation of RLRs in DCs using poly ( I:C ) Lyovec or 5’pppRNA-Lyovec also induced IL-27 secretion , TFH polarization and IgM and IgG production by B cells . These data strongly suggest that RLRs are efficient in the induction of TFH responses via their crosstalk with IFNα/βR signaling , and links viral recognition to induction of robust antibody responses . Viral RNA is a potent pattern-associated molecular pattern that can activate numerous receptors and induce strong immune responses; both TLRs and RLRs have been implicated in DENV sensing [8–10] . DENV particles contain positive single-stranded RNA that can be directly sensed by TLR7 as shown in macrophages [8] . However , we did not find a role for TLR7 in recognizing DENV by DCs as MYD88 silencing did not affect type I IFN responses , probably because TLR7 signaling requires IRF7 , which is constituently expressed in plasmacytoid DCs but minimally in other DC subsets [17 , 33] . Indeed , our data show that IRF7 is minimally expressed by DCS and that this transcription factor is induced by DENV infection of DCs via IFN-β . DENV replication leads to double stranded RNA intermediates , which can be sensed by TLR3 and RLRs . Studies in multiple cell lines have shown that TLR3 can be involved in IFN-β production in response to DENV infection although it is unclear how cytoplasmic RNA is transferred to endosomal TLR3 [9 , 34] . Our data strongly suggest that RIG-I and MDA-5 but not TLR3 are involved in sensing of DENV and subsequent induction of type I and type III IFN responses . Although DENV is known to block type I IFN responses by inhibiting RLR-MAVS interaction , TBK1 and IRF3 phosphorylation and IFNα/βR signaling via STAT2 degradation [35–38] , our data suggest that the inhibition does not effectively occur in DCs . Both type I IFN and type III IFN suppressed viral replication of DENV and preventing antiviral ISG induction by IFNα/βR or IFNLR increased DENV RNA replication . Notably , the increase in DENV RNA replication resulted in an increase in IFN-β suggesting that DENV RNA and IFN-β levels are correlated . It is likely that sensing of RNA products precedes activity of de novo produced viral proteins that block RLR signaling . Recently , it was shown that Measles virus ( MV ) directly affects RLR activation . RIG-I and MDA5 activation is tightly regulated and requires dephosphorylation by PP1 phosphates for activation [39] . MV replication is sensed by RLRs and leads to type I IFN responses . However , MV triggers C-type lectin receptor DC-SIGN signaling leading to kinase Raf-1 activation [23] . Raf-1 subsequently induces association of inhibitor protein I-1 with PP1 to lower RIG-I and MDA5 dephosphorylation and type I IFN induction [40] . Although DENV also binds to DC-SIGN [41 , 42] , we did not observe inhibition of RLR activation upon infection . In contrast to MV infection , which activates RLRs very rapidly [43] , DENV infection is only sensed after 18 hours when innate signaling by DC-SIGN is probably not effective anymore . Therefore , RIG-I and MDA5 activation is not only important for TFH polarization and antibody production but also to limit viral replication in DCs and possible viral transmission to other cells . Indeed , our data show that RLR-dependent induction of type I IFN and type III IFN suppressed DENV . We have recently shown that IL-27 is important in the differentiation of TFH and IL-27 induction is dependent on the formation of ISGF3 [7] . Although several TLRs can induce IL-27 transcription , the levels are not sufficient to induce TFH polarization and requires IKKε activation by other receptors for STAT1 phosphorylation and ISGF3 formation [7] . RLRs are therefore unique in their ability to activate both IKKε and IFNα/βR signaling for efficient IL-27 transcription and TFH polarization . These underlying mechanisms might also apply to other viruses as Measles virus , Influenza virus , Rubella virus , HIV-1 and Hepatitis C virus activate RLRs during infection [44–47] . Indeed , our data strongly suggest the use of RLR-ligands as adjuvants for human vaccination strategies , which has been shown to be successful in animal models [48–50] . In addition to IL-27 , Activin A and IL-12 have been identified as important cytokines to drive human TFH differentiation [24] . IL-12 expression is strongly induced by TLR signaling , while it is inhibited by RIG-I mediated IRF3 activation [51] . Our data also show that RLR triggering by synthetic ligands or DENV does not lead to IL-12p70 production , in contrast to strong IL-12p70 production by TLR4 activation . We obtained similar results for Activin A , suggesting that Activin A and IL-12p70 could be important for TLR-mediated TFH formation while IL-27 is crucial for RLR-mediated TFH formation . In the natural course of infection , skin DCs are the first immune cells to encounter DENV after a blood meal of an infected mosquito [31 , 52] . Effective control of viral propagation from the site of infection requires robust type I IFN responses to suppress viral replication . Our data show that both CD14+ and CD1c+ dermal DCs mount type I IFN responses against DENV . Interestingly , DENV specifically induced IL-27 expression in CD14+ dermal DCs , which are known to be effective inducers of TFH differentiation [32] . Our data show that the induction of IL-27 by DENV in CD14+ dermal DCs critically depends on DENV RNA replication and thereby supports a key function of cytoplasmic sensors of DENV RNA replication in the induction of TFH responses by primary human skin DCs in response to DENV infection . Our data shows that IFN-α/β induced IL-27 expression is pivotal for TFH formation by DCs while direct stimulation of naïve CD4+ cells with type I IFN-α/β is thought to inhibit TFH formation [24 , 53] . These studies indicate that IFN signaling , depending on the cell-type and time , can have different effects on TFH differentiation . In addition , direct IFN-α/β stimulation does not lead to IL-27 transcription without IKKε activation to modulate IFNα/βR signaling . Developing effective DENV vaccines has been hampered by the formation of non-neutralizing antibodies that have the potential to enhance disease [2 , 54] . A subunit vaccine based on the neutralizing epitope of DENV envelop protein could circumvent the formation of non-neutralizing antibodies [55 , 56] . However , subunit vaccines usually have low immunogenicity and induce only weak antibody responses . We propose that a subunit vaccine containing the neutralizing DENV epitope in combination with RLR-based adjuvants is a potent strategy to induce high levels of DENV neutralizing antibodies . In conclusion , we have identified an innate mechanism in DCs that drives TFH polarization during viral infection . Adjuvants targeting this innate mechanism have the potential to improve vaccination strategies for DENV and other pathogens .
This study was done in accordance with the ethical guidelines of the Academic Medical Center and human material was obtained in accordance with the AMC Medical Ethics Review Committee ( i . e . Institutional Review Committee ) according to the Medical Research Involving Human Subjects Act . Buffy coats obtained after blood donation ( Sanquin ) or skin tissue are not subjected to informed consent according to the Medical Research Involving Human Subjects Act and the AMC Medical Ethics Review Committee . All samples were handled anonymously . Peripheral blood monocytes were isolated from buffy coats of healthy donors ( Sanquin ) by Lymphoprep ( Axis-Shield ) gradient followed by Percoll ( Amersham Biosciences ) gradient steps . Monocytes were differentiated into immature DCs in the presence of 500 U/ml IL-4 and 800 U/ml GM-SCF ( both Invitrogen ) for 6–7 days in RPMI supplemented with 10% fetal calf serum , 10 U/ml penicillin , 10 mg/ml streptomycin ( all Invitrogen ) and 2 mM L-glutamine ( Lonza ) . This study was done in accordance with the ethical guidelines of the Academic Medical Center . Dermal DCs ( DDCs ) were isolated from human skin tissue obtained from healthy donors after corrective breast or abdominal surgery . A dermatome ( Zimmer ) was used to produce 0 . 3 mm skin grafts that were treated with dispase ( 1U/ml , Roche ) for 45 min at 37°C to separate dermis and epidermis . Dermal tissue was floated on medium for 16h . Migrated cells were collected and separated based on CD14 ( 130-050-201 , Miltenyi ) and CD1c ( 130-090-506 , Miltenyi ) expression using magnetic beads according to the manufactures instruction . Isolated cells were analyzed for HLA-DR-PE/Cy7 ( 1:200 , 560651 , BD ) , CD11c-Alexa647 ( 1:100 , 2108100 , BioLegend ) , CD14-PerCP ( 1:10 , 345786 BD ) and CD1c-APC/Cy7 ( 1:50 , 331519 , BioLegend ) expression on a BD Canto II ( S6 Fig ) . CD14+ DDCs were characterized as HLA-DR+CD11c+CD14+CD1c+ and CD1c+ DDCs as HLA-DR+CD11c+CD1c+CD14- . Purity of sorted cells was over 95% . DCs were stimulated with 1 μg/ml poly ( I:C ) LyoVec LMW or 10 μg/ml 5’ppp-dsRNA-LyoVec ( both Invivogen ) unless stated otherwise . DENV replication inhibitor SDM25N ( 10μM , Tocris Bioscience ) , blocking IFNα/βR antibody or blocking IFNLR antibody ( 20 μg/ml , both PBL Interferon Source ) were added simultaneous with DENV to DCs . DCs were transfected with 500 nM short interfering RNAs ( siRNAs ) using the Neon Transfection System ( ThermoFisher ) according to the manufacturer’s instructions . In brief , DCs were washed with PBS , resuspended in Buffer R ( ThermoFisher ) and divided over different siRNAs . DCs were transfected with a single pulse of 1500V for 20 ms , mixed with complete RMPI and incubated for 48h before stimulation . SMARTpool siRNA used were MAVS ( M-024237-02 ) , TRIF ( M-012833-02 ) , MYD88 ( M-004769-01 ) , RIG-I ( M-012511-01 ) , MDA5 ( M-013041-00 ) , TBK1 ( M-003788-02 ) , IKKε ( M-003723-02 ) , IRF9 ( M-020858-02 ) , IRF3 ( M-006875-02 ) , and non-targeting siRNA ( D-001206-13 ) as control ( all Dharmacon ) . Silencing was confirmed by real-time PCR , flow cytometry and immunoblot ( S3 Fig ) . Naive CD4+ T cells were isolated from buffy coats of healthy blood donors ( Sanquin ) with human CD4+ T-cell isolation kit II ( Miltenyi ) by negative selection and subsequent depletion of CD45RO+ memory T cells using phycoerythrin ( PE ) -conjugated anti-CD45RO ( 80μg ml-1; R0843; Dako ) and anti-PE beads ( Miltenyi ) . B cells were isolated from buffy coats of healthy blood donors ( Sanquin ) with human B-cell isolation kit II ( Miltenyi ) by negative selection . This study was approved by the Medical Ethics Review Committee of the AMC . DCs were either silenced for indicated proteins or treated with SDM25N and stimulated for 48h as indicated . DCs were combined with allogeneic naïve CD4+ T cells ( 5 , 000 DCs/20 , 000 T cells ) in the presence of 10 pg/ml Staphylococcus aureus enterotoxin B ( Sigma ) . SDM25N ( 1 μM , Tocris Bioscience ) was added to cocultures of SDM25N-treated DCs to maintain inhibition of DENV replication . Neutralizing antibodies against IL-27 ( 5 μg/ml , AF2526; R&D Systems ) or normal goat IgG ( AB-108-C; R&D Systems ) as isotype control was added at the start of DC-T cell coculture . After 3 days , cells were further cultured in the presence of 10 U/ml IL-2 ( Chiron ) . Resting T cells were restimulated with 100 ng/ml PMA and 1 μg/ml ionomycin ( both Sigma ) for 24h . For flow cytometry analysis of restimulated T cells , cells were stained with Alexa Fluor 647-conjugated anti-CXCR5 ( 1:800; 558113; BD Pharmingen ) and PerCP-Cy5 . 5-conjugated α-PD-1 ( 1:50; 561273; BD ) before fixation in 2% para-formaldehyde for 20 min , followed by permeabilization in 50% methanol at -20°C for 45 min . Cells were stained with anti-Bcl-6 ( 1:50; ab19011; Abcam ) , followed by incubation with PE-conjugated anti-rabbit ( 1:200; 711-116-152 , Jackson ImmunoResearch ) . Cells were analyzed on a FACS Canto II ( BD Biosciences ) . Supernatants of restimulated T cells were harvested after 24h and IL-21 expression was analyzed by ELISA ( eBioscience ) . T-cell dependent B-cell activation was assessed by coculturing resting differentiated T cells restimulated with 1 μg/ml anti-CD3 ( 1XE , Sanquin ) and 2 μg/ml anti-CD28 ( 15E8 , Sanquin ) with allogeneic B cells ( 100 , 000 T cells/50 , 000 B cells ) . Supernatants were harvested after 7 days for analysis of IgM and IgG production by ELISA ( eBioscience ) . DENV-2/16681 was added to 80% confluent C6/36 cells at an MOI of 0 . 01 in RPMI medium RPMI supplemented with 2% fetal calf serum , 10 U/ml penicillin , 10 mg/ml streptomycin ( all Invitrogen ) and 2 mM L-glutamine ( Lonza ) . After 5–7 days , supernatant was harvested and cleared from cellular debris by centrifugation and subsequent filtration using a 0 . 2 μM filter . Supernatant was aliquoted , snap-frozen in liquid nitrogen and stored at -80°C . Viral titers were determined as described previously[57] . DCs were infected with DENV at an MOI of 1 unless stated otherwise . Infection was determined after 36-48h by flow cytometry . Cells were fixed in 4% para-formaldehyde for 15 min followed by permeabilization in PBS supplemented with 0 . 1% saponin for 10 min . Cells were stained with anti-NS3 ( 1:800 , SAB2700181 , Sigma ) followed by PE-conjugated anti-rabbit ( 1:200; 711-116-152 , Jackson ImmunoResearch ) in combination with APC-conjugated CD83 ( 1:25 , 551073 , BD Pharmingen ) and FITC-conjugated CD86 ( 1:25 , 555657 , BD Pharmingen ) . Cells were analyzed on a FACS Canto II ( BD Biosciences ) . TBK1 , IKKε and STAT1 phosphorylation was determined by flow cytometry and immunoblot . For flow cytometry , cells were fixed in 4% para-formaldehyde for 15 min followed by permeabilization in 90% methanol at -20°C for 45 min . Cells were stained with phospho-specific antibodies against TBK1 Ser172 ( 1:50 , 5483S , Cell Signaling ) , IKKε Ser172 ( 1:50 , 06–1340 , Millipore ) , STAT1 Ser708 ( 1:100 , provided by M . Gale , Jr , University of Washington School of Medicine , Seattle , WA , ( Perwitasari et al . , 2011 ) or STAT1 Ser727 ( 1:200 , 9177; Cell Signaling ) , followed by PE-conjugated anti-rabbit ( 1:200; 711-116-152 , Jackson ImmunoResearch ) . Cells were analyzed on a FACS Calibur ( BD Biosciences ) . For immunoblot , whole cell extracts were prepared using Ripa lysis buffer ( Cell Signaling Technology ) and protein were resolved by SDS-PAGE and detected with anti-TBK1 Ser172 ( 1:1000 , 5483S , Cell Signaling ) anti-IKKε Ser172 ( 1:1000 , 06–1340 , Millipore ) , anti-STAT1 Ser708 ( 1:1000 , provided by M . Gale , Jr , University of Washington School of Medicine , Seattle , WA , ( Perwitasari et al . , 2011 ) or anti-STAT1 Ser727 ( 1:1000 , 9177; Cell Signaling ) . Membranes were also probed with anti-TBK1 ( 1:1000 , 3504 , Cell Signaling ) , anti-IKKε ( 1:1000 , 2905 , Cell Signaling ) or anti-STAT1 ( 1:1000 , 9172 , Cell Signaling ) as loading control . Primary antibody was detected using HRP-conjugated secondary antibody ( 1:2000 , 21230 , Pierce ) mRNA was isolated using mRNA capture kit ( Roche ) and cDNA was synthesized with reverse transcriptase kit ( Promega ) . PCR amplification was performed in the presence of SYBR Green in an ABI 7500 Fast PCR detection system ( Applied Biosystems ) . Specific primers were designed using Primer Express 2 . 0 ( Applied Biosystems; S1 Table ) . Expression of target genes was normalized to GAPDH ( Nt = 2Ct ( GAPDH ) –Ct ( target ) ) and set at 1 in DENV-infected DCs for each donor within one experiment . Nuclear translocation of IRF3 and IRF9 was determined by confocal microscopy , immunoblot and ELISA . For Confocal microscopy , cells were allowed to adhere to poly-l-lysine coated glass slides for 20 min at 37°C before fixation in 2% para-formaldehyde for 20 min followed by permeabilization using 0 . 2% Triton for 10 min . Cells were stained with anti-IRF3 ( 1:100 , D83B9 , Cell Signaling ) or anti-IRF9 ( 5 μg/ml , sc-496X , Santa Cruz ) and anti-DENV E protein ( 1:400 , 3H5-1 , Millipore ) followed by Alexa Fluor 488-conjugated anti-mouse ( 1:400 , A11029 , Invitrogen ) and Alexa Fluor 546-conjugated anti-rabbit ( 1:400 , A11035 , Invitrogen ) . Nuclei were stained using Hoechst ( 1:10 , 000 , Molecular Probes ) . Cells were analyzed on a Leica TCS SP8 X mounted on a Leica DMI6000 inverted microscope and data was processed using Leica LAS-X software . For immunoblot , nuclear and cytoplasmic extracts were prepared using NucBuster protein extraction kit ( Novagen ) . Proteins were resolved by SDS-PAGE and detected by immunoblotting with anti-iRF3 ( 1:1000 , 4302; Cell Signaling ) or anti-IRF9 ( 1:1000 , sc-496 , Santa Cruz ) . Membranes were also probed with anti-β-actin ( 1:2500 , sc-81178; Santa Cruz ) to ensure equal protein loading . Detection was performed as described above . Specific increase of IRF3 and IRF9 in the nuclear fraction without increase in the cytoplasmic fraction underscores specificity of the fractionation . IRF3 and IRF9 levels in nuclear extracts was also determined using ELISA ( IRF3 , SEB589HU , USCN Life Sciences; IRF9 , MBS921012 , MyBiosource ) . Statistical analyses were performed using the Student’s t-test for paired observations . Statistical significance was set at P<0 . 05 .
|
Strong antibody production is critical for effective immune responses against viral infections and is a primary factor in the development of successful vaccines . However , it is unclear how virus infection leads to effective antibody responses . Dengue virus ( DENV ) is known to induce potent antibody production , although the underlying mechanism is poorly understood . Dendritic cells ( DCs ) are professional sentinels of the immune system and crucial for induction of immune responses . Here we show that DENV infection of human DCs leads to robust antibody production by inducing a specific T helper cell type ( also called follicular T helper or TFH ) that specializes in stimulating antibody production by B cells . Our data show that DENV replication triggers a viral detection system consisting of sensors RIG-I and MDA5 , which specifically induce factors such as IL-27 that are essential for TFH induction . Our data demonstrate that this viral detection system is especially powerful to induce antibody production . Indeed , synthetic molecules that trigger this viral detection mechanism induced superior antibody production compared to other activation signals . Thus , we have identified a viral detection mechanism that leads to strong antibody production and its importance in DENV infection as well as its potential in vaccinations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"enzyme-linked",
"immunoassays",
"immune",
"physiology",
"gene",
"regulation",
"immunology",
"microbiology",
"cell",
"differentiation",
"developmental",
"biology",
"antibodies",
"immunologic",
"techniques",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"immune",
"system",
"proteins",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"gene",
"expression",
"t",
"cells",
"immunoassays",
"viral",
"replication",
"spectrophotometry",
"biochemistry",
"rna",
"antibody-producing",
"cells",
"cytophotometry",
"cell",
"biology",
"nucleic",
"acids",
"b",
"cells",
"virology",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"non-coding",
"rna",
"spectrum",
"analysis",
"techniques"
] |
2017
|
RIG-I-like receptor activation by dengue virus drives follicular T helper cell formation and antibody production
|
Infection of the genitourinary tract with Group B Streptococcus ( GBS ) , an opportunistic gram positive pathogen , is associated with premature rupture of amniotic membrane and preterm birth . In this work , we demonstrate that GBS produces membrane vesicles ( MVs ) in a serotype independent manner . These MVs are loaded with virulence factors including extracellular matrix degrading proteases and pore forming toxins . Mice chorio-decidual membranes challenged with MVs ex vivo resulted in extensive collagen degradation leading to loss of stiffness and mechanical weakening . MVs when instilled vaginally are capable of anterograde transport in mouse reproductive tract . Intra-amniotic injections of GBS MVs in mice led to upregulation of pro-inflammatory cytokines and inflammation mimicking features of chorio-amnionitis; it also led to apoptosis in the chorio-decidual tissue . Instillation of MVs in the amniotic sac also resulted in intrauterine fetal death and preterm delivery . Our findings suggest that GBS MVs can independently orchestrate events at the feto-maternal interface causing chorio-amnionitis and membrane damage leading to preterm birth or fetal death .
Preterm birth is the leading cause of neonatal mortality worldwide [1] . Globally , an estimated 13 million babies are born prematurely each year , out of which more than one million succumb to death [2] . In addition , being the leading cause of neonatal death , preterm birth also increases the risk of neonatal infections [3] . The survivors of preterm birth are also at increased risk of neurodevelopmental impairments , respiratory and gastrointestinal complications [4] . Amongst the various causes of preterm birth , intrauterine infections by various bacterial pathogens have been suggested to be one of the main reasons [5] . Group B Streptococcus ( Streptococcus agalactiae , GBS ) the β-hemolytic , gram-positive bacteria colonizes the genitourinary tract of almost 20–30% pregnant women [6–8] and is frequently associated with preterm births [9] . Epidemiological and clinical studies have shown that colonization of vagina and cervix with GBS significantly increases the probability of intra-amniotic infection , chorio-amnionitis , endometriosis , preterm premature rupture of amniotic membrane ( PPROM ) and preterm births [10–15] . It is postulated that ascending GBS infection from the vagina to the feto-maternal space decreases amniotic membrane integrity causing PPROM [16] . This effect is attributed to the hemolytic pigment of GBS which causes chorio-amnionitis in vivo [17] . While bacterial infections have been strongly associated with preterm births , it is not clear how preterm labor-related infections occur . Although ascending infections are postulated to be the main reason of preterm births , recent studies have suggested that intra-amniotic inflammation associated with spontaneous preterm labor occurs even in the absence of detectable microorganisms in the feto-maternal interface and amniotic fluid , a phenomenon , referred to as ‘sterile intra-amniotic inflammation’ [18] . Similar observations were made in an experimental model of rhesus monkeys where GBS was not detected in the amniotic fluid despite extensive inflammation [19] . These observations led us to postulate that the physical presence of the bacteria in the amniotic fluid and/or the chorio-decidua may not be necessary for intra-amniotic inflammation and preterm birth . Interaction with the environment and other units of life forms an important cellular phenomenon and is mediated via the action of either cell surface associated or secreted molecules . The latter bypasses the need for physical presence of the cell at the site of interaction which often might not be possible due to limitations of size , distance , presence of hostile molecules etc . Prokaryotes have a wide variety of secretion system which includes the classical secretory ( Sec ) system , the TAT system , accessory Sec system and ABC transporters . Apart from these , outer membrane vesicles secreted by gram-negative bacteria have been proposed to be an ancillary secretory mechanism . These bilayered structures were found to be secreted almost ubiquitously by most , if not all gram negative bacteria wherein they perform a wide range of functions including quorum sensing [20] , biofilm formation [21] , nutrient acquisition , defense [22] and stress resistance [23] . Lately , extracellular membrane vesicles ( MVs ) are also reported to be produced by a number of gram positive bacteria . These include Staphylococcus aureus [24] , Bacillus anthracis [25] , Clostridium perfringens [26] , Bacillus subtilis [27] and very recently in Streptococcus pneumoniae [28] and Streptococcus suis [29] . Loaded with toxins and other virulence factors [25] , adhesins and immuno-modulatory substances [30] , these MVs contribute to the survival , virulence and dissemination of the pathogens in the host . While GBS is not known to produce similar vesicular structures; based on the observations in other pathogenic organisms , we hypothesized that GBS may also produce MVs , which at the feto-maternal interface/or amniotic fluid cause tissue damage resulting in PPROM and/or preterm delivery . In this report , we demonstrate for the first time that independent of the strains , GBS produces MVs . These MVs are capable of anterograde transport in mouse reproductive tract , have collagenase activity and reduce the stiffness of mouse chorio-decidual membrane ex vivo . In vivo injection of GBS MVs in mouse amniotic sacs causes chorio-amnionitis and inflammation resulting in premature delivery and fetal demise . Collectively , these findings provide a novel insight into how GBS can orchestrate events at the fetal membrane leading to premature birth .
Visualization of ultracentrifuged culture supernatant from GBS strain A909 ( serotype IA ) by Transmission Electron Microscopy ( TEM ) revealed numerous spherical structures of varying sizes resembling MVs ( Fig 1A ) . The budding and release of MVs by GBS cells were observed by Scanning Electron Microscopy ( SEM ) where the vesicles were found to be aggregated both at the time of secretion and following discharge ( Fig 1B and 1C ) . We also determined the presence of MVs associated with GBS by Atomic Force Microscopy ( AFM ) ( S1A Fig ) . Along with A909 , other GBS strains , including COH1 ( serotype III ) , NEM316 ( serotype III ) and 2603 V/R ( serotype V ) were also found to secrete MVs ( S1B Fig ) . We stained the putative MVs from GBS strain A909 with DiO , a lipophilic fluorescent dye . Flow cytometry of the labeled MVs revealed that almost 99% of the vesicles were DiO positive ( Fig 1D ) . Dynamic Light Scattering ( DLS ) of the vesicle preparations revealed two distinct populations , one < 50 nm and the other in the range of 150–300 nm ( Fig 1E ) . Some larger structures ( > 300 nm ) observed in DLS analysis was thought to be vesicular aggregates , supporting our observations from SEM analysis . Quantitatively , 1 ml of late exponentially grown GBS culture was estimated to produce 1 . 7x104 vesicles . The number of vesicles increased in a growth phase dependent manner , mostly due to increase in cell numbers . Maximum number of vesicles were attained at an OD of 1 . 2 corresponding to late-exponential phase of growth for GBS ( S1C Fig ) . We characterized the MVs isolated from GBS strain A909 with respect to its composition . Comparison of the fatty acid profiles of the lipids isolated from MVs and GBS cells using GC-MS revealed that both intact bacterial cells and vesicles had similar lipid compositions with palmitic acid being the major fatty acid in both cases ( Fig 2A ) . We next analyzed the protein composition of GBS MVs by separating the total proteins in SDS-PAGE followed by peptide mass fingerprinting by MALDI mass spectrometry ( Fig 2B ) . Eight out of 22 protein bands in a Coomassie stained SDS-PAGE gel used for GBS strain A909 MV proteins , could be successfully sequenced ( Fig 2B ) . The list of identified proteins , their MASCOT scores and other details are provided in Table 1 . In silico analysis revealed that all these proteins were either secreted or cell wall/membrane associated virulence factors ( Table 1 ) . The vesicles from strain A909 also contain DNA at a concentration of ~33 ± 5 ng/μg of MV protein . We attempted to amplify few genes encoding various GBS virulence factors , such as , cylE ( acyl CoA acyl transferase involved in hemolysin synthesis ) , cfb ( CAMP factor ) , pepB ( metallopeptidase ) , zooA ( zoocin ) and gapN ( glyceraldehyde 3-phosphate dehydrogenase ) from the DNA isolated from MVs . A PCR product ( 750 bp ) was obtained when primers specific for cfb gene was used ( Fig 2C ) . No PCR products were detected when primer sets of other genes ( cylE , pepB , zooA and gapN ) were used for PCR , revealing the absence of the cognate DNA from isolated MVs . However , all these genes could be successfully amplified when genomic DNA of GBS was used as template in PCR . We next attempted to determine if MVs can interact with host cells by studying the effects of purified MVs from GBS strain A909 on HeLa cells . A significant dose dependent reduction in the viability of HeLa cells was observed within 24 h of challenge with MVs . The reduction in viability was statistically significant at concentrations greater than 100 μg/ml ( Fig 2D ) . BSA entrapped liposomes that were used as negative control had no effect on HeLa cells viability even at higher concentrations ( Fig 2D ) , revealing that the cytotoxicity was specific for GBS MVs . To determine if MVs can enter host cells , HeLa cells were incubated with fluorescently labeled MVs and visualized using a confocal microscope . FITC labeled MVs could be readily detected inside HeLa cells ( Fig 2E ) . Extracellular MVs could be distinguished from the internalized MVs by double immunofluorescence where the internalized vesicles remain green and the extracellular MVs are red ( Fig 2E ) . Since GBS MVs contain few ECM degrading proteases ( Table 1 ) , we tested the presence of collagenolytic activity in MVs from GBS strain A909 by gelatin zymography which resulted in a single clear band ( Fig 3A ) . This underlines the presence of gelatinolytic activity in GBS MVs . To test if the GBS MVs degrade collagen in the host tissues , mouse chorio-decidual membranes were incubated with MVs from strain A909 and stained for collagen ( Fig 3B ) . In liposome ( BSA entrapped ) treated membranes ( negative control ) , intense collagen staining was observed in the region below the epithelial layer and the collagen fibers were continuous . However , in the membranes treated with MVs , collagen was fragmented and thread like structures were observed indicative of collagen degradation . The collagen degrading effect could be attributed to the activity of proteases present in MVs since this could be attenuated when the chorio-decidual tissue was incubated with MVs in presence of protease inhibitors ( Fig 3B ) . We next tested the mechanical properties of the fetal membranes following treatment with MVs . By AFM analysis ( Fig 3C ) , as compared to liposome challenge ( 2 . 85 ± 0 . 66 kPa ) ; treatment of mouse chorio-decidual membranes with MVs revealed a significant reduction in the stiffness of the membranes ( 1 . 19 ± 0 . 03 kPa ) . This detrimental effect of MVs was not observed in presence of protease inhibitors as compared to MV treatment , incubation of MVs in presence of protease inhibitors had significantly higher stiffness ( 2 . 24 ± 0 . 04 kPa ) . Significant reduction in the stiffness was also observed when the membranes were treated with collagenase ( 1 . 40 ± 0 . 09 kPa ) that was used as positive control ( S2 Fig ) . To test if the MVs in the vagina can travel anterogradely in the uterus , FITC labeled MVs from GBS strain A909 were instilled in mouse vagina and 6 h later , the uteri were imaged using confocal microscope . Bright fluorescent signals were detected abundantly at the utero-cervical junction where the MVs seemed to be lodged in the folds and crevices of the tissue . Intense fluorescent signals were also detected in the distal uterus where MVs appeared to be layered on to the cells . In the anterior most segment of the uterus , although less frequent , high intensity signals were detected as clusters lodged on the walls of the uterus . These signals were specific to FITC labeled MVs as no such fluorescent signals were detected in the PBS injected controls ( Fig 4 ) . These observations imply that GBS MVs could anterogradely move along the female reproductive tract and reach distant sites . To determine the in vivo effect of MVs , we injected MVs from strain A909 in individual mouse fetal sacs and 24 h post injection , mice were sacrificed and chorio-decidual membranes were collected . Histopathogically , in the tissues obtained from MV injected amniotic sacs , extensive disruption of chorio-decidual epithelium and cell sloughing was observed; no such changes were seen in tissues derived from PBS or BSA entrapped liposomes injected controls . The collagen distribution in the mice feto-maternal tissues was studied histologically using Picrosirius red . In the controls ( PBS or liposome injected groups ) , bundles of collagen fibers were detected all along the length of the sub-epithelial layer . In the tissues obtained from MV injected sacs extensive collagen fragmentation was observed , the collagen bundles appeared disrupted and a large number of disrupted collagen fibers were detected in the sub-epithelial zone ( Fig 5 ) . We next examined the effect of GBS MVs on apoptosis in chorio-decidual membrane . Large numbers of TUNEL positive cells were detected along the entire area of the chorio-decidua derived from amniotic sacs injected with MVs indicative of extensive apoptosis . Very few apoptotic cells were observed in the tissues derived from liposome or PBS injected sacs ( Fig 5 ) . Similar apoptotic effects were observed in HeLa cells following treatment with GBS MVs in vitro ( S3A Fig ) . Examination of the chorio-decidual tissue derived from amniotic sacs injected with MVs of GBS strain A909 revealed infiltration of neutrophils and lymphocytes underneath amniotic epithelium , a hallmark of chorio-amnionitis ( Fig 6A and 6B ) . Scoring of inflammation revealed presence of significantly higher numbers of neutrophils and lymphocytes in tissues derived following MV injections compared to PBS or liposome injected tissues ( Fig 6C and 6D and S4A and S4B Fig ) . A large number of F4/80 positive macrophages were also seen infiltrated in case of MV injected tissues ( Fig 7A ) . In case of control sacs ( injected with PBS or liposomes ) relatively fewer F4/80 stained macrophages and neutrophilic/leukocytic infiltration was observed ( Fig 7A and S4 Fig ) . We next examined if GBS MVs induce an inflammatory response in the host tissues . For this we injected MVs from GBS strain A909 in individual mouse fetal sacs on E14 . 5 and 24 h later the mRNA levels for selected inflammatory cytokines were estimated by qPCR in the decidua . In the decidua derived from MV injected sacs there was a significant increase in the transcript levels of Kc , Il-1β , Il-6 and Tnf-α compared to decidua derived from PBS injected amniotic sacs ( Fig 7B ) . However , the levels of Ifn-γ remained unchanged between the two groups . Amongst all the studied cytokines , a maximum increase was observed in the mRNA levels of Il-1β and Kc . The inflammatory response was specifically driven by injected MVs , as no difference in transcript abundance of inflammatory cytokines was observed in tissues derived from liposome compared to PBS injected sacs ( S4G Fig ) . Considering our findings on the widespread effect of GBS MVs on fetal membrane components we tested if MVs from GBS strain A909 are capable of inducing preterm birth . For this purpose amniotic sacs of mice were injected with either 5 or 10 μg of GBS MVs or PBS and animals were monitored every 6 h up to 72 h for signs of preterm birth ( vaginal bleeding , pups in cage ) . Administration of 5 μg and 10 μg of MVs in the amniotic fluid resulted in preterm births ( by day 18 of pregnancy ) of 55% and 68% of fetuses , respectively . Only 10% of such events were observed in the PBS treated group . Along with preterm birth , increased frequency of intrauterine fetal death ( IUFD , defined as resorption or presence of a macerated fetus in amniotic sac ) was observed in the group treated with GBS MVs as compared to PBS treated controls ( 36% and 29% vs 3% ) ( Fig 8A and S2 Table ) . Interestingly , in the groups where MVs were injected , fetuses were either resorbed or delivered by 48 h . Moreover , the delivered fetuses from MV injected sacs exhibited significant reduction in size and abnormal morphology ( Fig 8C ) . Collectively these suggest that GBS MVs lead to fetal demise , preterm birth and cause fetal injury .
Membrane vesicles are bilayered structures , found to be secreted almost ubiquitously by gram negative bacteria and a large number of gram positive bacteria [24 , 25 , 28 , 29 , 31–34] . In the current study , we demonstrate for the first time that GBS ( a gram positive bacteria ) is capable of secreting MVs in a strain independent manner . Employing different microscopic techniques we could not only detect the presence of vesicles in GBS culture supernatant , but also demonstrate the budding of vesicular structures from GBS surface . Interestingly , similar budding structures from GBS surface resembling MVs have been reported in GBS colonized in the murine genital tract , implying that GBS could also produce MVs in vivo [35] . We isolated and characterized these MVs both physically and biochemically and examined their pattern of interaction with host cells in order to explore their potential contribution in GBS pathogenesis . Intriguingly , we could extract DNA and amplify the cfb gene , an important virulence factor of GBS , using the MV DNA as template . Occasionally , DNA has been found to be associated with MVs from other microorganisms such as P . aeruginosa [36] , S . vesicuolsa [37] and C . perfringens [26] . Though DNA is detected in GBS MVs , we believe this DNA is fragmented as we could only detect the cfb gene , but genes representing other virulence factors ( cylE , pepB , zooA ) or housekeeping gene ( gapN ) could not be amplified . This implies that the DNA packaged in MVs is perhaps not a random phenomenon but may involve a regulated mechanism . While the functional significance of the DNA packaging in MVs is yet unknown , since the MVs fuse with the host or other bacterial cells , these might act as carriers of DNA from one cell to another [38] . It will be of interest to determine if the MV DNA specifically enters the cells to induce a pathological activation of host nuclease machinery . Proteomic analysis of the GBS MV proteins revealed presence of numerous ECM degrading enzymes . The preferential packaging of ECM degrading enzymes compared to other abundant membrane proteins implies presence of a selective sorting mechanism . Such a sorting mechanism have been identified in P . gingivalis which not only facilitates preferential packaging of important virulence factors but also enables it to exclude other abundant outer membrane proteins from the cargo [39] . Similar enrichment of acidic glycosidases and proteases were observed in Bacteroides species suggesting presence of a species specific machinery devoted to selectively pack proteins into the vesicles to do specific jobs , in this case securing nutrients for the benefit of the whole bacterial community present in the microbiota [40] . The presence of various ECM degrading enzymes in GBS MVs therefore points towards its significant role in GBS pathogenesis . Indeed , we observed that the MVs could not only interact with HeLa cells extracellularly but also internalize and cause cell death . These observations prompted us to hypothesize that the presence of various virulence factors and ECM degrading proteins in GBS MVs might lead to tissue degradation and cell death that may contribute to PPROM and preterm births . We next performed a series of experiments to investigate this hypothesis . The fetal membranes ( amnion and chorion ) rest upon a collagenous basement membrane of type II and IV collagen and beneath this layer lays a fibrous layer that contains collagen types I , III , V , and VI . Collagen , therefore provides major structural strength for the membranes and degradation of it leads to loss of membrane integrity [41–43] . The enrichment of ECM degrading proteases , coupled with the presence of gelatinolytic activity in GBS MVs suggested that these might cause loss of fetal membrane integrity by ECM degradation . Indeed , ex vivo treatment of chorio-decidual membrane with GBS MVs led to collagen fragmentation . This collagen fragmentation had profound effect on fetal membrane integrity as the stiffness of the membranes challenged with MVs were significantly reduced . It is likely that during pregnancy , GBS via its MVs lead to collagen degradation and reduction in tissue stiffness which together would resist further expansion of the amniotic sac , a prerequisite to accommodate the growing fetus . Thus our findings suggest that GBS MVs lead to loss of ECM and weaken the amniotic membrane making it susceptible to rupture upon pressure from the growing fetus . Epidemiological data suggest that colonization of vagina and cervix with GBS increases the probability of chorio-amnionitis . We suspected that the MVs produced by GBS while colonizing can move upwards in the female reproductive tract and affect cells and tissue at distant sites . Budding structures resembling MVs have been observed in GBS colonized in the murine genital tract [35] . Extending this data , herein we show that GBS MVs when vaginally instilled could traverse anterogradely upto anterior most segment of the uterus . Recently , it has been shown that GBS mutant for hyaluronidase is not capable of ascending infection in a mouse model [44] . Since , hyaluronidase is also enriched in our GBS MV preparations; it is possible that it could aid ascend of MVs into the female tract . While it needs to be demonstrated that like GBS , MVs can also anterogradely transport into the gravid uterus; it is tempting to hypothesize that GBS colonization at a distant site such as vagina can affect the feto-maternal tissues in the uterus by the virtue of secreting MVs . We next asked if GBS MVs have any pathogenic effects in vivo . Corroborating our ex-vivo observations; we observed that instillation of GBS MVs intra-amniotically led to extensive disruption of the chorio-decidual structure . This disruption is also associated with collagen fragmentation which could be due to the proteases present in the MVs . Beyond collagen , extensive alterations in the expression of genes encoding for extracellular matrix proteins has been reported in chorio-decidua of rhesus monkeys infected with GBS [16] . While it will be of interest to see the alterations in other ECM molecules in response to MVs , we believe that the changes in extracellular matrix and collagen degradation would contribute to the mechanical weakening of the tissue resulting in the loss of membrane stiffness . Indeed we did see a reduction on stiffness of the chorio-decidua tissue challenged with MVs ex vivo . It will be of interest to test the effects of MVs derived from GBS mutants for various ECM degrading enzymes and determine the mechanistic basis of this phenomenon . Along with loss of collagen , we also detected extensive apoptosis in the chorio-decidua derived from sacs injected with MVs . These observations suggested that coupled with mechanical weakening , cell death might further contribute to loosening of the membrane . Such tissue damage by MVs could make the membrane prone to rupture leading to PPROM . A leading feature of PPROM and preterm births due to inflammation is chorio-amnionitis . Chorio-amnionitis is acute inflammation of the fetal membrane and chorion which is typically caused due to ascending microbial infection especially in case of infection with genital mycoplasmas [45] . Histologically , chorio-amnionitis is characterized by leukocyte infiltration in the chorion/amnion and the decidua . In the context of GBS , in vivo administration of the live bacteria in pregnant mice led to histopathologic characteristics resembling chorio-amnionitis [16] . Herein for the first time we demonstrated that chorio-amnionitis can be caused by MVs even in absence of live bacteria . We observed that intra-amniotic administration of GBS MVs caused extensive leukocytic infiltration in the subchorion layer and in decidua resembling chorio-amnionitis . Beyond leukocytic infiltration , the decidua of mice injected with GBS MVs also had macrophage infiltration . Moreover , these vesicles also induced inflammatory cytokines such as , Kc , Il-1β , Il-6 , and Tnf-α in the decidua . Levels of these pro-inflammatory cytokines are also known to be elevated in the amniotic fluid of patients in independent contexts of GBS infections and PPROM [16 , 46] . Since HeLa cells also showed similar response upon treatment with MVs ( S3B Fig ) , we believe that the elevation in levels of these cytokines in the decidua may not be exclusively due to higher numbers of immune cells present in these tissues , but because of the ability of MVs to generate an inflammatory response in the decidual cells . We therefore conclude that GBS by the virtue of production of MVs activates the host immune system which might trigger the immune cell homing and activation leading to chorio-amnionitis . These observations are novel as for the first time we have shown that even in absence of active infection in the chorio-decidua , features resembling clinical chorio-amnionitis could be mimicked by MVs . Clinically this observation is highly relevant as 50–80% women with chorio-amnionitis do not have bacteria in their amniotic fluid or the decidual tissue [47–50] . Based on our findings we can hypothesize that MVs secreted by the pathogens residing in lower genital tract may be responsible for cases with unexplained chorio-amnionitis . It would be imperative to study the presence of MVs in amniotic fluid and chorio-decidual tissues of woman with culture negative chorio-amnionitis . Since , GBS MVs lead to tissue damage , inflammation and chorio-amnionitis , resulting in mechanical weakening of the fetal membrane , we finally asked if GBS MVs can lead to premature birth . Indeed , we observed that intra-amniotic injection of GBS MVs lead to preterm delivery . More than 50% of mouse pups were delivered preterm ( almost 2 days prior to their expected day of delivery ) when challenged with MVs . Finally , we also observed that like live GBS [17] , the MVs are also highly pathogenic to the fetus . Almost 30–40% of fetuses died in utero amounting to resorption and some fetuses were still born . The fetuses that were stillborn or recovered from uterine sacs after GBS MV challenge , were smaller and had major damage to its organs . These results together imply that GBS MVs can cause IUFD or preterm births . To our knowledge this is the first report describing the direct role of membrane vesicles produced by any pathogenic bacteria in disease pathogenesis . In the light of the fact that intrauterine infections can lead to autism like changes in the brain [51] , it is possible that GBS MVs might have a similar effect . It will be of interest to study the effects of sub-lethal dose of GBS MVs on fetal development and physiology . In summary , the results of the present study have shown that GBS MVs cause host cell death , membrane weakening and inflammation of the feto-maternal interface which causes preterm birth and IUFD . Coupling this with the fact that pathogen derived vesicles can function as vehicles for long distance delivery of virulence factors [52] , our results imply that the production of extracellular membrane vesicles serves not only as a tool for secretion but also arms GBS with an additional weapon by which while colonizing it can orchestrate events at distant sites including the fetal membrane . Our findings provide a plausible explanation for the occurrence of PPROM and premature delivery in woman with chorio-amnionitis without detectable bacteria in their amniotic fluid or the decidual tissue . We conclude that GBS utilizes MVs as a surrogate to spread its virulence factors in the host which are responsible for the clinical features of GBS infection during pregnancy . Prevention of vesicle biogenesis may therefore be a viable therapeutic option to prevent GBS mediated preterm birth .
All the experimental work on animals was done as per the guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals ( CPCSEA ) , India . The study protocol has been reviewed and approved by the Institutional Animal Ethics Committee of National Institute for Research in Reproductive Health ( NIRRH ) under the project number 10/15 . S . agalactiae strains A909 ( serotype IA ) , COH1 ( serotype III ) , NEM316 ( serotype III ) and 2603V/R ( serotype V ) ( kindly provided by Dr . Lakshmi Rajagopal , Department of Pediatric Infectious Diseases , University of Washington School of Medicine , Univ . of Washington , Seattle , USA ) were cultured in Todd Hewitt broth ( THB ) at 37○C . Unless otherwise stated MVs from GBS strain A909 was used for further experiments . MVs were purified from growing GBS cultures as per the standard protocol used for many other bacteria [25 , 27 , 32 , 53 , 54] . Briefly , GBS cells grown upto optical density ( OD600nm ) 1 . 2 , were harvested by centrifugation ( 12 , 000 x g , 30 min , 4○C ) and the supernatant was passed through 0 . 22μm filter . The filtrate was concentrated using Amicon ultrafiltration system ( 10 kDa ) and ultracentrifuged ( 150000 x g , 3 h , 4○C ) to pellet down MVs . The MV pellet was resuspended in PBS ( pH 7 . 2 ) and protein content was determined by Bradford assay following lysis of MVs by 0 . 05% Triton X-100 . To quantitate the number of MVs produced from GBS culture , MVs were labeled with 20 μM of Vybrant DiO cell labeling solution ( Molecular Probes ) and quantified by flow cytometry ( FACS Aria II , BD Biosciences , USA ) using fluorescent counting beads ( CountBright Absolute counting beads; Invitrogen ) as standards as described earlier [25] . The varying diameters and size distribution of MV preparations were measured using a Goniometer ( Brookhaven Instrument Co . , USA ) as described earlier [55] . For Scanning Electron Microscopy ( SEM ) , GBS cells were air dried , desiccated overnight and visualized with Field Emission Gun Scanning Electron Microscope ( JEOL , USA ) at an accelerating voltage of 5 kV . For Transmission Electron Microscopy ( TEM ) , MV samples were applied to Formvar/Carbon film coated 200-mesh copper grids ( Pacific Grid-Tech ) and negatively stained with 2 . 5% uranyl acetate followed by visualization under Transmission Electron Microscope ( FEI Technai , USA ) ( 120 kV ) . For Atomic Force Microscope ( AFM ) , bacterial suspension was loaded onto poly-L-lysine coated coverslips and visualized under an atomic force microscope ( Asylum Research , USA ) under contact mode at a scanning rate of 1 Hz using silicon nitride cantilevers . Lipids were extracted from intact GBS cells and MVs using protocol described earlier [25] . Following extraction , lipids were silylated using 1:1 ratio of N , O-Bis ( trimethylsilyl ) trifluoroacetamide ( BSTFA ) and pyridine at 75○C for 30 min . Silylated lipids were then analyzed using a GC-MS ( Agilent , USA ) fitted with a HP-5MS fused silica capillary column ( 30 m × 0 . 25 mm i . d . , 0 . 25 μm film thickness ) with Helium as carrier gas ( flow rate 1ml/min ) . Identification of compounds was based on mass spectra including comparison to a library ( NIST ) . 150 μg of MV proteins were separated on 12% SDS-PAGE . After staining with Coomassie Brilliant Blue R-250 ( CBB ) , bands were excised and processed for in-gel trypsin digestion [56] . The eluted oligopeptides were co-crystallized with CHCA ( 5 mg/ml ) and spectra were acquired using MALDI-ToF mass spectrometer ( UltraFlex III , Bruker Daltonics ) . Mascot ( Version 2 . 2 . 04 , Matrix Science ) searches were conducted using the NCBI non-redundant database with the following settings: 1 missed cleavage; Carbamidomethyl on cysteine as fixed modifications , methionine oxidation as variable modification and 100 ppm error ( 150 ppm error for band No . 4 ) . A match with S . agalactiae protein with the best score in each Mascot search was accepted as successful identification ( p< 0 . 05 ) . Subcellular locations were predicted by LocateP database and confirmed by pSORT and TMHMM algorithms . MVs were initially treated with 50 μg/ml DNaseI in presence of 10 mM MgCl2 at 37°C for 1 h to remove any surface bound DNA . Following heat inactivation ( 10 min at 80°C ) and lysis of MVs using Triton X-100 ( 0 . 05% ) at 37⁰C for 30 min , DNA was extracted by phenol-chloroform-isoamyl alcohol , precipitated using ammonium acetate as described earlier [26] . The purified DNA was quantified and used for PCR using primers specific for cylE , cfb , pepB , zooA and gapN genes ( S1 Table ) . Genomic DNA isolated from GBS strain A909 served as control template . HeLa cells ( procured from National Center for Cell Science , Pune , India ) were grown in DMEM media ( Invitrogen ) supplemented with 10% fetal bovine serum ( Invitrogen ) at 37○C in 5% CO2 . HeLa cells were incubated with increasing protein concentrations of MVs ( 10 μg/ml to 300 μg/ml ) for 24 h and the viability was assessed using MTT assay kit ( HiMedia , India ) . To study internalization , MVs were labeled with FITC ( Sigma ) in 0 . 1 M sodium carbonate buffer ( pH 9 . 0 ) for 1 h at room temperature and washed to remove unbound FITC [57] . FITC labeled MVs ( 30 μg MV protein ) were then allowed to interact with HeLa cells for 6 h . After washing to remove unbound MVs , cells were fixed and stained with mouse polyclonal anti-FITC antibody ( Invitrogen ) followed by an anti-mouse secondary antibody conjugated to AlexaFluor 555 ( Invitrogen ) . Images were acquired with an oil immersion Plan-Apochromat 63X/1 . 4 NA objective using a confocal laser scanning microscope ( Zeiss ) . 10 μg of MV proteins were separated ( 150 V , 5 h , 4°C ) on 12% SDS-PAGE with 0 . 1% co-polymerized gelatin . After treatment in 1% Triton X-100 , gelatinolysis was promoted by further incubation for 16 h at 37°C in developing buffer ( 50 mM Tris-HCl , 50 mM Tris base , 200 mM NaCl , 5 mM CaCl2 , 0 . 02% Brij 35 , pH 7 . 6 ) . The gel was then stained with Coomassie Brilliant Blue R-250 and destained to intensify the digestion halos . BSA entrapped liposomes were prepared by thin film hydration method [58] . Briefly , phosphatidylcholine ( 10 mg ) was dissolved in a mixture of chloroform and methanol ( 2:1 ) and the solvent was evaporated under vacuum using a rotary evaporator to form a thin layer of lipid film in a round bottom flask . The dried lipid film was hydrated using 1 ml of PBS containing BSA ( 10 mg/ml ) at 45°C for 1 h to form multilamellar vesicles ( MLV ) . The liposome suspension was then sonicated using a probe sonicator at 40% amplitude to obtain unilamellar vesicles . Chorio-decidual membrane was collected from pregnant mice on E14 . 5 . Immediately after dissection , the membranes were incubated in DMEM ( Invitrogen ) containing 10% FBS ( Invitrogen ) at 37°C and treated for 24 h with either PBS , or BSA entrapped liposomes ( 100 μg/ml ) or MVs ( 100 μg/ml ) in presence or absence of protease inhibitor cocktail ( Sigma ) . Collagenase ( 10 μg/ml ) was used as positive control . Following incubation , the membranes were washed , carefully spread on a slide layered with double adhesive tape , allowed to air dry for 5 min . Subsequently , the membranes were hydrated with PBS , and their mechanical properties were probed with an Atomic Force Microscope ( AFM ) . A 5 μm diameter spherical probe with a nominal spring constant of 32 . 66 pN/nm was used . Using a custom-written code , force-indentation curves were fitted with Hertz model to estimate the Young’s modulus of elasticity of each of the membranes . Each sample was probed randomly multiple number of times ( greater than 50 ) at multiple different positions to estimate average stiffness of the membranes . The vagina of mice in estrus phase were flushed thrice with 40 μl of 0 . 2% Triton X-100 in 0 . 9% saline followed by 40 μl 0 . 9% saline . MVs were labeled with FITC as described earlier [57] and 100 μg of FITC labeled GBS MVs in 100 μl PBS were vaginally instilled in 25–30 μl aliquots using a micropipette . Control mice received PBS only . After 6 h mice were euthanized and the reproductive tract ( cervix to uterus ) was collected . Tissues were briefly fixed in 4% paraformaldehyde and mounted on slides with Vectashield containing DAPI ( Vector Laboratories ) . Images of different parts of the tissue such as utero-cervical junction , distal uterus and proximal uterus were captured with an oil immersion Plan-Apochromat 40X/1 . 3 NA objective of a confocal laser scanning microscope . For pregnancy-outcome experiments , C57BL6/J- were bred and maintained at the Experimental Animal Facility of NIRRH under constant temperature and 12 h light and dark cycles were used . Female mice in estrus were impregnated naturally by a male of proven fertility and mating was confirmed by the presence of a vaginal plug ( E0 . 5 ) . Intra-amniotic injections were performed on E14 . 5 of a 19–20 d gestation . Briefly , animals were anesthetized and a 1 . 5-cm midline incision was made in the lower abdomen . The mouse uterus is a bicornuate where the fetuses are arranged in a “beads-on-a-string” pattern . Both the horns were exposed and individual fetal sacs were injected with 100 μl of MVs ( 5 or 10 μg protein ) or equivalent amount of PBS or BSA-liposomes . After injection the uterus was returned to the abdomen , muscle and skin layers were sutured and dams were returned to their cages and monitored on regular intervals . Surgical procedures lasted ∼10 min and post-operative care was taken as per standard protocols at NIRRH . Delivery of one or more pups in the cage or lower vagina within 48 h was considered preterm . For collection of tissues , animals were euthanized 24 h after surgery . The inoculated horns were incised longitudinally along the anti-mesenteric border . Gestational tissues ( full-thickness biopsies from the middle region ) and fetal membranes were harvested and frozen in Trizol reagent ( Invitrogen ) and stored at −80°C for RNA extraction or fixed in 4% paraformaldehyde for histopathology . Fixed tissues were paraffin embedded and 5 μm thick paraffin sections were collected on poly-lysine coated glass slides and processed for routine hematoxylin and eosin staining . RNA from HeLa cells or mouse tissue was isolated using Trizol reagent as detailed previously [59] . RNA was treated for 30 min with DNaseI to remove any DNA contamination and processed for reverse transcription . The details of qRT-PCR have been described previously [59] . Briefly , 1 μg of RNA was reverse-transcribed using SuperScriptTM First-Strand Synthesis System ( Invitrogen ) . The cDNA was further used for quantitative RT-PCR ( qRT-PCR ) for various cytokines . β-actin ( for HeLa cells ) and 18S ( for mouse tissues ) were used as the house keeping genes . The sequences of the primers are as mentioned in S1 Table . Care was taken to design primers that spanned an intron to eliminate any amplification due to genomic DNA contamination . The specificity of amplicons was confirmed by performing dissociation / melt curve analysis . Only those primer pairs that resulted in a single sharp melt peak with a consistent melt temperature were included in the study . The relative changes in the expression of above genes was analyzed by 2-ΔΔCt method [60] . Hematoxylin and Eosin stained slides of the chorio-decidua were examined under 40X objective for neutrophils and lymphocytes . The cells were identified based on their morphology . The number of neutrophils and lymphocytes per field were counted in 8–10 random fields per section . Five random sections from each fetus were analyzed . The analysis was done in three biological replicates . Immunohistochemistry on tissue sections was performed as detailed previously [61] . For detection of macrophages , paraffin embedded 5 μm sections of the chorio-decidual membrane ( with or without MV treatment ) were stained with rabbit polyclonal anti-F4/80 Ab ( 1:100; Santa Cruz Biotechnology ) and detected using HRP conjugated goat anti-rabbit secondary antibody ( 1:100; Dako ) and 3 . 3’ Diaminobenzidine with H2O2 . Sections were counterstained with hematoxylin and mounted in DPX . Total collagen staining in chorio-decidual membrane sections were performed using Picrosirius red [62] . Paraffin sections were deparaffinized , rehydrated in graded methanol series and stained with hematoxylin followed by 0 . 5% Direct Red 80 ( Sigma ) prepared in picric acid . Following dehydration with 100% methanol and clearing in xylene , the slides were mounted with DPX . Induction of apoptosis by GBS MVs was analyzed by In Situ Cell Death Determination Kit ( Roche ) according to manufacturer’s protocol . Briefly , following deparaffinization , tissue sections were rehydrated and digested using proteinase K . After permeabilization with 0 . 25% Triton X 100 , the sections were incubated with terminal deoxynucleotidyl transferase and fluorescein-dUTPs for 1 h at 37°C . Following washing to remove unbound dUTPs , the sections were mounted with Vectashield containing DAPI ( Vector Laboratories ) . Images were acquired with an oil immersion Plan-Apochromat 40X/1 . 3 NA objective of confocal laser scanning microscope ( Zeiss ) . Graphpad Prism version 4 . 03 was used for statistical analysis . Statistical tests undertaken for individual experiments are mentioned in the respective figure legends . Statistical significance was accepted at p< 0 . 05 .
|
Preterm birth is a major health concern globally as it is not only a leading cause of neonatal death , but also has long term consequences including defective brain development . Infection of vagina and cervix of pregnant women with the bacteria , Group B Streptococcus ( GBS ) , causes chorio-amnionitis that significantly increases the probability of preterm births . We report that , GBS produces small extracellular membrane vesicles ( MVs ) which are toxic to both fetal and maternal cells . In animal studies , we found that the MVs disrupt the connective tissue of the fetal membrane reducing its mechanical strength which may cause premature rupture of amniotic sac . Further we show that even in absence of the bacteria , the MVs directly led to extensive inflammation in the mouse resulting in chorio-amnionitis , preterm births and still births . Collectively , our findings reveal how GBS while colonizing the lower genitourinary tract might orchestrate events at the fetal membrane leading to premature birth .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"uterus",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"vesicles",
"pathology",
"and",
"laboratory",
"medicine",
"maternal",
"health",
"obstetrics",
"and",
"gynecology",
"hela",
"cells",
"immunology",
"biological",
"cultures",
"collagens",
"preterm",
"birth",
"women's",
"health",
"signs",
"and",
"symptoms",
"microscopy",
"pregnancy",
"cell",
"cultures",
"cellular",
"structures",
"and",
"organelles",
"pregnancy",
"complications",
"birth",
"research",
"and",
"analysis",
"methods",
"scanning",
"probe",
"microscopy",
"inflammation",
"atomic",
"force",
"microscopy",
"proteins",
"liposomes",
"cell",
"lines",
"immune",
"response",
"biochemistry",
"diagnostic",
"medicine",
"cell",
"biology",
"anatomy",
"biology",
"and",
"life",
"sciences",
"cultured",
"tumor",
"cells"
] |
2016
|
Membrane Vesicles of Group B Streptococcus Disrupt Feto-Maternal Barrier Leading to Preterm Birth
|
Wolbachia is currently being developed as a novel tool to block the transmission of dengue viruses ( DENV ) by Aedes aegypti . A number of mechanisms have been proposed to explain the DENV-blocking phenotype in mosquitoes , including competition for fatty acids like cholesterol , manipulation of host miRNAs and upregulation of innate immune pathways in the mosquito . We examined the various stages in the DENV infection process to better understand the mechanism of Wolbachia-mediated virus blocking ( WMVB ) . Our results suggest that infection with Wolbachia does not inhibit DENV binding or cell entry , but reduces virus replication . In contrast to a previous report , we also observed a similar reduction in replication of West Nile virus ( WNV ) . This reduced replication is associated with rapid viral RNA degradation in the cytoplasm . We didn’t find a role for host miRNAs in WMVB . Further analysis showed that the 3’ end of the virus subgenomic RNA was protected and accumulated over time suggesting that the degradation is XRN1-mediated . We also found that sub genomic flavivirus RNA accumulation inactivated XRN1 in mosquito cells in the absence of Wolbachia and led to enhancement of RNA degradation in its presence . Depletion of XRN1 decreased WMVB which was associated with a significant increase in DENV RNA . We also observed that WMVB is influenced by virus MOI and rate of virus replication . A comparatively elevated blocking was observed for slowly replicating DENV , compared to WNV . Similar results were obtained while analysing different DENV serotypes .
Dengue is the most important mosquito-transmitted viral disease of humans in terms of the global burden of disease[1] . Current control methods focus almost entirely on vector control through either preventative source reduction or insecticide spraying in response to outbreaks . The increasing global incidence of dengue is a testament to the shortcomings of the current approach to dengue control . The introduction of the endosymbiotic bacterium Wolbachia pipientis into Ae . aegypti has been shown to interfere with the replication of RNA viruses like dengue ( DENV ) , Chikungunya virus ( CHIKV ) , Yellow fever virus ( YFV ) , West Nile virus ( WNV ) , Semliki Forest Virus ( SFV ) and Zika virus [2–6] and thus potentially reduce their transmission by mosquitoes . A number of mechanisms have been proposed to contribute to Wolbachia-mediated virus blocking ( WMVB ) . The presence of Wolbachia has been shown to up regulate reactive oxygen species ( ROS ) -dependent activation of Toll pathway genes and associated anti-microbial effectors as well as genes involved in melanization and methyltransferase [7–9] . Studies in D . melanogaster flies and cell lines , however , have shown that WMVB is independent of the Toll , Imd and RNAi pathways , indicating that immune activation is not required for blocking , though it may enhance it [5 , 10 , 11] . Alternatively , Wolbachia may compete with viruses for key host intracellular molecules such as fatty acids , especially cholesterol or amino acids , thus reducing viral replication [12 , 13] . Finally , WMVB may be mediated via manipulation of the expression of host miRNAs . The miRNA aae-mir-2940 , for example , is highly expressed in both mosquitoes and cell lines infected with Wolbachia and reduces the expression of AaDnmt2 and induced the expression of metalloprotease gene that affects the replication of Wolbachia and viruses [14 , 15] . On the other hand , Wolbachia is able to block viral replication in D . melanogaster Jw18 cells without upregulating host miRNAs [5] . The exact mechanism responsible for WMVB remains unknown . Previous studies have typically examined the response of the host to the presence of Wolbachia to attempt to dissect the mechanism of WMVB . Here , we have instead investigated the fate of the virus itself . We tracked the different stages of virus replication including its binding to cellular receptors , internalisation , replication and egress in the Ae . aegypti-derived Aag2 cell line [16 , 17] . Our study showed that WMVB is not associated with inhibition of virus binding or internalisation . We further analysed the fate of viral RNA and found that it degrades and that the 3’ end of the virus subgenomic RNA accumulates over time , indicating the potential involvement of exoribonuclease XRN1 . sfRNA accumulation also inhibited XRN1 activity and in turn enhanced XRN1 mediated degradation in the presence of Wolbachia . The presence of Wolbachia does , however , efficiently block the replication of WNV and all DENV serotypes . This blocking , however , was dependent on the virus MOI and the rate of viral genome replication .
DENV enters a host cell by binding to its receptors followed by endocytosis [18 , 19] . We speculated that Wolbachia-mediated receptor blocking or change in receptor expression could decrease viral binding and internalisation . We tested each of these steps separately . To test whether the presence of Wolbachia affects the number of virus particles bound to cellular receptors , we incubated DENV with Aag2 and Aag2 cells containing the wMel strain of Wolbachia ( Aag2-wMel ) at 4°C for 1 hour . At this temperature , the virus can bind to cellular receptors , but is not internalised [20 , 21] . After washing away unbound viral inoculum , we quantified DENV RNA ( Fig 1A ) and found that levels of DENV do not differ significantly on cells with and without Wolbachia . This suggests that the presence of Wolbachia does not influence the attachment of virus to cells . The experiment was repeated at different MOIs of DENV , and also with WNV , and similar results were obtained ( S1 Fig ) in all cases . We then tested whether Wolbachia affects DENV internalisation by raising the temperature to 25°C . At this temperature , the virus enters cells through receptor-mediated , clathrin-dependent endocytosis , which allows for virus uncoating and replication [18 , 19] . We quantified intracellular DENV RNA at 25 minutes and 1 hour post-infection and found viral levels were similar in Aag2 and Aag2-wMel cells at both time points ( Fig 1A ) , indicating that virus internalisation and early replication was not affected by Wolbachia . After internalisation by endocytosis and uncoating , the DENV RNA genome is released into the cytoplasm , and the process of viral RNA replication starts at the rough endoplasmic reticulum . We tracked viral replication by quantifying DENV RNA at multiple time points across 8 days’ post-infection . Significant differences in DENV RNA levels between Aag2 and Aag2-wMel cells became apparent almost immediately ( Fig 1B ) . As early as 0 . 5 days post-infection ( dpi ) , cell lines with Wolbachia contained ~2- to ~5-fold less intracellular DENV RNA copies than Aag2 cells . By day 5 , the difference between cell types had increased to up to 388-fold ( Fig 1B ) . This pattern of DENV increase in Aag2 cells and decrease in Aag2-wMel cells is also observed for negative strand RNA ( Fig 1B ) . In parallel , DENV titre in the media , after its egress from the cell , reached ~106 PFU/ml in Aag2 cells , but only ~102 PFU/ml in the presence of Wolbachia ( Fig 1C ) . Analysis of DENV infected cells using immunofluorescence microscopy also shows that Wolbachia reduces DENV replication ( Fig 1D ) . This is evident from the absence of non-structural protein 1 ( NS1 ) in the presence of Wolbachia in Aag2-wMel cells and further supports the observation that Wolbachia strongly diminishes DENV replication . Increasing differences in DENV levels were not only due to replication of DENV in Aag2 cells , but also to a steady decrease in DENV in Aag2-wMel cells over time . At 0 . 5 dpi , DENV levels in Aag2-wMel cells were ~6-fold less than the primary inoculum at 0 dpi . By days 5 and 8 post-infection , DENV levels were ~40 to 60-fold lower than at day 0 , respectively ( Fig 1B ) . These results suggest that the viral RNAs which are unable to engage in replication are subjected to degradation in the presence of Wolbachia . The experiments were also repeated at an MOI of 1 and similar results were observed ( S2A Fig ) . To confirm that above results are not cell-line specific , we repeated the experiment using RML-12 cells of Aedes albopictus origin , and observed a similar steady decrease of DENV in RML-12-wMel cells over time ( S3 Fig ) . To further analyse the degradation process of viral RNA , we quantified DENV RNA using primers that span different regions of the DENV genome ( Fig 2A ) . The increase in all regions of DENV genomic RNA correlates well with an increase in viral replication after 1 dpi in Aag2 cells ( Compare Fig 1B with Fig 2B ) . In contrast , in the presence of Wolbachia , there was a substantial and consistent decrease in the levels of all regions of the DENV genome over time ( Fig 2C ) . There were no significant differences in the rate of accumulation of any region of the DENV genome , with the exception of 0 . 5 dpi in Aag2 cells , where the levels of 3’ UTR were significantly higher than other regions ( Fig 2B ) . We also observed minor non-significant increase in 3’ UTR at all time points in Aag2-wMel cells ( Fig 2C ) . The 3’ UTR of flavivirus genomes produces subgenomic flavivirus RNA ( sfRNA ) as a by-product of incomplete RNA digestion by XRN1[22–27] . This Subgenomic RNA accumulates in cells and is an important factor in flavivirus pathogenicity [28] . Since the primers used above could detect only genomic RNA , we next examined levels of 3’ UTR at different time points using primers that detect both gRNA and sfRNA . Consistent with our previous results , DENV 3’ UTR gRNA levels increased in Aag2 cells over time , but steadily decreased in Aag2-wMel cells ( Fig 2D ) . In contrast , sfRNA levels increased over time in both cell types , although the increase was far less dramatic in the presence of Wolbachia ( only ~6-fold compared to ~26-fold higher at 8 dpi; Fig 2E ) . This happens due to incomplete RNA digestion by XRN1 were sfRNA is not digested . As a result , levels of sfRNA increase over time , but not the gRNA , which is sensitive to XRN1 and gets degraded . It should also be noted that WMVB does not completely inhibit virus replication as is evident from the low levels of positive and negative strands ( Fig 1B ) throughout the time course . This also explains the reduced accumulation of sfRNA in Aag2-wMel cells . Accumulation of sfRNA in DENV-infected cells can lead to inactivation of the exoribonuclease XRN1 , thereby increasing the overall stability of viral and host RNAs [29] . The much lower accumulation of sfRNA in Aag2-wMel cells compared to Aag2 cells raises the possibility that reduced levels of sfRNA might not be sufficient to inhibit XRN1 in infected cells . If so , this could contribute to a positive feedback cycle supporting Wolbachia’s ability to reduce dengue RNA levels . To determine the effect of Wolbachia on RNA decay , we tested degradation of host mRNA during dengue infection [29 , 30] . Aag2 and Aag-wMel cells , mock or infected with DENV , were treated with actinomycin D to stop cellular transcription . We then estimated the half-life of the transcripts of two host genes , ECR and La , which have previously been studied in the insect cells for mRNA decay [31] . In mock-infected Aag2 cells , the half-life of La and ECR ( 3 . 60 ± 0 . 28 and 1 . 13 ± 0 . 17 hours , respectively; Fig 3A–3D ) were similar as previously reported [31] . DENV infected Aag2 cells showed approximately twofold increase in the half-life of both La and ECR transcripts reaching 7 . 63 ± 0 . 98 and 3 . 49 ± 0 . 40 hours , respectively ( Fig 3A–3D ) consistent with an increased accumulation of DENV-derived sfRNA inhibiting XRN1 . However , in DENV infected Aag2-wMel cells , this increase in half-life of host RNAs was not observed , suggesting that DENV is unable to inhibit Xrn1 activity in the presence of Wolbachia . It should be also noted that inhibition of XRN1 stabilises various cellular RNAs and results in changes to cellular gene expression . This may be also a strategy by flaviviruses to escape cellular surveillance mechanisms that can otherwise inhibit their replication [29 , 30] . XRN1 is known to be associated with DENV replication regulation and its downregulation can positively enhance virus replication [32] . To investigate this further , we directly tested the effect of DENV RNA following siRNA knockdown of XRN1 . We observed ~70% reduction in XRN1 mRNA levels when XRN1 specific siRNAs were used ( Fig 3E ) . In Aag2 cells , XRN1 knockdown resulted in ~ 2-fold increase in DENV RNA ( Fig 3F ) as expected . Whereas XRN1 knockdown showed a greater increase in DENV RNA levels ( ~4-fold , p < 0 . 0001 ) in Aag2-wMel cells ( Fig 3F ) . These results further support our observation that XRN1 activity is not fully blocked in Aag2-wMel cells . Thus , an interaction between Wolbachia , viral RNA and XRN1 may be central to WMVB . Even though lower levels of sfRNA accumulated in DENV-infected Aag2-wMel cells , we observed a considerable amount of degradation of the viral 3’ UTR region ( Fig 2C ) . This could be caused by other RNA-degrading factors , including miRNAs even if RNA is eventually degraded by XRN1 and exosome [33] . Previous studies have suggested that upregulation of some host miRNAs is an important factor in WMVB [14 , 15] . We used qPCR to analyse the expression of a number of host miRNAs that have either been previously reported to be upregulated in the presence of Wolbachia [14 , 15] or which we identified as having the potential to bind to the 5’ or 3’ UTR of the DENV genome ( S2 Table ) . We saw no significant change in the expression of any of these miRNAs in the presence of Wolbachia ( Fig 4 ) , suggesting that they are not involved in Wolbachia-mediated inhibition or degradation of viral RNA . Our results support similar observations recently reported for Drosophila cells , in which host miRNA expression was not affected by Wolbachia [5] . As it has previously been reported that intracellular West Nile virus replication is enhanced by Wolbachia [4] , we also investigated the replication of this virus in our system . As expected , after infecting Wolbachia-free cells with an MOI of 10 , the titre of WNV of both polarities increased over time , with RNA levels peaking at 5 dpi ( Fig 5A ) . In contrast to the previous report [4] , however , we did not see an increase in WNV RNA levels in the presence of Wolbachia . Instead , we observed that intracellular WNV RNA copies were reduced for both positive and negative strands ( ~7-fold and ~6-fold reduction at 8 dpi compared to 0 dpi ) . The difference in positive strand WNV RNA levels between cells with and without Wolbachia was ~46-fold at 1 dpi and ~4×105-fold at 8 dpi . Analysis of cell culture supernatant showed similar results , with WNV titre reaching ~108 PFU/ml at 5 dpi in Aag2 cells , but only ~105 PFU/ml in Aag2-wMel cells . We observed differences between DENV and WNV in the extent of its replication ( compare Figs 1B , 1C , 5A and 5B ) . Early inhibition at 0 . 5 dpi in the presence of Wolbachia , as seen for DENV , was not seen for WNV at the same MOI . Similarly , we saw only ~7-fold overall reduction from the initial inoculum of WNV in Aag2-wMel cells at 8 dpi , compared to a ~60-fold reduction of DENV . Similar results were observed at different MOIs ( S2C–S2F Fig ) . WNV reached a titre of ~108 PFU/ml in Aag2 cell culture supernatant after its egress , compared to ~106 PFU/ml with DENV on 5dpi . This difference was further analysed by comparing the WMVB of DENV and WNV through one phase decay up to 8 dpi . Even though both viral RNAs were degraded during WMVB , the time to reach 50% viral RNA reduction differed ( Fig 5C and 5D ) . A 50% reduction of DENV RNA was seen as early as ~ 6 hours compared to ~ 16 hours for WNV . This can be attributed to the difference in the rate of replication of both viruses . In DENV , due to its slower replication , new viral RNAs are not synthesised at the same rate as in WNV . This results in fast depletion of viral RNA resulting in 50% degradation happening as early as ~ 6 hours . RNA degradation occurs similarly in WNV but due to its comparatively fast replication , new RNA strands are synthesised at the same time . This also further extends the 50% reduction time for WNV RNA from primary inoculum . We next wanted to confirm variation in WMVB and its relationship with intrinsic replication rate by comparing different strains of DENV virus from different serotypes , which might differ in the rate of replication but still utilise a similar replication mechanism . We first analysed the intrinsic replication rate of DENV serotypes 1–4 , all passaged up to 7 times in C6/36 cells and infected Aag2 and Aag2-wMel cells at an MOI of 1 . Analysis of virus copy number at different time points showed that virus replication was different for all these four serotypes ( S4 Fig ) . Regardless of serotype differences in DENV , they all were inhibited in the presence of Wolbachia . DENV-3 ET which reached the highest titre of ~107 genome copies/ ml showed the lowest WMVB ( S4 Fig ) . To analyse this further , we then compared DENV-1 Viet , DENV-2 NGC and DENV-3 08/09 along with DENV 1–4 ET isolates . All the virus infection , RNA isolation and quantification were carried out at the same time . Analysis of virus copies from cell culture supernatant and intracellular RNA after infecting Aag2 and Aag2-wMel cells shows the extent of intrinsic virus replication ( Fig 5E and 5F ) . Viral RNA in cell culture supernatant reached a titre of ~107 genome copies/ ml for DENV-3 ET , DENV-2 NGC and DENV-3 08/09 compared to DENV-1 ET , DENV-2 ET , and DENV-4 ET and DENV-1 Viet which reached a slightly low titre of ~106 genome copies/ ml . Again , regardless of serotype difference , all seven DENV isolates were inhibited in the presence of Wolbachia with a higher viral inhibition observed among DENV-1 ET , DENV-4 ET and DENV-Viet which were below the detection limit of 102 genome copies/ ml in cell culture supernatant . Intracellular analysis of viral RNA also showed similar results ( Fig 5F ) . The greatest difference in blocking was observed between DENV-3 ET and DENV-4 ET ( *** p < 0 . 0001 , Fig 5G ) which , grew to the highest and lowest copies respectively . These results show that Wolbachia blocks DENV irrespective of its serotype and that the WMVB is influenced by the intrinsic replication rate .
The insect endosymbiont Wolbachia is currently being used as a tool to reduce transmission of dengue and other Aedes transmitted viruses [34] . Even though various studies have been carried out to understand the exact mechanism behind WMVB [7 , 8 , 12 , 14] , none has examined the detailed virus infection process in the Wolbachia context . DENV , a positive-stranded RNA virus enters the host cell through host receptors and releases its RNA into the cytoplasm where it is translated by host ribosomes . The translated products include polyproteins encoding the replication machinery used to produce positive and negative RNA strands [35 , 36] . In the current study , we saw no effect of Wolbachia on the early stages of virus binding or cellular internalisation . Instead , we found that Wolbachia reduced virus replication that is followed by enhanced viral RNA degradation . RNA viruses use different strategies to stabilise transcripts and evade the cellular RNA decay machinery and so maintain a continuous infection . In flaviviruses , some of the RNAs produced are degraded through a 5’ to 3’ exonucleolytic pathway [37 , 38] . This process leads to an accumulation of sfRNA due to incomplete degradation . This XRN1-mediated degradation typically stalls near the 3’UTR , and leads to inhibition of the XRN1 enzyme , as it remains bound to sfRNA . This inhibition has been reported before in mammalian cells [29 , 30] and now in insect cells to cause increased stabilisation of cellular mRNA and thus giving a better environment for the survival of viral RNAs . In the presence of Wolbachia , we observed that DENV replication is inhibited and thus doesn’t lead to accumulation of sfRNA making the cells not ideal for virus growth , as an active XRN1 enzyme will be regulating the RNA level . We are not assuming that XRN1 mediated viral RNA degradation is the only mechanistic explanation for WMVB but our results suggest that it plays an important part . The role of other decay factors associated with XRN1 including UPF1 , SMG5 , SMG6 and SMG7 also cannot be ruled out [39] . It is now well established along with this work that Wolbachia inhibits replication of most of the positive stranded RNA viruses . This is not the case with negative stranded RNA viruses such as Phasi Charoen-like bunyavirus which are unaffected by the presence of Wolbachia [40] . It is also known that depletion of RNA decay factors like XRN1 and UPF1 doesn’t have any effect on infection of negative stranded RNA viruses of the Paramyxoviridae and Bunyaviridae families [41–43] . Furthermore , they cap-snatch the 5′ end of host mRNAs using a virally encoded endonuclease . This makes the 5′ end of the viral mRNA indistinguishable from endogenous mRNAs and helps in protection from degradation [43 , 44] . These differences between positive and negative stranded viruses and the correlation with WMVB further suggests the significance of viral RNA decay in contributing to WMVB . Comparison of the extent of WMVB by comparing WNV with DENV or by comparing different strains of DENV that show variation in replication rate and subsequently the titre they attain at a given time point , shows that the degree of WMVB is highest with slower replicating viruses . Ideally , RNA viruses should replicate quickly and be recruited to cellular membranes to evade the host RNA decay machinery . A slowly replicating virus is more prone to RNA degradation [41 , 45 , 46] , which can occur rapidly: the half-life of a cellular mRNA is on average only hours [47] . Hence , slowed replication in the presence of Wolbachia should make viral RNA more prone to host RNA decay mechanisms , resulting in viral RNA degradation . At higher viral MOI , the presence of Wolbachia may prevent an equally high rate of replication , leading to sudden viral RNA degradation as we observed for both DENV and WNV . Thus , it is possible that Wolbachia may block slowly replicating virus more efficiently than viruses that replicate rapidly to higher titre . We predict , therefore , that a slowly replicating virus should be more severely affected by WMVB . As an example , DENV , which grows relatively slowly and to lower titre ( ~106 PFU/ml ) is more efficiently blocked by Wolbachia compared to WNV , which grows robustly and to a titre of ( ~108 PFU/ml ) . In addition to the extent of virus replication , WMVB is directly affected by Wolbachia density in the host [48 , 49] . It is thus possible that competition for shared cellular resources may be exacerbated at higher densities , increasing the negative impact of Wolbachia on viral replication and subsequent RNA degradation by XRN1 .
Aag2-wMel cells containing the wMel strain of Wolbachia were prepared by infecting Aag2 cells with wMel ( S1 Text ) . Aag2-Tet ( hereafter referred as Aag2 ) was prepared by treating Aag2-wMel cells with tetracycline [49] . Cells were maintained in complete medium containing 44% Schneiders media ( Life Technologies ) , 44% M&M ( CaCl2 0 . 151 gm , MgCl2 0 . 047 , KCl 0 . 2 gm , NaCl 7 gm , NaH2PO4 0 . 174 gm , Glucose 4 gm , Yeast extract 5 gm , Lactalbumin hydrolysate 6 . 5 gm , and NaHCO3 0 . 12 gm per litre ) 10% fetal bovine serum ( FBS ) ( Gibco ) and 1% penicillin-streptomycin solution ( Gibco ) . C6/36 cells of Aedes albopictus origin were supplied by the European Collection of Cell Cultures ( ECACC; Salisbury , United Kingdom ) and was purchased from CellBank Australia ( Westmead , NSW , Australia ) and were grown in RPMI media ( Gibco ) containing 10% FBS , 1X glutamax ( Gibco ) , 20 mM HEPES ( pH 6 . 9 ) and 1% penicillin-streptomycin solution ( Gibco ) . Vero cells of African green monkey kidney origin were supplied by the ECACC and was purchased from CellBank Australia and were grown in DMEM medium ( Gibco ) containing 5% FBS and 1% penicillin-streptomycin solution ( Gibco ) . Cell lines were routinely screened for any Mycoplasma contamination . Wolbachia densities in cell lines were measured by qPCR and FISH ( S1 Text , S6 Fig ) . Cell lines were also screened for any other flavivirus infection as described before ( S1 Text , S7 Fig ) Dengue virus ( DENV ) 2 ET 300 was passaged up to 7 times in C6/36 cells , and West Nile virus ( WNV ) ( Kunjin subtype ) of unknown passage history was further propagated in C6/36 cells [50] . The viruses used for infection were produced after infecting C6/36 cells containing RPMI media ( Gibco ) with 2% FBS , 1X glutamax ( Gibco ) , and 20 mM HEPES ( pH 6 . 9 ) , and harvested 5 dpi . The harvested virus was aliquoted and stored at -80°C until use . Virus titre in cell culture media was measured by plaque assay as reported previously [51] . Briefly , Vero cells were seeded towards 90% confluency and infected with the virus of different dilutions . 1% carboxymethyl cellulose containing DMEM medium , 2% FBS and 1% penicillin-streptomycin was added as an overlay and incubated at 37°C for 5–7 days . Plaques were visualised by crystal violet staining . Aag2 and Aag2-wMel cells seeded in 12 well plates were incubated at 4°C for 5 minutes , followed by washing once with cold PBS . Cells were infected with DENV-2 while kept on ice at specified MOI and further incubated for 1 hour at 4°C with gentle rocking every 15 minutes . Virus inoculum was then removed and the cells washed three times with cold PBS . Cells were then collected and stored in -80°C till use . For the internalisation study , media at 25°C was added to the cells followed by continued incubation at 25°C . Cells were washed at each time point with acid glycine ( pH 3 ) followed by PBS ( pH 7 . 4 ) wash . Cells were collected at different time points and stored in -80°C until use . All other virus infection experiments were carried out by infecting cells at specified MOI and incubating for 1 hour at 25°C followed by washing with PBS ( pH 7 . 4 ) three times and continued incubation at 25°C . Unless otherwise stated , all RNA extraction was carried out from cells using Trizol ( Life Technologies ) , according to manufacturer’s instructions . To synthesise cDNA , RNA was DNase ( Roche ) treated and converted to cDNA using SuperScript III Reverse Transcriptase ( Invitrogen ) according to manufacturer’s instructions . RNA integrity was analysed through bioanalyser ( Micromon , Monash University , Clayton ) ( S5 Fig ) . qPCR was run using the LightCycler 480 Probes Master ( Roche ) in a LightCycler 480 real-time PCR machine . Strand-specific analysis of DENV was done as described previously [52] using LightCycler 480 SYBR Green I Master ( Roche ) with 45 cycles followed by melt curve analysis and was normalised with RPS-17 . To analyse West Nile virus RNA strands , positive strand cDNA was synthesised using primer TAGWNKUNJ-E-R and for the negative strand with TAGWNKUNJ-E-F . qPCR was performed using LightCycler 480 SYBR Green I Master ( Roche ) and with primers Tag_only and WNKUNJ-E-F for the positive strand . To detect negative strands , primers Tag_only and WNKUNJ-E-R were used . The PCR conditions were as follows: Step 1 , 95°C for 5 min; Step 2 , 95°C for 10 s , 60°C for 10 s , 72°C for 10 s with 45 cycles; followed by melt curve analysis . Both strands were normalised with RPS-17 . Aag2 cells grown on coverslips in 6 well plates were infected with DENV at an MOI of 5 . 24 hpi , cells were fixed with 4% formaldehyde in PBS for 10 min and then permeabilised by incubating with 4% formaldehyde and 0 . 2% triton-X-100 . Cells were then blocked with 1% BSA in PBS-T ( 0 . 2% Triton X-100 ) for one hour at room temperature and probed with anti-NS1-4G4 mouse monoclonal antibody and anti-wsp rabbit polyclonal antibody at a dilution of 1:100 for an hour followed by washing with PBS-T . Secondary antibody anti-rabbit-alexa 488 and anti-mouse alexa 594 were then added and incubated for an hour . Coverslips were further washed with PBS-T and stained with DAPI for 10 minutes , followed by a 10-minute wash in PBST and rinsed briefly in distilled water . After removing excess water , slides were fixed with antifading reagent ( ProLong Gold Antifade Mountant , Thermo Fisher ) and were viewed using an immunofluorescence microscope ( Zeiss Imager . A1 ) . Primers that span from 5’ to 3’ of the DENV genome ( Fig 2A , S1 Table ) were used to perform qPCR . 1 μg of DNase-treated total RNA was reverse-transcribed using SuperScript III Reverse Transcriptase ( Invitrogen ) according to the manufacturer’s instructions . qPCR reaction was performed on LightCycler 480 ( Roche ) using the FastStart Universal SYBR Green Master Rox ( Roche ) reagent mix . The data analysis was performed using LinReg software [53] . DENV-1 ET , DENV-2 ET 300 , DENV-3 ET , DENV-4 ET , DENV-1 Viet , DENV- 3 08/09 were all passaged up to 7 times . DENV-2 NGC was produced by in vitro transcription from an infectious clone [54] ( S1 Text ) . All viruses were amplified at the same time in C6/36 cells and harvested 5 dpi and stored in -80°C as aliquots until use . Aag2 and Aag2-wMel cells were infected with various DENV isolates at an MOI of 1 and harvested 5 dpi . The virus was isolated from the media using QIAamp Viral RNA Mini Kit ( qiagen ) and quantified using RT-qPCR using primers DENV-G-F , DENV-G-R and DENV-G- FAM Probe [55] . Virus copy number was calculated from the standard curve generated from serial dilutions of reverse transcribed DENV RNA prepared with carrier RNA . The minimum detection limit was determined as 102 copies [56] . Total cellular RNA was isolated using directzol RNA miniprep kit ( zymo research ) and quantified by RT-qPCR using primers DENV-G-F , DENV-G-R , DENV-G- FAM Probe and normalised with RPS-17 . We screened 124 previously annotated Ae . aegypti miRNAs [57] to identify those with the potential to bind to the 5’ or 3’ UTR of the DENV genome . For each miRNA-UTR pair , we first calculated the minimum free energy ( MFE ) of the best potential binding site between the sequences , using the program RNA-hybrid [58] . We then used the shuffleseq function within the Emboss package to generate 100 simulated sequences of the same length and base composition as the real miRNA and calculated the MFE of each simulated sequence to the UTR . We considered Ae . aegypti miRNAs as candidates for binding to the DENV 5’ or 3’ UTR if their MFE for binding was better than 95% of the simulated sequences for that miRNA-UTR pair . Aag2 and Aag2-wMel cells resuspended in Trizol ( Life technologies ) were used to purify RNA using Direct-zol RNA MiniPrep kit ( Zymoresearch ) according to the manufacturer’s instructions , which included on column DNase digestion . RNA was reverse transcribed using miScript II RT Kit ( Qiagen ) according to manufacturer’s instruction and amplified using LightCycler 480 SYBR Green I Master ( Roche ) with each reaction involving forward primer sequence of mature miRNA ( S1 Table ) and universal reverse primer ( Qiagen ) . Additional adenosine sequence was added to the mature miRNA sequence in case it was required to increase the primer Tm value . The PCR condition was as follows: Step 1 , 95°C for 10 min; Step 2 , 95°C for 10 s , 60°C for 15 s , 72°C for 10 s for 45 cycles with 45 cycles followed by melt curve analysis . The data analysis was performed using LinReg software [53] . miRNAs were further selected according to primer efficiency and expression value and were normalised with RPS-17 . Subgenomic flavivirus RNA ( sfRNA ) of DENV was quantified as described before [22 , 26] with modifications . Briefly , primers QG-F and QGSF-R were used to detect genomic RNA , which gives a PCR product of 309 bp . Primers QGSF-F and QGSF-R were used to detect both genomic and sfRNA with a product size of 184 bp ( See S7 Fig in [26] . qPCR was performed using LightCycler 480 SYBR Green I Master ( Roche ) in a LightCycler 480 real-time PCR machine with 45 cycles and melt curve analysis . Relative levels of genomic RNA were calculated by normalising with RPS-17 . Relative levels of sfRNA were calculated by normalising product produced from QGSF primers ( gRNA + sfRNA ) with product produced from QG and QGSF primers ( gRNA ) and normalised to RPS-17 . The expression levels were shown as fold change in sfRNA level at each time point compared to its level at 0 dpi . The Half-life of cellular mRNA was analysed as reported before [29–31] . Briefly , Aag2 and Aag2-wMel cells were mock infected or infected with DENV at an MOI of 0 . 1 . After 5dpi , the media was changed and replaced with media containing actinomycin D ( 5 μg/ml ) and RNA collected after 30 minutes to ensure cellular transcription shutoff . Total cellular RNA was isolated using directzol RNA miniprep kit ( Zymo research ) and treated with DNase ( Roche ) . Total RNA isolated was reverse transcribed using random primer ( Thermofisher ) and abundance of mRNA was analysed through qPCR using the FastStart Universal SYBR Green Master Rox ( Roche ) reagent mix with primers ECR-F , ECR-R [31] , aae La-F , aae La-R and normalised with RPS-17 . siRNAs ( Sigma ) containing pool against XRN1 ( siXRN1 ) and control siRNA having scrambled sequence ( siControl ) were transfected in Aag2 and Aag2-wMel cells using Lipofectamine RNAiMAX ( Invitrogen ) as reported before [59] . Transfections were carried out twice at 24 hour intervals followed by DENV infection at MOI 0 . 1 . The cells were lysed 48 hpi using trizol ( Life Technologies ) and RNA isolated using Direct-zol RNA isolation kit ( zymo research ) . RNA was DNase ( Roche ) treated and quantified by RT-qPCR . Unless noted otherwise , all data are presented as mean ± SEM from at least three biological replicates . All statistical analysis was performed using t-tests with Welch’s correction unless noted otherwise . Graphs were log10 transformed [60 , 61] before statistical analysis unless noted otherwise . All analysis and graphs were generated using prism-6 software and considered significant if P ≤ 0 . 05 ( * ) , **P ≤ 0 . 01 , *** P ≤ 0 . 001 , ****P ≤ 0 . 0001 .
|
Dengue virus ( DENV ) is a human pathogen transmitted by Aedes mosquitoes . Infection with DENV causes dengue fever and may develop into life-threatening dengue hemorrhagic fever . Dengue disease is increasing globally and current control methods are proving ineffective in curtailing this growing problem . A novel strategy to stop DENV transmission is currently being trialled in five countries through the introduction of Wolbachia , an insect endosymbiont , into wild Aedes aegypti populations . Various mechanisms have been proposed to explain Wolbachia-mediated virus blocking ( WMVB ) including the response of the host to Wolbachia and factors like cholesterol , immune genes and miRNAs . Here we followed the fate of virus in mosquito cell lines and found that Wolbachia does not alter virus binding or internalisation . Further tracking of the virus shows that its replication is reduced in the presence of Wolbachia . The reduced replication is associated with increased viral RNA degradation for both DENV and West Nile virus ( WNV ) . Unlike earlier reports , we didn’t find any evidence for miRNA involvement in WMVB . Analysing the viral RNA further shows that the 3’ region of viral RNA is not fully degraded , indicating that the degradation is likely due to the cellular enzyme XRN1 . Accumulation of DENV 3’ regions inhibited XRN1 in the absence of Wolbachia and reduced the activity of XRN1 but not in the presence Wolbachia . Knockdown of XRN1 using siRNA resulted in decreased WMVB associated with increased DENV RNA . The magnitude of WMVB is also dependent on the infectious virus dose and the intrinsic rate at which the virus strain replicates . Similar results were seen for different DENV serotypes confirming that slowly replicating viruses are blocked more efficiently by Wolbachia .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"pathogens",
"biological",
"cultures",
"microbiology",
"wolbachia",
"viruses",
"micrornas",
"rna",
"viruses",
"cell",
"cultures",
"bacteria",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"viral",
"replication",
"biochemistry",
"rna",
"west",
"nile",
"virus",
"flaviviruses",
"nucleic",
"acids",
"virology",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"organisms"
] |
2018
|
Wolbachia-mediated virus blocking in mosquito cells is dependent on XRN1-mediated viral RNA degradation and influenced by viral replication rate
|
HIV-1 escape from the cytotoxic T-lymphocyte ( CTL ) response leads to a weakening of viral control and is likely to be detrimental to the patient . To date , the impact of escape on viral load and CD4+ T cell count has not been quantified , primarily because of sparse longitudinal data and the difficulty of separating cause and effect in cross-sectional studies . We use two independent methods to quantify the impact of HIV-1 escape from CTLs in chronic infection: mathematical modelling of escape and statistical analysis of a cross-sectional cohort . Mathematical modelling revealed a modest increase in log viral load of 0 . 051 copies ml−1 per escape event . Analysis of the cross-sectional cohort revealed a significant positive association between viral load and the number of “escape events” , after correcting for length of infection and rate of replication . We estimate that a single CTL escape event leads to a viral load increase of 0 . 11 log copies ml−1 ( 95% confidence interval: 0 . 040–0 . 18 ) , consistent with the predictions from the mathematical modelling . Overall , the number of escape events could only account for approximately 6% of the viral load variation in the cohort . Our findings indicate that although the loss of the CTL response for a single epitope results in a highly statistically significant increase in viral load , the biological impact is modest . We suggest that this small increase in viral load is explained by the small growth advantage of the variant relative to the wildtype virus . Escape from CTLs had a measurable , but unexpectedly low , impact on viral load in chronic infection .
The CTL response is thought to play a role in firstly reducing [1] , [2] and then controlling the level of HIV-1 viraemia . Evidence for the protective role of CTLs in humans includes HLA class I heterozygous advantage [3] , the evolution of multiple viral mechanisms to evade CTL surveillance [4]–[6] and the observation of statistically significant associations between possession of certain HLA class I alleles and the rate of disease progression [7] although the size of this CTL-mediated protective effect is contentious [8]–[10] . The pressure exerted by CTLs on HIV-1 leads to frequent [11] , [12] selection of HLA class I-associated viral polymorphisms that result in reduced or abolished CTL recognition , a phenomenon known as HIV-1 “escape” from the CTL response [11] , [13]–[17] . It seems likely that HIV escape would be detrimental for the host , yet the impact of escape on viral load or CD4+ T-cell count is unclear . Longitudinal studies following viral escape in HIV-1 infected individuals have shown no conclusive results [18]–[21] , largely because of data scarcity . Cross-sectional studies show a relationship between HLA-associated polymorphisms in pol and high viral load [11] , and between HLA-associated polymorphisms in gag and low CD4+ count [22] and high viral load [23] . However , as acknowledged by the authors themselves [23] , it is impossible to interpret these results because the direction of causality is unknown . That is to say , it is unclear whether the negative correlation between the number of viral polymorphisms and CD4+ T-cell count arises because a high number of polymorphisms leads to poor CTL control and greater immunosuppression , or simply because these individuals have been infected for a longer time , resulting in a lower CD4+ T-cell count and , simultaneously , a greater opportunity to develop mutations . Similarly , the association between high viral load and HLA-associated polymorphisms may be due to viral strains escaping the immune response which results in an increase in viral load , or it may be that high viral load and high viral replication allows for more mutations to accumulate . Furthermore , recent work demonstrated a correlation between rate of progression to AIDS and the replication rate of the infecting strain [24] , [25] . This introduces another potential confounding factor into associations between HLA-associated polymorphisms and surrogate markers of disease progression . Therefore , in order to interpret the relationships between HLA-associated polymorphisms and viral load and/or CD4+ count , it is essential to adjust for the number of mutation events that have occurred in individuals . Hence , because of limited longitudinal data and difficulties in interpreting cross-sectional data , the consequences of HIV-1 escape from the CTL response are still unclear . The aim of this work was to quantify the impact of HIV escape on viral load in chronic infection .
We used an extension of the Perelson/De Boer model of HIV-1 dynamics [26] to simulate the outgrowth of an HIV-1 variant which has escaped a single CTL response . This enabled us , firstly , to check our intuition that escape must lead to an increase in viral load; secondly , to obtain an order of magnitude estimate of the predicted increase in viral load upon escape; and thirdly , to test whether a small increase in viral load upon escape could be explained solely by the outgrowth rate of the variant , or whether it was necessary to postulate additional mechanisms ( e . g . the generation of de novo CTL responses which recognise the variant and subsequently reduce viral load ) . The variant was assumed to have reduced replicative ability compared to the wildtype ( i . e . although the variant is fitter in the presence of a CTL response , it is less fit in the absence ) . The system was modelled under the “worst case” scenario that the CTL response is unable to target the escaped epitope again and so does not regain control of viraemia . In all 100 , 000 runs of the model , covering a wide range of biologically plausible parameters , there was an increase in viral load upon escape . The increase in viral load upon escape was significantly correlated with the outgrowth rate of the variant ( Spearman's rank correlation: p<0 . 0001; rho = 0 . 59 ) . For outgrowth rates representative of what is observed in chronic HIV infection ( i . e . the variant replaces the wildtype at rate 0 . 01–0 . 04 per day , corresponding to variant outgrowth in 230–900 days [4] , [8] , [13] , [15] , [16] , [20] , [27]–[31] ) , the median increase in viral load was 0 . 051 log copies ml−1 ( 95% confidence interval: 0 . 050–0 . 052; Figure 1; Figure S1 ) . A second scenario , in which the infection and virion production rates of the variant were set independently of the wildtype such that the variant was permitted to be fitter than the wildtype at the time of escape , was also considered . In this model , the median increase in viral load was 0 . 050 log copies ml−1 ( 95% confidence interval: 0 . 044–0 . 056 ) . Even if the outgrowth rate of the escape variant was an order of magnitude higher ( e . g . 0 . 4 per day , corresponding to the variant replacing the wildtype in 23 days ) , the median predicted increase in viral load was still only 0 . 27 log . As before , there was a significant positive correlation between the outgrowth rate of the escape variant and the resulting increase in viral load ( Spearman's rank correlation: p<0 . 0001; rho = 0 . 59 ) . Interestingly , for this model there was a fraction ( 12% ) of cases in which escape resulted in a decrease in viral load . To test the robustness of the prediction that escape leads to a small increase in viral load , additional mathematical models were simulated which tested different assumptions and explicitly modelled the CTL response as a separate dynamic population ( see Table S1 for further details ) . Median changes in viral load for these models ranged from an increase of 0 . 017 to 0 . 037 log copies ml−1 ( see Text S1 for full details ) . Following Moore et al . [11] and Kiepiela et al . [32] , we define an “escape event” as a coding change in an epitope that is statistically significantly associated with possession of the restricting HLA allele . We introduce two differences into our analysis . Firstly , we cap the number of escape events per epitope at 1 ( so that we do not double-count additional mutations required for escape from the same CTL clone ) . Secondly and more importantly , we adjust for the background level of mutation . Variation in the length of infection and viral replicative rates between individuals are confounding effects which can increase both the number of mutations and viral load making it impossible to infer the impact of HIV escape on viral load simply by studying raw cross-sectional data [22] . In order to correct for the background level of mutation , we performed multivariate regression . This allowed us to test whether the number of escape events ( NEE ) is a significant independent predictor of viral load after a possible relationship between viral load and the background level of mutation ( quantified by NSE , the number of synonymous changes in epitopes per individual ) had been taken into account . The analysis showed that the number of escape events is a significant independent predictor of log viral load when the number of synonymous changes is taken into account ( Figure 2 , multiple linear regression: two-tailed p = 0 . 0021 , gradient = 0 . 11 , r2 = 0 . 060; Figure S2 shows alternative representations; linear regression of NEE alone on log viral load: p = 0 . 029 , gradient = 0 . 055 ) . Neither changing the definition of “escape event” to include only mutations with a decrease in predicted binding affinity ( multiple linear regression: p-value = 0 . 42 , gradient = 0 . 13; see Text S1 ) , nor stratifying the cohort based on frequently occurring alleles ( see Table S2 ) , altered these results . Repeating this analysis with CD4+ T-cell count and the number of escape events showed a trend toward a negative association ( multiple linear regression: two-tailed p = 0 . 062 , gradient = −23 ) . In order to investigate the relationship between escape from protective alleles and viral load , we performed three analyses . First , we removed individuals with protective alleles from the multiple linear regression ( see Table S2 ) . The “unprotected” group still has a statistically significant positive association between log viral load and NEE ( multiple linear regression p = 0 . 00024 , coef = 0 . 19 ) , although no association was observed in the “protected” group ( p = 0 . 20 , coef = 0 . 060 ) . Second , we added the number of protective alleles and the number of not-protective alleles possessed by an individual as ( two ) additional predictors of log viral load to the multiple linear regression . We find that all four factors are significant independent predictors of log viral load ( multiple linear regression: NEE p = 0 . 000444 , coef = 0 . 12038; NSE p = 0 . 009526 , coef = −0 . 09110; Protective p = 0 . 006509 , coef = −0 . 47809; Not-protective p = 0 . 006715 , coef = 0 . 28016 ) . Third , we wanted to further explore whether alleles associated with a lower viral load were protective because they presented epitopes where escape was rare . To do this we calculated the NEE of epitopes presented by protective alleles by individual and compared it to the NEE of epitopes presented by not-protective alleles . To ensure equal sample size in the “protective” and “not-protective” groups , we selected a number of alleles in either group such that 10% of the cohort possessed a protective allele , and 10% of the cohort possessed a not-protective allele . We observe that epitopes restricted by protective HLA alleles had a median of 1 . 0 additional escape events than not-protective alleles ( Wilcoxon rank sum test: p = 0 . 013 , difference = 1 . 0 ) . These three pieces of analysis show that escape in epitopes presented by not-protective HLA class I alleles is detrimental for the host , and that protective HLA class I alleles do not have fewer escape events than ‘detrimental’ alleles . The “protectiveness” of an allele is therefore not attributable to an inability to escape such responses . To check for robustness , we performed a bootstrap analysis by sampling 10 , 000 times with replacement from the Full cohort . We found that 92% of the bootstrap runs resulted in a significant positive association between log viral load and the number of escape events , the median two-tailed p-value of all runs was 0 . 0019 ( 95% confidence interval: 0 . 0018–0 . 0021; Figure S3A ) . By calculating the gradient of the line of best fit through the data , after correcting for the background level of mutation , we find that each escape event causes a median increase in viral load of 0 . 11 log copies ml−1 ( 95% confidence interval: 0 . 040–0 . 18 ) , or an absolute median increase of 8 , 500 copies ml−1 . Overall , the number of escape events could only explain about 6% of the variation in viral load in the cohort . Next we investigated whether escape events in any particular gene were associated with the increase in viral load , so we repeated the multi-linear analysis on each gene independently ( Figure 3; Table S3 ) . We found a significant positive association between log viral load and coding changes in pol ( multiple linear regression: two-tailed p-value = 0 . 0059 , gradient = 0 . 14 ) , but not in any other gene . Additionally , a significant positive association was found between viral load and escape events in pol in 83% of bootstrap runs ( Figure S4 and Table S4 ) . Amongst the other genes , we noted that there was a decrease in log viral load with escape events in gag , although this was not statistically significant ( multiple linear regression: two-tailed p-value = 0 . 91 , gradient = −0 . 011; Table S3 ) . As Gag-specific CTL responses have been shown as protective [33]–[37] , we had expected escape to lead to a large rise in viral load . To check that power was not an issue , we repeated the analysis using a longer HLA-gag epitope list derived using a phylogenetically corrected method [38] . This confirmed the negative association between escape events in gag and log viral load ( multiple linear regression: two-tailed p-value = 0 . 013 , gradient = −0 . 091 ) . Next , we investigated which gene ( or genes ) was driving the effect in the 92% of bootstrap runs that gave a significant association between the total number of escape events in all genes and viral load ( see previous section “Relationship between number of escape events and viral load” ) . This analysis showed that escape events in most genes were detrimental for the host , but that escape events in Pol epitopes were most frequently ( 81% ) behind the increase in viral load ( Table S5 ) . Finally , we investigated individual gene effects in the Extended cohort ( N = 347 ) in order to increase power . We could not study this cohort in the previous analysis as we did not have sequence data for all genes for all individuals; however this is unnecessary for the purposes of individual gene analysis . As the Full cohort is a subset of the Extended cohort , the results must not be considered independent . Analysis of this extended cohort strengthened and extended our previous results . Escape events in pol were still significantly associated with an increase in viral load ( at a more significant two-tailed p-value of 0 . 0032 ) . The results of a bootstrap run on this extended cohort are shown in Figure 4 . Genes appear to fall in one of two distinct groups: genes in which escape events were associated with a high viral load ( env , nef , pol and vif ) and genes in which escape events were associated with a low viral load ( gag and rev ) .
To date , it has not been possible to quantify the impact of HIV-1 escape on viral load . Longitudinal datasets are too sparse to allow firm conclusions and studies of cross-sectional datasets have been impossible due to confounding factors . Furthermore , previous studies [39]–[41] have concentrated on escape from CTLs associated with protection , e . g . Gag-specific CTLs , or on protective alleles , e . g . B*57 , which although interesting and important for determining mechanisms of protection , are unlikely to be representative and may give a distorted view of the importance of escape . Here , we have investigated the impact of “typical” ( i . e . average ) escape events in seven genes in chronic HIV-1 infection using two independent methods: mathematical modelling and cross-sectional cohort analysis , using multiple regression to remove confounding factors . Firstly , we simulated the emergence of HIV-1 escape variants using an extension of the Perelson/De Boer model of HIV-1 dynamics [26] , using parameters selected from a wide range of biologically realistic values . We modelled the system under the “worst case scenario” that the CTL response does not regain control of viraemia following escape . This modelling showed that the increase in viral load per escape event was significantly positively correlated with the outgrowth rate of the variant ( Spearman's rank correlation: p<0 . 0001; rho = 0 . 59 ) . The median increase in viral load was 0 . 051 log copies ml−1 per escape event . Adapting the model to allow CTLs to partially regain control would lead to a smaller long-term increase in load per escape event . A range of alternative models based on different assumptions led to similar conclusions . Secondly , we analysed a cross-sectional cohort of 157 HIV-1 C-clade infected individuals . We used a novel method to remove confounding factors which permits the analysis of data-rich cross-sectional cohorts . By treating the number of synonymous changes accrued by an individual as a clock that keeps count of the number of “mutation events” , we corrected for differences in individual mutation rates , regardless of whether the increased mutation frequency resulted from a higher viral set point , higher viral replication rate or longer length of infection ( and hence lower CD4+ count ) . We show a statistically significant positive association between the number of mutated epitopes and viral load . We found a small increase in viral load of 0 . 11 log per escape event . To put this into context , the median standard deviation of random fluctuations in log viral load over time in a chronically HIV B-clade infected cohort is 0 . 32 ( see Text S1 ) . The association between the number of mutated epitopes and viral load was mainly accounted for by mutations in pol . The escape events studied could have occurred at any time point prior to sampling , i . e . evolved during the acute [42] or chronic phases , or were already present in the transmitted infecting virus . The small increase in viral load per escape event that we observed in the cohort could be explained by two ( non-exclusive ) hypotheses . Firstly , the impact of escape on viral load is transient due to the flexibility of the CTL response which adapts to recognise either the escape variant or a previously sub-dominant epitope , resulting in renewed and long-term suppression of viral load . Secondly , the impact of escape is small because the variant has a slow outgrowth rate ( i . e . a small net fitness advantage compared with the wildtype ) , so the accompanying increase in viral load is not large . Under the first hypothesis , escape initially causes a large increase in viral load but then the CTL response adapts and partially regains control of viraemia , either by targeting previously unrecognised or sub-dominant epitopes [16] , [43] , [44] or by developing a new CTL response against the variant in the case of escape due to loss of TCR recognition [45] . The transient nature of the increase would unlikely be reflected in a cross-sectional study , and so the observed increase in viral load would be small . Under the second hypothesis , the increase in viral load per escape event is small simply because the outgrowth rate of the escape variant is slow ( i . e . CTL selection pressure on the wildtype minus fitness cost of the variant is small ) . The mathematical modelling we performed allowed us to test these two hypotheses . For outgrowth rates observed in chronic HIV-1 infection [4] , [13] , [15] , [16] , [20] , [27]–[31] , equivalent to a variant replacing the wildtype in 230 to 900 days from first appearance of the variant , we found that the median increase in viral load was predicted to be around 0 . 051 log copies ml−1 . Whilst we do not rule out that existing CTL responses are strengthened or that new CTL responses can arise to regain control of viraemia after an escape event , this mechanism is unnecessary to explain the observed modest increase in viral load per escape event . Rather , the small increase in viral load is exactly what would be predicted given the observed outgrowth rate of escape variants in late primary and chronic infection . Additionally , the modelling showed that the increase in viral load per escape event was significantly positively correlated with the outgrowth rate of the variant: that is , the faster the variant outgrows the wildtype , the larger the resulting increase in viral load . This is consistent with the observation that protective HLA class I alleles associated with slow progression to AIDS tend to present epitopes where escape variants have a slower outgrowth rate [46] . These two observations predict , somewhat counter-intuitively , that escape from protective HLA alleles will result in a smaller increase in viral load than escape from non-protective alleles . A similar argument could explain our finding that escape events in gag did not cause a measurable increase in viral load . Given the protective role of Gag-specific CTLs [33]–[37] , it might be expected that escape from these CTLs would be particularly detrimental . However , some escape variants in gag have also been shown to carry very heavy fitness costs [47]–[49] . Consequently , the outgrowth rate of escape variants in gag will not necessarily be fast . In fact , recent preliminary data suggests the opposite: that escape variants in gag have a significantly slower outgrowth rate than variants in other proteins [46] . This would explain why escape mutations in gag have not been found to lead to large increases in viral load . Results from the modelling suggest one potential mechanism how escape can lead to a lower viral load . Intuitively , we would expect that variants must have a net fitness advantage over the wildtype , which results in a higher viral load upon escape . However , in Model 2 , we observed that 12% of all runs resulted in a decrease in viral load and the outgrowth of the variant . Our analysis of these runs showed that in all cases , the variant was more infectious than the wildtype ( higher β′ ) but able to produce fewer virions ( lower h′ ) , resulting in an increase in the number of variant-infected cells but not an increase in the viral load . This intriguing result suggests that although the variant has a net fitness advantage over the wildtype , it may only be manifest at the level of infected cells and not necessarily at the level of free virus . It is therefore not unfeasible for an ‘attenuated variant’ to emerge . A prediction of the model is that proviral load would always be higher after escape . Alternatively , the negative association between escape events in gag and viral load may simply be due to individuals with protective HLA class I alleles , which are associated with lower viral load , mounting more Gag-specific CTL responses and therefore the emergence of more escape variants in gag . A number of studies have reported a relationship between protective HLA class I alleles and low levels of viral escape , suggesting that escape is an important determinant of progression to AIDS [46] , [50]–[52] . However , we did not find any evidence that protective HLA class I alleles were associated with lower levels of viral escape . Instead , we observed that individuals possessing a protective allele had more escape events than individuals with an allele associated with a higher viral load . The protectiveness of an allele therefore appears unrelated to its ability to prevent viral escape ( despite the high fitness cost of escape mutations ) , but rather suggests that protective alleles present epitopes where variants only have a small net fitness advantage . We find that escape only determines approximately 6% of variation in viral load . How can we reconcile the low impact of escape events on viral load with disease progression , given the strong association between HLA and disease progression ? As viral load is subject to high levels of random fluctuations within an individual over time , one possible explanation is that viral load is not always strongly associated with progression [53] . However , we also failed to find a strong association between escape events and CD4+ T cell count , which is usually a better correlate of progression . Another possible explanation is that HLA-mediated protection may not be completely dependent on CTL , as other HLA-mediated factors could play a greater role in HIV infection , e . g . NK cells or linkage disequilibrium with a protective gene . It is therefore entirely plausible that CTL escape only partially contributes to HLA-mediated protection . Furthermore , a much greater number of CTL-independent factors potentially determine viral load , for example , CCR2 polymorphisms and CCR5 deletions [54]–[56] , Nef deletions [57] , NK cell receptor polymorphism [58] , viral tropism [59] and host activation status [60] . A recent study also showed significant genetic and immunological heterogeneity between people who naturally control HIV-1 viremia [33] . Other studies have estimated that HLA class I type determines about 3%–15% of viral load [33] , [61]–[63] and that escape mutations determine about 30% of HLA class I protection [46] . If we assume in the most extreme case that that HLA-associated protection is solely due to the ability to target epitopes where escape is rare , then this would indicate that only around {3%–15%}×30% = 0 . 9% to 4 . 5% of variation in viral load could be explained by escape , which is very similar to our estimate of 6% . Given the large number of possible determinants of viral load it is perhaps not surprising that HIV escape from CTL can only explain a small proportion of an individual's viral load . We predicted mathematically and also showed by analysis of an HIV-1 infected cohort that HIV escape from the CTL response is associated with a small increase in viral load . Although the finding that a typical escape event has a modest impact on viral load is surprising , it is consistent with firstly , the failure to find a clear association between escape and viral load in longitudinal studies [18]–[21] , secondly , the low outgrowth rate of escape variants in chronic infection [8] , and lastly , the weak positive correlation between uncorrected viral load and HLA-associated polymorphisms in gag observed by Brumme et al . [23] . This study shows that although the impact of HIV-1 escape from CTLs is highly statistically significant , the effect in clinical terms is mild: an increase in viral load of 0 . 11 log copies ml−1 will have few consequences for patient health [64] . This unexpected result demands further independent studies to corroborate it . If replicated , this result would suggest that , with the crucial exception of vaccine design , the focus on HIV escape may be out of proportion to its importance with other factors playing a more significant role in determining viral load in chronic HIV-1 infection .
The Perelson/De Boer model of HIV-1 dynamics [26] was adapted to include HIV-1 escape variants , which have escaped a single CTL response ( Equation 1 ) . ( 1 ) The five populations ( all measured per mm3 ) are as follows: uninfected CD4+ cells ( S ) ; CD4+ cells infected with wildtype virus ( X ) ; free wildtype virus ( W ) ; CD4+ cells infected with variant virus ( Y ) ; free variant virus ( V ) . Parameters are described in Table 1 . The model was simulated using lsoda in R . Each run sampled parameters from ranges obtained from the literature [65]–[71] , and the variant was assumed to be attenuated compared to the wildtype , such that β>β′ and h>h′ . The median outgrowth rate of the variant ( k , see below ) observed in chronic HIV-1 infection [4] , [13] , [15] , [16] , [20] , [27]–[31] is 0 . 01 d−1 to 0 . 04 d−1 [8]; we were interested in typical escape events and so runs with an outgrowth rate outside this range were not included in the analysis and the model was run until 100 , 000 simulations had been accumulated . From these runs the difference in viral load of the wildtype ( W+V at t0 ) and the variant ( W+V at tend ) was calculated . A full description of this model , as well as other models considered , can be found in Text S1 . The outgrowth rate ( net growth advantage ) of an escape variant is the rate at which the variant outgrows the wildtype virus . It is defined as the growth rate of the variant minus the growth rate of the wildtype [8] , [72] , or: ( variant replication rate – variant death rate ) – ( wildtype replication rate – wildtype death rate ) . As this quantity is not easily expressed in terms of the model parameters , we use the results of the model simulation and Equation 2 to calculate the outgrowth rate of the variant . ( 2 ) Vavg is the average value of variant free virus ( V ) from the time when the variant first emerges ( t0 ) to the time when the variant constitutes 99% of the virus population ( tfix ) . Similar to Asquith et al . [8] , we do not consider the time for the variant to emerge when calculating the outgrowth rate . Xavg ( average number of wildtype-infected cells ) , Wavg ( wildtype free virus ) , Yavg ( variant-infected cells ) and Savg ( uninfected cells ) are similarly defined as in Equation 3 . ( 3 ) Alternatively , k can be estimated by fitting the solution of a simplified model of escape virus dynamics [8] to the model data using nonlinear least-squares regression ( using the Levenberg-Marquardt algorithm , nls . lm , in the minpack . lm package in R ) . The two approaches to calculating k are in strong positive agreement ( Pearson's product moment correlation coefficient: p<0 . 0001 , r = 0 . 998 ) . The time for the variant to reach fixation is given by approximately 9/k ( see Text S1 for details ) . This list , reproduced in Table S6 , is taken from Supplementary Table S3 of Kiepiela et al . 2007 [73] , and consists of 37 HLA-epitopes pairs which are statistically significantly associated with coding changes in the epitope region ( plus three amino acids either side of the epitope region ) in individuals with the same HLA class I allele . Viral RNA sequence , HLA class I genotype and viral load were taken from a treatment naïve HIV-1 C-clade infected cohort attending a prenatal clinic in Durban , as described in [32] . Individuals with a viral load greater than 106 copies ml−1 were classed as acutely infected [74] , [75] and removed from the analysis . This cohort consisted of two overlapping groups: the Full cohort which consisted of individuals with virus sequence data for all epitopes in the K37 epitope list ( N = 157 ) ; and the Extended cohort which included all individuals with complete viral sequence data for a single gene ( N = 347 ) . The Extended cohort includes all individuals in the Full cohort . All cohort data was collected and anonymised by researchers external to the authors of this paper . Following Moore et al . [11] and Kiepiela et al . [32] , a viral coding change is classed as a ‘HLA-associated polymorphism’ if the variant amino acid is significantly associated with possession of the restricting HLA allele . Polymorphisms will include mutations which disrupt immune surveillance ( by altering processing , MHC binding affinity or T-cell recognition ) , and compensatory mutations which restore fitness to the virus . For each gene in the HIV-1 genome , all viral amino acid sequences in the cohort were aligned using TCoffee [76] , and the most frequent amino acid assigned as the cohort consensus . To obtain the nucleotide consensus , the amino acid sequence was used as an alignment guide for the program tranalign ( available from http://emboss . sourceforge . net ) and the most frequently occurring triplet of nucleotides coding for the most frequent amino acid was chosen as the cohort nucleotide consensus . The number of epitopes with a coding change ( NEE ) was calculated for each patient , either across all genes , or by separating the epitopes by gene and calculating a NEE score for each gene . HLA-matched epitopes , plus flanking regions of 3 amino acids either side of the epitope , were taken from the K37 HLA-epitope list and compared to the epitope in the cohort consensus amino acid sequence . Any deviation from the cohort consensus was marked as a putative escape mutation and the number of putative escape events ( NEE ) calculated . The escape event could have happened at any time point prior to sampling ( i . e . during the acute or chronic phases , or was already present in the transmitted infecting virus ) . The number of escape events per epitope is capped at one , i . e . if there are two HLA-associated polymorphisms in the same epitope then this is counted as a single , and not multiple , escape event . We use NSE to quantify the background level of mutations in an individual . NSE is the equivalent of NEE for synonymous nucleotide changes . The method for obtaining a score for NSE , i . e . epitopes with a silent coding change , is similar to the method for obtaining NEE , save that the individual's viral nucleotide sequence is compared with the cohort consensus nucleotide sequence and only non-coding changes enumerated . To calculate the impact of HLA polymorphisms on viral load after correcting for the “background” number of mutations , we performed multiple linear regression with log viral load as the dependent variable and NEE and NSE as predictors: ( 4 ) The data plotted in Figures 2 and 3 are partial residual plots , that is , they show viral load after removing the background level of mutation for each individual . The left-hand side of Equation 4 was plotted on the y-axis against NEE on the x-axis , where is the regression coefficient for NSE from the fitted full model . The fraction of variation in viral load in the cohort that could be explained by this model was calculated using the R2 coefficient . Patients were selected from the Full cohort at random with replacement 10 , 000 times . For each bootstrap iteration in which total escape events over all genes was a significant predictor of log viral load , escape events were then split by gene to determine escape events in which gene were significant predictors of log viral load . When multiple genes were found to be significant predictors , we checked if they were independent predictors or correlated with escape events in other genes . Sets of independently significant genes contributing less than 1% of the total runs are grouped together as “Minor effects” and excluded from further analysis . Multiple linear regression was used to determine significance , and the number of synonymous changes was always included as a co-predictor . See Text S1 for full details of the method . To estimate the fraction of viral load variable that can be explained by the HLA class I genotype , we calculated the “Explained Fraction” ( EF ) metric [61] on data reported by Nelson et al . [61] , Pereyra et al . [33] and Emu et al . [62] separately . We categorised the various HLA class I genotypes into two classes: protective alleles and not protective ( or susceptible ) alleles . Disease states were categorised either using AIDS endpoints ( for the Nelson data ) or viral load ( for Pereyra and Emu data ) . Methods and results are detailed in full in Text S1 .
|
HIV , like many viruses , has evolved multiple strategies to evade immune surveillance . One of these strategies is the evolution of escape mutations which reduce the ability of the immune response to kill HIV-infected cells . But does HIV escape matter ? Some believe that the accumulation of escape mutations leads to AIDS; many more believe escape is likely to be highly detrimental to human health . Yet , to date , it has not been possible to measure the impact of escape . We developed two independent methods to quantify the impact of escape on HIV viral load . Both methods showed that escape does lead to a detectable increase in viral load , but is unlikely to have a major impact on patient health as the increase is small . Indeed , only 6% of between-individual variation in viral load could be attributed to HIV escape . This work suggests that the current research focus on escape in chronic HIV infection might be out of proportion to its importance with other factors playing a more significant role in determining viral load .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology",
"computational",
"biology"
] |
2010
|
Quantifying the Impact of Human Immunodeficiency Virus-1 Escape From Cytotoxic T-Lymphocytes
|
The sensitivity of massively-parallel sequencing has confirmed that most cancers are oligoclonal , with subpopulations of neoplastic cells harboring distinct mutations . A fine resolution view of this clonal architecture provides insight into tumor heterogeneity , evolution , and treatment response , all of which may have clinical implications . Single tumor analysis already contributes to understanding these phenomena . However , cryptic subclones are frequently revealed by additional patient samples ( e . g . , collected at relapse or following treatment ) , indicating that accurately characterizing a tumor requires analyzing multiple samples from the same patient . To address this need , we present SciClone , a computational method that identifies the number and genetic composition of subclones by analyzing the variant allele frequencies of somatic mutations . We use it to detect subclones in acute myeloid leukemia and breast cancer samples that , though present at disease onset , are not evident from a single primary tumor sample . By doing so , we can track tumor evolution and identify the spatial origins of cells resisting therapy .
Cancer is a disease largely driven by accumulated somatic mutations . Many of these are clonal mutations and occur in the founding cell to initiate disease . These become uniformly present in the tumor by propagating to that cell's progeny during clonal expansion . Others are subclonal events , which occur in an existing neoplastic cell and are then passed on only to the subpopulation of cells derived from it . The result of this accumulation of mutations is that tumors are composed of a heterogeneous mixture of cells . These subpopulations compete and evolve [1]–[3] , and the mutations “captured” [4] in subsets of cells during this evolution serve as a genetic signature of the resulting ( sub ) clones . Recently , high resolution glimpses of this clonal heterogeneity have been provided by next-generation sequencing [4]–[14] , SNP array [6] , [10] , [13] , [15] , and array comparative genomic hybridization [16] , [17] platforms . Single-cell sequencing [16] , [18]–[20] may eventually address this heterogeneity directly without the confounding effects of mixing cell types , but technical challenges , such as allele dropout [21] , remain . There are also pragmatic concerns about the large number of cells that must be sequenced to establish the heterogeneity of a given sample . The emerging picture from these studies , across a diversity of solid [6] , [8] , [9] , [11] , [13] , [16] and hematological [4] , [5] , [7] , [10] , [12] , [14] , [15] , [17] , [22] disorders , is that tumors are both spatially [9] , [13] , [16] and temporally [4]–[17] heterogeneous and are frequently comprised of a single founding clone and several subclones . Increasing evidence suggests that intra-tumor heterogeneity and clonal architecture have clinical implications [3] , [23] , [24] and contribute to therapy resistance [25] . Several studies have linked the presence of subclones to poor clinical outcome , as in chronic lymphocytic leukemia ( CLL ) [10] , or to increased risk of progression to malignancy , as in Barrett's esophagus [26] and multiple myeloma ( MM ) [17] . Subclonal mutations can drive resistance as well , as shown in EGFR-mutant non-small cell lung cancers [27] , [28] . Studies in chronic myeloid leukemia have also demonstrated that drug-resistant subclones may harbor aggressive mutations that are restrained by more fit , but indolent clones [2] , [3] . In these cases , therapeutic application of imatinib leads to competitive release of BCR-ABL mutant subclones , which renders the therapy ineffective [3] , [29] . Thus , designing effective second line therapies requires a deep understanding of both a cancer's underlying mutations and how it's clonal structure evolves in response to treatment . Existing methods [6] , [10] , [11] , [22] , [30]–[32] have been useful for inferring clonal architecture and its consequences , e . g . , that putative driver mutations in SF3B1 and TP53 in CLL [10] and in PIK3CA and PTEN in triple-negative breast cancer [11] may arise during late-emerging , subclonal diversification [6] rather than as founding lesions . Recent results suggest that accurately describing the subclonal composition and evolution of tumors requires sampling cancer cells across multiple time points or spatially-distinct regions [3] , [23] , [25] . Current studies of distant metastases and of spatial heterogeneity are collecting as many as six to twenty samples [9] , [13] , [16] , [18] . The scale of these ambitious studies will challenge existing methods . For example , histogram-based approaches to representing clonal markers [10] , [31] are attractive in avoiding model assumptions in low dimensions ( i . e . , with few samples ) , but with many samples will suffer from exponential computational complexity . Several approaches [6] , [10] , [11] , [22] , [30] , [31] leverage Markov chain Monte Carlo ( MCMC ) techniques , but these , too , are computationally demanding and rely on assumptions about chain convergence . Most existing methods [6] , [10] , [22] , [31] inferring clonality from copy number alterations ( CNAs ) avoid additional computational overhead by making the simplifying assumption that the tumor sample is “monogenomic” [31] and does not harbor subclonal copy-number events . Contrary to their assumptions , these methods have detected such subclonal events in CLL [10] , though without being able to correct for them . Similar subclonal events have been detected in MM ( Ref . [33] and B . S . W . , R . V . , and M . H . T , data not shown ) . These methods introduce uncertainty , through the probabilistic inference of allele-specific copy numbers , and error , by ignoring subclonal CNAs . A recent approach [32] does generalize to subclonal CNAs , but also suffers from computational inefficiencies when extended beyond the simple case of clonal CNAs . A method which could avoid the uncertainty of deconvolving subclonal CNAs and operated with significantly lower computational demands would benefit studies aiming to understand the evolution of cancer . To address these needs , we introduce SciClone , a method for estimating the number and content of subclones across one or many samples . It focuses primarily on variants in copy-number neutral , loss of heterozygosity ( LOH ) -free portions of the genome , which allows for the highest-confidence quantification of variant allele frequencies ( VAF ) and inference of clonality . As with other tools [6] , [11] , [22] , [30] , regions of CNA and LOH are provided as inputs after having been inferred from whole-exome sequencing ( WES , e . g . , via ASCAT [22] , [34] , cn . MOPS [35]–[37] , or VarScan 2 [38] ) , whole-genome sequencing ( WGS , e . g . , via HMMcopy/APOLLOH [11] , [39] or VarScan 2 [38] ) , or SNP arrays ( e . g . , via ASCAT [6] , [22] , [34] ) . SNVs of sufficient depth are provided by WES or first discovered by WGS and subsequently deeply sequenced in a targeted fashion . The approach is not limited to SNVs , but is amenable to any event that can be described as a frequency . In particular , we demonstrate the integration of copy number events and discuss how copy-altered VAFs could be accommodated . Computational efficiency is achieved by clustering the VAFs using a variational Bayesian mixture model [40] ( VBMM ) , which differs substantially from the Dirichlet process models previously used to infer subclones [6] , [10] , [11] , [22] , [30] , [31] . VBMMs similarly automatically infer the number of clusters and provide a probabilistic interpretation of clustering , but their deterministic nature allows them to scale to high dimension , where they enjoy efficiency advantages [40] over stochastic MCMC techniques employed by existing clonality detection methods [6] , [10] , [11] , [22] , [30] , [31] . Further , the variational Bayesian approach provides a computational termination condition more straightforward than monitoring techniques [41] required of MCMC . Though VBMMs are heuristic , and their approximations occasionally result in sub-optimal solutions [42] , we demonstrate their effectiveness here through simulation and application to several real tumor data sets . In particular , SciClone advances our preliminary [14] variational Bayesian beta mixture modeling approach for clustering VAFs in a single sample by: ( 1 ) applying the standard technique of factorizing the density over samples [43] , [44] to extend applicability to an arbitrary number of samples , ( 2 ) replacing our previous ad hoc notion of cluster overlap with a quantitative measure [45] , [46] , ( 3 ) leveraging the probabilistic nature of VBMMs to quantify a variant's likelihood of belonging to a cluster via a -value , and ( 4 ) offering alternative binomial and Gaussian mixture models . We demonstrate SciClone by inferring low-frequency subclones from a single MM sample and by quantitatively assessing the clonality of driver mutations in ( potentially noisy ) WES-derived data . We extend this approach to accommodate multiple samples and apply it to track clonal evolution of an acute myeloid leukemia ( AML ) tumor to relapse in response to therapy . As a further example of SciClone's scalability and utility in correlating mutations across samples , we examine spatial heterogeneity and aromatase-inhibitor resistance within three samples from a single breast cancer patient . In both the AML and breast cancer data sets , our analysis reveals subclones present in the primary tumor but not discernible from a single primary tumor sample . This reinforces the necessity of analyzing multiple patient-derived tumor samples to elucidate the full complexity of a cancer's heterogeneity . The SciClone package is available at http://github . com/genome/sciclone .
Many tumors are highly heterogeneous and visualizations of somatic VAFs reveal high-density aggregations that correspond to specific subpopulations of cells ( Fig . 1 ) . To test our ability to detect and segregate these clusters , we used 2 , 018 validated , deep-sequenced ( median depth 188x ) , genome-wide somatic SNVs from a primary MM tumor ( M . H . T . , et al . , in preparation ) . These formed a high density region near 50% VAF , as expected of heterozygous SNVs in the founding clone of a nearly pure tumor sample ( Figs . 1a and c ) . The actual average VAF of this founding clone is slightly less ( 46 . 1% ) , reflecting a small amount of normal cell admixture and a tumor purity of 0 . 922 . Lower VAFs correspond to subclone-specific mutations that arose later in the tumor's expansion; a cluster of such VAFs thus represents a subclonal population , whose cells contain all of the founding clone mutations , as well as these subclonal mutations . This tumor is hyperdiploid with characteristic trisomies of chromosomes 3 , 5 , 7 , 9 , 11 , 15 , 19 , and 21 . Founding clone mutations in these ( and other ) copy number altered regions are shifted in a predictable pattern , with a doubling of VAFs in regions of single-copy deletion where only the variant allele remains ( Fig . 1b ) . In regions of single-copy amplification , mutations group near the expected frequency of 31% where the wild-type allele is amplified ( i . e . , ∼1 variant allele/3 total alleles ) or 62% when the mutant allele is amplified ( i . e . , 2 variant alleles/3 total alleles; Fig . 1d ) . A more careful calculation that incorporates the tumor purity gives an expected frequency of the mutant-amplified cluster , , very close to the observed frequency of 63% . The same holds for the wild-type-amplified cluster . In this patient , these wild-type amplified SNVs occur at similar VAFs to subclonal events that are copy number neutral ( red circles in Fig 1c ) . Disambiguating the two would require inference of allele- ( and subclone- ) specific copy number profiles , with its attendant uncertainty . To confirm that the two cases are in fact distinct and not an artifact of inaccurate copy number calls , we verified that none of the subclonal event VAFs in putative copy-number neutral regions reside instead on trisomic chromosomes . Further , they are not restricted to one or a few chromosomes , whose amplification or deletion might not have been detected , thus leading to apparent subclonal events . To obtain a more objective view of this sample's clonal architecture , we identified low-frequency subclones by clustering the VAFs from copy number neutral , LOH-free , non-repetitive regions of the genome using SciClone , which uses an approach based on variational Bayesian beta mixture modeling [43] , [44] , [47] . The method automatically infers the optimal number of clusters , based on an initial overestimation of their expected number ( ten , unless specified otherwise ) . In this MM sample , SciClone detected three clusters ( Fig . 1 ) : the first with average VAF of 46 . 1% , and two lower-frequency clusters with average VAFs of 32 . 0% and 11 . 9% ( Fig . 1 ) . Each cluster is represented by a posterior predictive density ( see Materials and Methods ) , which provides the probability of a VAF given the observed data ( and subsequent model fit ) . These densities probabilistically define boundaries between clusters , including the visually ambiguous separation between the highest-frequency cluster and the cluster with average VAF of 32 . 0% . As a comparison for our results , we applied an MCMC method , PyClone [11] , [30] , to the same variants in copy-number neutral , LOH-free regions . PyClone recapitulated the presence of the minor clusters , though with increased computational demands ( Table S1 ) . We tested PyClone's ability to integrate copy-number altered SNVs expected to be in the founding clone and found that it often assigned these sites to independent clusters . As a second point of comparison , we applied THetA [32] , a method that infers clonal architecture based on copy number events alone . For reasons of computational complexity , we applied it to a limited number of segments with representative copy number states . THetA detected clonal amplifications that occurred in 89 . 8% of the cells ( 44 . 9% VAF ) and a subclonal deletion that occurred in 54 . 1% of the cells ( 27 . 0% VAF ) . We integrated these data and SciClone clustering of all CNA and SNV VAFs revealed that the THetA-inferred CNA VAFs support the presence of subclones originally inferred from SNVs alone ( Fig . S1 ) . In some samples , ordering of subclonal VAFs may reveal the clonal phylogeny of the tumor [48] . However , in this sample , the data are insufficient to distinguish between a branched phylogeny , in which the two subclones arose from independent cells within the founding clone , and a linear phylogeny , where the lower VAF subclone is descended from the higher VAF subclone . The latter case implies that all mutations in the higher VAF subclone are also present in the lower VAF subclone , as are all founding clone mutations . Using WES to both discover variants and obtain deep read counts for defining VAFs may be an attractive , direct approach to clonality analysis [10] , as it avoids the additional time and expense of WGS followed by targeted sequencing . However , while WES data captures the coding variants likely driving disease progression , their number may be insufficient to reliably infer clonal architecture , particularly for cancers with relatively low somatic mutation rates . To begin to address these considerations , we applied SciClone to whole-exome sequenced breast [49] and endometrioid carcinoma [50] cancer samples from The Cancer Genome Atlas ( TCGA ) project . To obtain a sufficient number of variants , we relaxed the minimum depth of coverage requirement to 50x , resulting in 29 copy-number neutral variants from the breast sample and 53 from the endometrial sample . The breast tumor has a high-VAF cluster corresponding to its founding clone as well as a subclonal cluster , with most variants occurring in the latter ( Fig . 2a ) . The endometrial sample is more complex , with both a high-VAF cluster and three tightly-grouped and poorly-separated subclonal clusters ( Fig . 2b ) . Drawing inferences about mutation clonality ( e . g . , assessing whether mutations generally occur in the founding clone and hence are likely to be early , disease-initiating events [14] or attempting to correlate subclonal mutations with clinical outcome [10] ) requires accurately and confidently assigning individual VAFs to clusters . Our variational Bayesian approach does so via “fuzzy” cluster assignments , which describe the ( conditional , posterior ) probability that a VAF belongs to a particular cluster ( given that it belongs to one of them ) . In particular , the likely driver PIK3CA mutations in the endometrial sample are assigned confidently to the highest-frequency cluster one , with probabilities of 93 . 1% for the lower VAF variant and 97 . 1% for the higher . In contrast , the potential driver ATM mutation is nearly as likely to belong to cluster one ( 42 . 1% probability ) as to the lower VAF cluster two ( 57 . 8% probability ) to which it was “assigned” ( i . e . , that which maximized its posterior probability ) . Given the relatively few SNVs , this ambiguous assignment indicates that the data are insufficient to accurately define the clonal structure and that the separation between cluster one and cluster two may be an artifact of sparse data . This uncertainty might be resolved by increased depth of sequencing or by additional clonal markers ( e . g . , as discovered by WGS ) . Nevertheless , the strong assignments of the PIK3CA mutations to a cluster with average VAF near 50% suggest that , despite the relatively high level of noise in the data , they belong to the founding clone . Tumors evolve in response to treatment , both through loss of specific mutations and acquisition of new ones . Understanding this process in the context of a tumor's clonal architecture is critically important in defining mechanisms of resistance and in informing therapeutic decisions . To better understand mechanisms of therapy resistance , we extended our method to accommodate multiple samples and applied it ( Fig . 3 ) to samples from a primary AML tumor and post-treatment relapse occurring 26 months after chemotherapy [5] . These primary and relapse tumors were initially sequenced to depths of 26 . 7x and 31 . 5x , respectively , with subsequent capture validation providing deep read counts for all discovered variants ( median depth: 753x ) . All variants were analyzed , as no CNAs are present in either sample . Analysis of the primary tumor sample in isolation ( Fig . 3c ) suggests a simple organization consisting of a single subclone and a founding clone containing an IDH2 R140L mutation . Mutations in this residue may play a role in oncogenesis , given their recurrence in AML [51] and resulting neomorphic enzymatic activity [52] . Hence , this clonal mutation is an attractive target for small molecule inhibitors , such as those reactive against IDH2 R140Q [53] . However , simultaneous analysis of the relapse genome further dissects the apparently homogeneous highest-frequency cluster harboring IDH2 R140L into three distinct subpopulations of cells ( Fig . 3a ) : one that is effectively eliminated by chemotherapy ( cluster three , average relapse VAF ) , a second diminished by treatment ( cluster two , average relapse VAF 11 . 6% ) , and a third largely unaffected by treatment ( cluster one , average relapse VAF 41 . 3% ) . As further evidence of their high degree of overlap in the tumor sample , their respective average VAFs in the tumor are 42 . 7% , 43 . 1% , and 44 . 9% . The additional resolution provided by the relapse sample distinguishes these subpopulations to expose a more complex clonal architecture ( Fig . 3d ) and indicates that the IDH2 R140L mutation in cluster two is subclonal . Thus , targeting it therapeutically would be unlikely to eradicate the founding clone . We do observe that the subclonal mutations in cluster five were eliminated by treatment , suggesting that it carried a lower proclivity for resistance than the surviving clones . Remarkably , there is a second , independent IDH2 mutation ( R140W ) in the relapse sample . But , as above , defining its clonality from this sample alone ( Fig . 3b ) is confounded by an inability to associate its VAF ( 32 . 8% ) unambiguously with either the founding clone or a subclone . This uncertainty is resolved through multidimensional analysis that incorporates the tumor sample and places the mutation in cluster four . Mutations within this cluster , including IDH2 ( R140W ) , were either present in the primary tumor below the level of detection or are new mutations , possibly induced by cytotoxic chemotherapy [5] . In either case , they are potential drivers of disease progression . Given the clonal complexity of this sample , we next asked how many variants were required to capture this complexity and whether we were likely to have missed additional complexity . To address these concerns , we randomly selected a subset of the original variants and performed clustering . The number of clusters inferred as a function of the number of variants analyzed is fairly constant for variants ( Fig . 4a ) , whereas it drops precipitously for variants . As sequencing detected a total number of variants within the flat regime of this curve , we can be confident that no subclones with a higher VAF than the most infrequent cluster identified ( average VAF ∼12% ) were missed . Further , this suggests that ∼200 variants would have been sufficient to reveal this sample's clonal architecture . To assess the sensitivity of our approach in inferring the separation of clusters , we performed one-dimensional analysis of VAFs from relapse sample clusters one and two ( Fig . 3 ) after varying their inter-cluster separation ( Fig . 4b ) . While the results are sample-dependent , they indicate that clusters can be reliably distinguished if they lie greater than ∼7% away from one another . To ensure that the inferred number and composition of clones were not overly sensitive to our computational method , we varied both the number of initial clusters and the clustering approach itself . Consensus clustering indicated that the ( subjectively ) correct number of clusters ( five ) was inferred by the variational Bayesian beta mixture modeling for a range of initial number of clusters from six to 15 ( data not shown ) . We next used SciClone to cluster using a variational Bayesian binomial mixture model and a previously-published [40] , [54] variational Bayesian Gaussian mixture model ( see Text S1 ) . Consensus clustering indicates that the results are stable for the majority of variants as both the number of initial clusters and the method ( beta , binomial , or Gaussian ) are varied ( Fig . 4c ) . The few variants that clustered differently between methods ( Fig . S2 ) were situated near cluster boundaries or between clusters . A similar effect was seen when clustering the data with PyClone ( Table S1 ) , though in this case variants along cluster boundaries tended to coalesce into independent clusters ( clusters six and eight in Fig . S3 ) : PyClone's default hyperparameter settings lead it to overdissect the founding clone . After increasing the number of iterations from 10 , 000 ( with a burn-in of 1 , 000 iterations ) to 100 , 000 iterations ( with a burn-in of 10 , 000 iterations ) , PyClone results were even more similar to those obtained with SciClone , but the former still split the highest-VAF cluster into two ( data not shown ) . Despite these differences , the results are largely consistent between SciClone and PyClone and we have increased confidence in variants that are similarly assigned by both approaches . Spatial heterogeneity complicates the analysis of solid tumors , as distinct regions of a tumor may harbor different subclonal populations [9] , [13] , [16] . Assaying multiple regions of heterogeneous tumors should assist in uncovering the full spectrum of mutations and subclones present in a tumor and help identify the spatial origins of subclones that give rise to therapy resistance . To investigate this effect , we analyzed two pre-treatment biopsies from the same breast tumor and added a temporal dimension by examining a single sample from the tumor collected 16 weeks after aromatase-inhibitor ( AI ) therapy . Mutations in the three samples had median coverage of 130 . 5x from deep capture sequencing . Three-dimensional clustering with SciClone revealed five distinct groups of mutations and a fairly low purity , resulting in reduced VAFs in all samples ( Fig . 5 , Movie S1 ) . Differences between the two pre-treatment biopsies were captured in clusters four and five , containing region-specific mutations . Cluster two cannot be identified from pre-treatment tumor one alone , but the second biopsy reveals it as a distinct subpopulation of cells with higher VAF in the first biopsy ( 36 . 03% vs . 8 . 13% ) . The effect of AI therapy is revealed by inclusion of the post-treatment sample , in which clusters two and four are eliminated . These likely represent AI-responsive subpopulations of cells , though additional spatial heterogeneity leading to their apparent removal cannot be discounted . Cluster five contains mutations specific to the second biopsy; while some of the cells harboring them expanded in the post-treatment sample , others appear to have been eradicated completely . The heterogeneity in response observed in this cluster suggests that it actually encapsulates several distinct , but overlapping subclonal populations that occur at similar VAFs in the pre-treatment biopsy and are difficult to separate without additional data . Application of PyClone to these data ( Table S1 , Fig . S4 ) reveals several significant differences . While it infers two distinct clusters from the heterogeneous cluster five , it also partitions variants in the founding clone into two clusters . This separation is likely a clustering artifact , since ( 1 ) the two clusters are merged when all of the data ( in copy-altered and -neutral regions ) are clustered using 34 , 000 iterations ( data not shown ) and ( 2 ) the presence of two large , independent clusters comprising 70% of the cellular population each is biologically unreasonable . The discordance between methods suggests that the limited number of variants affected require special attention .
Clonal heterogeneity complicates both our understanding of the biology of tumors and the design of effective treatment strategies . While an individual tumor sample provides a glimpse of this complexity , additional temporally or spatially distinct samples allow higher resolution mapping of subclonal architecture , including the isolation of drug-sensitive clones and small subpopulations driving relapse . To leverage the increasingly commonplace and cost-effective opportunities to sequence multiple samples from an individual , we developed SciClone , which scalably analyzes large numbers of samples to provide an unbiased , probabilistic dissection of a cancer's clonal landscape . To do so , SciClone employs variational Bayesian mixture modeling of beta , binomial , and Gaussian distributions . Each of these may have advantages ( see Text S1 ) in certain situations , though our tests suggest that the beta mixture model works best in practice . We have previously used related techniques in analyzing FACS data [55] and expect them to be of general interest to those requiring methods that automatically and efficiently infer the number of clusters from high-dimensional biological data . Application of SciClone to primary and relapse AML tumors identified subclonal populations with dramatically divergent response to conventional therapy . Such analyses are the first step towards inferring driver mutations responsible for both resistance to therapy and clonal expansion following treatment . Insight into the spatial origins of treatment response was provided by analysis of three samples from a breast tumor , two of which were obtained from distinct regions of a single tumor at the same time point . The AML and breast cancer cases highlight an inherent limitation of bulk sequencing of tumor cells: subclonal populations cannot be distinguished if they occur at similar frequencies . Single-cell sequencing may eventually offer a solution , but will require dramatic improvements in fidelity and throughput . Using currently available data , we demonstrated that temporally or spatially distinct samples from the same tumor can be used to tease apart these overlapping subclones . This is demonstrated in AML , where the apparent founding clone in the primary tumor is dissected into two additional subclones by incorporating the relapse sample . The breast cancer samples exhibit two-fold complexity . As in the AML primary tumor , a cryptic subclone is revealed in the pre-treatment breast tumor when multiple samples are considered; in this case , from two spatially-isolated biopsies . Additionally , each pre-treatment sample exhibits a clone not detected in the other . This suggests that manipulation of the patient's tumor-derived cells ( e . g . , passage within culture or as mouse xenografts ) may be a viable method for identifying additional subclones and predicting those with differential responses to therapy . Our analysis of exome-sequenced cases showed that SciClone can be useful on samples with as few as 29 SNVs ( Fig . 2a ) , but our simulations ( Fig . 4a ) showed that in more complex cases , such as the AML tumor/relapse pair , establishing subclonal boundaries may require two hundred or more variants and that subclones be separated by VAFs of ∼7% or more ( Fig . 4b ) . This downsampling approach may be applied to any data set to establish a baseline sensitivity . Cases with poorly defined cluster boundaries ( e . g . , due to a paucity of mutations ) , such as the endometrial case ( Fig . 2b ) , benefit from SciClone's probabilistic formalism . In particular , by assigning an ATM mutation similar probability of belonging to the founding clone and a subclone , SciClone reflected the lack of certainty inherent in the data and indicated that their sparsity may poorly characterize the tumor's clonal diversity . The sensitivity of any clustering method in dissecting clonal boundaries is dependent on cluster overlap , which we have characterized via the “uncertainty” of their probabilistic assignments ( Refs . 45 and 46 and Materials and Methods ) . An additional , qualitative means of detecting high-confidence variant/cluster assignments involves taking the consensus , or intersection , across clustering methods ( Fig . 4c ) . Confidence in detecting all major subclones increases with the number of variants , including passenger mutations more likely to be missed by exome sequencing . Thus , WGS followed by deep validation sequencing is most likely to capture the full spectrum of mutations and yield high-quality characterization of subclonal entities . Next-generation sequencing of variants within copy-number neutral regions of autosomal chromosomes leads to a straightforward interpretation of the inferred VAFs as half the cellular frequency harboring the corresponding variant . Because of the widespread availability of variants to serve as clonal markers and the relative reliability of their bioinformatic analysis and quantification , our initial clonality analyses have focused on SNVs . Nevertheless , other genomic events have been used to identify clonal dynamics . For example , the alternate “waxing” and “waning” of subclonal CNAs has been observed in multiple myeloma [17] , [33] . However , the analysis and discovery of CNAs pose several challenges for clonality: ( 1 ) Cancer types may be described hierarchically in terms of their propensity to elicit either mutations or copy number changes [56] . For mutation-dominated cancer types , such as the cytogenetically-normal AML analyzed here , few CNA events may be available for analysis . The converse does not apply: since SNVs accumulate with age [4] , an abundance of SNV clonal markers are expected in all malignant , as well as in non-malignant , tissues . Given the density of SNVs , clonality analyses that rely solely on them may well capture the full clonal architecture , while missing specific ( copy number ) events of pathogenic interest; clonality analyses relying solely on copy number events are likely to miss both . ( 2 ) There is no digital readout of CNAs , rather observed copy number reflects the admixture of subclonal populations and is a ( linear ) combination of the copy number state of each subclone , weighted by the fractional subpopulation of the subclone . In principle , the correctness of the analysis requires the simultaneous inference of this admixture and the number of copies ( 0 , 1 , 2 , 3 , or greater ) of each chromosomal segment in each subclone . Though such an analysis would infer the clonal hierarchy directly , rather than the clusters of variants that serve to identify them as in a SNV-based analysis , inference in the latter case is simplified since there are fewer mutational states ( presence or absence , at least of the vast majority of variants , which are heterozygous ) and the correctness of inferring one cluster is independent of a second cluster . For these reasons , we prefer to overlay CNA events on the higher confidence copy-number neutral SNV VAFs . Incorporating such events is important: ( 1 ) to include SNVs from CNA regions , which are especially likely to be involved in disease , and ( 2 ) to ameliorate the loss of SNVs from copy-number neutral regions occurring as the number of genomic regions perturbed by CNA ( in some sample ) increases with the number of samples analyzed . Accommodating these events may be accomplished: ( 1 ) by determining the fractional population harboring the event ( as in Fig . S1 ) or ( 2 ) by adjusting a SNV's VAF based on it's inferred copy number states across subclonal populations . One approach to the latter involves inferring copy number states from the B-allele frequencies of germline SNPs ( e . g . , using ASCAT [34] or APOLLOH [39] ) and phasing these to somatic variants ( e . g . , by detecting a single sequencing read spanning both ) to impute subclone-specific copy numbers to each variant [6] . After adjusting the SNV VAFs , they could be clustered by SciClone in a manner completely analogous to the analysis of unadjusted VAFs ( using the beta or Gaussian mixture model approaches ) . We are currently pursuing this approach . MCMC techniques , such as PyClone [11] , [30] , offer an alternative approach to clustering variants . However , our comparisons of SciClone and PyClone ( Table S1 ) reinforce the computational inefficiencies of MCMC approaches relative to variational Bayesian techniques [40] and show that SciClone is between one and two orders of magnitude faster . SciClone inherits the simple variational Bayesian ( computational ) convergence condition of monitoring monotonic changes in a lower bound ( see Text S1 ) . While this approach may converge to a local extremum , more subtlety is required to ensure the ( theoretical ) asymptotic convergence to the global extremum guaranteed by MCMC , e . g . , monitoring variance within a Markov chain relative to variance between independent Markov chains [41] . PyClone provides no direct facilities to monitor convergence . Regardless , the theoretical convergence properties of MCMC seem unjustified given the involved computational overhead for a clustering application , such as clonality detection , where error estimates of the parameters are of marginal interest . SciClone has already contributed to the understanding of biological mechanisms underlying cancer and has the potential for increased utility with the advent of clinical sequencing . Towards this end , we are developing methods that cross-reference the clonal status of specific mutations with databases of targeted therapeutics . As an example , the Drug-Gene Interaction Database [57] identifies three genes in the AML sample as as potentially druggable: ( DRD2 , KCNQ2 , and P2RY2 ) . The fact that each of these mutations lies in a subclone complicates their interpretation , and suggests that careful study is needed to understand how specific subclonal populations respond to different therapeutics . While clinical decisions of which ( sub ) clones to target and how remain controversial , it is clear that making these decisions will require accurate assessment of clonal architecture using tools such as SciClone .
A VAF is defined with respect to the number of reads , , supporting the variant allele and the number of reads , , supporting the reference ( or non-major-variant , in the case of multiple variants ) allele: . Our previous method [14] considered variants in a single sample and modeled the probability of a VAF belonging to cluster aswhere is the gamma function . Here , we extend this to the case with samples by defining the -vector , whose component , , is the VAF of that variant in the sample . We make the assumption that , within a cluster , the VAFs are independent across samples , so that the cluster may simply be modeled as ( 1 ) where and are the -vectors whose components are and , respectively . This implies only that within a cluster there is no correlation between samples . The validity of this assumption is indicated by the visually-observed orthogonality of the VAF principal component axes to the coordinate ( i . e . , sample ) axes . We have rarely , if ever , seen evidence for such intra-cluster correlation . Nevertheless , this assumption may be relaxed through use of a mixture of multivariate Gaussian distributions ( see Text S1 ) , each of which has a full-rank covariance matrix . In considering a mixture of ( multi-dimensional ) beta components [Eq . ( 1 ) ] , we introduce a -dimensional latent ( or unobserved ) binary random variable indicating whether VAF does ( ) or does not ( ) belong to component and satisfying a 1-of- representation in which . The marginal probability that a VAF belongs to component is given by its mixing coefficient , subject to the probabilistic constraints Given the 1-of- representation of , this may be written as ( 2 ) Similarly , the conditional distribution that a VAF arises from the mixture may be written ( 3 ) in terms of the shape parameter vectors and of the beta component , with aggregate parameters and . To accommodate binomial and Gaussian mixture models in addition to the beta mixture model , we introduce abstract notation used below to define quantities ( e . g . , -values ) independently of the concrete representation of likelihoods and posterior distributions . We begin by defining abstract parameters , which differ according to the model , i . e . , beta , binomial , or Gaussian . For example , , with . Further , while the Gaussian mixture model is also a function of VAFs , the binomial mixture model is defined with respect to the variant and reference count vectors , and , respectively . To abstract away these details , we use the notation to denote the VAFs of a beta or Gaussian mixture model or the counts and of a binomial mixture model , when convenient . In particular , , while , with . Eqns . ( 2 ) and ( 3 ) extended across the entire set of VAFs ( or , more abstractly , of data ) and their associated latent variables are combined to express the complete-data ( i . e . , including the latent variables ) likelihood ( 4 ) which may be summed over to give the incomplete likelihood These equations could be used to fit the beta parameters using expectation maximization ( EM ) or Markov chain Monte Carlo ( MCMC ) techniques . We instead use a previously described [43] , [44] variational Bayesian approach [40] , [42] , [54] , [58] , [59] to modeling a mixture of beta distributions . The general variational Bayesian theory and its application to mixture modeling are described in depth in several excellent references [40] , [42] , [54] , [58] , [59] . To establish consistent notation , we provide an abridged , but self-contained , introduction to this general theory and to its application to Gaussian [40] , [54] , [58] , [59] and binomial mixture models in Text S1 . Here , we summarize its application to beta mixture models to provide sufficient context for its use in and extension for clonality analysis . For full details of this derivation , the reader is referred to the original references [43] , [44] . Variational Bayesian beta mixture modeling approximates the posterior distribution , , over the model parameters , , and and the latent variables with a distribution . The form of this approximate distribution is a consequence of choice of prior distribution , whose product with the likelihood [Eq . ( 4 ) ] defines the posterior according to Bayes' theorem , and of the mild and standard [40] assumption that the latent variables factorize from the model parameters , i . e . , . This further simplifies , without assumption , to . Finally , the authors assume the and variables are independent and factorize to ultimately give . Ma and Leijon used four synthetic one-dimensional data sets ( Fig . 4 of Ref . 43 ) , including two with highly overlapping beta distributions , to demonstrate the high accuracy of this method despite its assumption that the parameters of the beta distribution are independent . Fan et al . [44] additionally analyzed six three-dimensional data sets and similarly found that accuracy was not adversely effected by this factorization approximation ( Table I of Ref . 44 ) . Our own extensive simulation results further support these findings . We generated data sets by sampling a mixture of beta distributions in one , two , or three dimensions and having between two and five clusters ( 100 data sets for each dimensionality/number of clusters pair ) . Fig . S5 shows the concordance ( i . e . , fraction of correctly assigned items ) between the clustered and known results for each simulated data set . The average concordance is , , and in one , two , and three dimensions , respectively . Performance increases with dimensionality as the clusters become increasingly separated . This sparsity may be quantified by the minimum cluster self-overlap ( see below ) . Data sets having a relatively small minimum cluster self-overlap have a relatively large overlap between clusters , which leads to uncertainty and degrading performance . The prior distributions are generally selected to be conjugate to the likelihood for analytic convenience ( e . g . , see the derivations of the variational Bayesian Gaussian and binomial mixture models in Text S1 ) . While a conjugate prior to the beta likelihood is formally available , its use would lead to an analytically intractable integration [43] . Therefore , Ma and Leijon [43] instead propose use of the following prior distribution ( 5 ) where and are gamma distributionsand is the Dirichlet distributionover proportions , with the normalizing constant and The parameters of the approximate posterior distribution are now determined by iteratively minimizing the Kullback-Leibler divergence , a measure of the difference , between it and the posterior distribution , following the general prescription of variational Bayesian inference ( see Text S1 ) . The authors make a non-linear approximation to an expectation value arising during the iterative procedure so that the resulting , approximate posterior distribution has the form of a gamma distribution , despite the fact that the above gamma prior distribution is not conjugate to the beta likelihood . Significantly , the authors show that this additional approximation can be used to minimize the original , desired Kullback-Leibler divergence between the posterior distribution and the approximate , non-gamma posterior distribution . This results in the approximate posterior distribution ( 6 ) where , , , , and are defined in Eqns . 47–51 , respectively , of Ref . 43 . These parameters are updated from the corresponding initial hyperparameter values , , , , and as in a traditional EM iterative algorithm . It will also be convenient to define the posterior density with respect to the component ( 7 ) Variational Bayesian mixture models provide probabilistic assignments of variant ( i . e . , a VAF for beta or Gaussian mixture models or variant counts for a binomial mixture model ) to cluster according to the posterior probabilities . The act as “responsibilities” and satisfy In the case of the beta mixture model , the are defined by Eqns . 31 and 32 of Ref . 43 . A more general derivation is provided in Text S1 , along with specific calculations for binomial and Gaussian mixture models . For visualization purposes , for example , we occasionally transform these probabilistic assignments into hard assignments , which assign to one and only one cluster according to The posterior predictive density gives the probability of a new ( i . e . , unobserved ) variant , , given the observed data and all possible assignments of that variant to a cluster . Evaluating the sum over , making use of Eq . ( 2 ) , gives We next approximate the true posterior distribution , , with the variational approximation to give Since for all mixture models considered in this manuscript , with andthis evaluates to ( 8 ) Ma and Leijon [43] assumed that , where is the Dirac delta function and are the converged parameter values , i . e . , that the posterior distribution has negligible probability when any of the parameters differ from their converged values . In this case , which may be efficiently evaluated . We instead use Eq . ( 8 ) , which avoids any assumption on the approximate posterior distribution . In the case of binomial and Gaussian mixture models , Eq . ( 8 ) may be evaluated analytically . In the case of a beta mixture model , for which is given by Eq . ( 1 ) and is given by Eq . ( 7 ) , we instead resort to numerical integration , evaluating data sampled from Eq . ( 7 ) via Eq . ( 1 ) . We choose hyperparameters resulting in prior distributions sufficiently broad to ensure that the number of clusters and their posterior parameterization are determined primarily from the data rather than from prior assumptions . In particular , following Ma and Leijon [43] , we choose for all . We also choose and for all . Given the latter choice , the gamma distributions and collapse to exponential distributions . The resulting variances of these distributions , e . g . , , are large given our choice of hyperparameters and hence provide a broad prior . We initialize the according to the hard assignments computed by -means ( provided in the R stats package and using default parameters , except with and ) . We initialize the parameters , , and to their respective hyperparameter values , , and . Finally , we initialize the such that the expected means of the cluster centers , , with and , are set to the values returned by -means . We then perform the variational E step ( i . e . , calculate the expectations immediately following Eq . 51 of Ref . 43 ) without updating the , followed by the variational M step to update the parameters , , , , and ( via Eqns . 47–51 of Ref . 43 ) . For the AML28 data set , this initialization results in the clusters shown in Fig . S6a . Initialization is followed by iteratively applying the variational E step ( including updating the ) and M step . To avoid undefined behavior in evaluating the beta distribution , we shift VAFs at zero or one by or , respectively , with equal to machine precision . Variational Bayesian mixture modeling generally discards clusters that do not contribute to the model , as determined by the data and strength of the prior distribution . Specifically , following convergence of the variational iteration and hard assignments of variants to clusters , we remove any clusters having less than the larger of three variants or 0 . 5% of , the total number of variants , assigned to them , a condition similar to our earlier approach [14] . If clusters are removed , the algorithm is again executed until convergence . For the beta mixture model , convergence is achieved when the absolute difference between all across consecutive iterations is less than . This condition differs slightly for binomial and Gaussian mixture models ( see Text S1 ) . The minimum cluster membership is motivated by the requirement of needing at least two proportions to fix the two degrees of freedom , and , of a beta distribution . More intuitively , clustering is effectively a separation of intra- and inter-cluster distances . Defining an intra-cluster distance requires at least two items be assigned to that cluster . To be conservative in our assessment of subclonality , we require clusters be well separated . Previously [14] , we used a condition on overlapping cluster standard error of the means to detect and remove overlapping clusters . Here , we instead adopt a quantitative notion of cluster overlap [45] , [46] , in which overlap between clusters and results in uncertain assignments of some variants , causing them to have appreciable and . This in reflected in a large relative ( to the “size” , , of cluster ) cluster overlap Minimizing this quantity for all is equivalent to maximizing the “self-overlap” of cluster with itself , which satisfies , . Hence , we remove any cluster having a less than a threshold . Overlap between ( independent ) clusters will be more likely in lower dimensional problems; hence , to determine a dimensionality-dependent we clustered simulated data sets by sampling a mixture of beta distributions in one , two , or three dimensions and having between two and five clusters . Average concordance ( across data sets of a given dimensionality ) between the clustered and known results ( in terms of fraction of correctly assigned items ) was stable for a wide range of within each dimension: in the range of to achieved the maximal concordance ( of ) in one dimension , in the range of to achieved a concordance of in two dimensions , and in the range of to achieved a concordance greater than or equal to in three dimensions . Intuitively , we anticipate that the probability of clusters overlapping scales inversely with the number of dimensions . Hence , we define for an -dimensional problem as , where was selected so that passes through the above optimal regions defined by the simulation . Namely , and . Results for these settings of across all simulated data sets are shown in Fig . S5 . We detect outliers using a more formal approach than our previous method [14] , by calculating the -value of a variant with the respect to the cluster to which it has been assigned ( via a hard assignment ) . If the probability of the variant belonging to that cluster is less than , the variant is removed from the analysis . The default used in this manuscript is ( which is not corrected for multiple testing ) . The -value of a variant is calculated with respect to the predictive posterior distribution [Eq . ( 8 ) ] aswhere is the Heaviside step function with for and for . In the case of beta mixture models , this integral is evaluated numerically by sampling from the predictive posterior distribution and then evaluating sampled variants with that distribution , which again involves numerical integration . For computational efficiency , we only calculate this integral for variants likely to be outliers , which we heuristically define as variants whose VAF in each sample lies outside of the narrowest interval containing of the fluctuation in the mean of cluster . This interval is determined as the narrowest such interval satisfyingand involves integrating the mean , , with respect to the posterior distribution . Several iterations of the AML28 data set following the -means initialization ( above and Fig . S6a ) are shown in Figs S6b and c , with the complete run shown in Movie S2 . Sequencing , alignment , and variant calling were performed as previously described [5] . Somatic copy number events were detected using copyCat ( http://github . com/chrisamiller/copycat/ ) . Copy-number neutral LOH was detected using VarScan 2 [38] and filtered to retain regions with 95% LOH and at least 10 sites . PyClone version 0 . 12 . 3 was downloaded from http://compbio . bccrc . ca/software/pyclone/ . VarScan 2-detected regions of LOH were excluded from analysis . Copy number events detected by copyCat were quantized and passed to PyClone as major_cn , with minor_cn set to zero; additionally , PyClone was run with ––var_prior total_copy_number , since allele-specific copy number calls were not provided . PyClone clustering used the beta-binomial mixture model . Initially , we attempted to cluster using 10 , 000 iterations and 1 , 000 burn-in iterations , as suggested by the authors ( Ref . 30 and https://bitbucket . org/aroth85/pyclone/wiki/Tutorial ) . However , these parameters yielded discordant clusterings across three runs . Therefore , we varied the number of total iterations ( and additionally varied the number of burn-in iterations to be 10% of the total iterations ) and for each configuration assessed concordance across three independent runs . The authors have suggested similar approaches based on visual inspection of convergence across randomly-initialized runs [30] . We choose the number of iterations at which the concordance across the three runs stabilized . These are given in Table S1 . Concordance was evaluated for each of the three pairs and was calculated as the maximal fraction of items assigned to the same cluster across permutations of the cluster labels of one of the two runs being compared . THetA version 0 . 51 was downloaded from http://compbio . cs . brown . edu/projects/theta/ . After failing to successfully run the program on the complete set of copy number events in the MM sample , we selected seven copy number regions , representing neutral , amplified , and subclonally deleted chromosomes , and ran THetA as described in the manual ( parameters: –n 3 –k 4 –m 0 . 10 ––NUM_PROCESSES 2 ) . The resulting population frequencies and copy number assignments were used to infer the VAF at which a SNV in that region would appear . These sites were added to the list of SNV inputs to SciClone and clustered with default parameters .
|
Sequencing the genomic DNA of cancers has revealed that tumors are not homogeneous . As a tumor grows , new mutations accumulate in individual cells , and as these cells replicate , the mutations are passed on to their offspring , which comprise only a portion of the tumor when it is sampled . We present a method for identifying the fraction of cells containing specific mutations , clustering them into subclonal populations , and tracking the changes in these subclones . This allows us to follow the clonal evolution of cancers as they respond to chemotherapy or develop therapy resistance , processes which may radically alter the subclonal composition of a tumor . It also gives us insight into the spatial organization of tumors , and we show that multiple biopsies from a single breast cancer may harbor different subclones that respond differently to treatment . Finally , we show that sequencing multiple samples from a patient's tumor is often critical , as it reveals cryptic subclones that cannot be discerned from only one sample . This is the first tool that can efficiently leverage multiple samples to identify these as distinct subpopulations of cells , thus contributing to understanding the biology of the tumor and influencing clinical decisions about therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"cancer",
"genetics",
"breast",
"tumors",
"obstetrics",
"and",
"gynecology",
"cancer",
"treatment",
"cancers",
"and",
"neoplasms",
"hematologic",
"cancers",
"and",
"related",
"disorders",
"oncology",
"genome",
"sequencing",
"women's",
"health",
"genome",
"analysis",
"molecular",
"biology",
"techniques",
"breast",
"cancer",
"molecular",
"biology",
"hematology",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"computational",
"biology",
"genomic",
"medicine"
] |
2014
|
SciClone: Inferring Clonal Architecture and Tracking the Spatial and Temporal Patterns of Tumor Evolution
|
One of the most obvious phenotypes of a cell is its metabolic activity , which is defined by the fluxes in the metabolic network . Although experimental methods to determine intracellular fluxes are well established , only a limited number of fluxes can be resolved . Especially in eukaryotes such as yeast , compartmentalization and the existence of many parallel routes render exact flux analysis impossible using current methods . To gain more insight into the metabolic operation of S . cerevisiae we developed a new computational approach where we characterize the flux solution space by determining elementary flux modes ( EFMs ) that are subsequently classified as thermodynamically feasible or infeasible on the basis of experimental metabolome data . This allows us to provably rule out the contribution of certain EFMs to the in vivo flux distribution . From the 71 million EFMs in a medium size metabolic network of S . cerevisiae , we classified 54% as thermodynamically feasible . By comparing the thermodynamically feasible and infeasible EFMs , we could identify reaction combinations that span the cytosol and mitochondrion and , as a system , cannot operate under the investigated glucose batch conditions . Besides conclusions on single reactions , we found that thermodynamic constraints prevent the import of redox cofactor equivalents into the mitochondrion due to limits on compartmental cofactor concentrations . Our novel approach of incorporating quantitative metabolite concentrations into the analysis of the space of all stoichiometrically feasible flux distributions allows generating new insights into the system-level operation of the intracellular fluxes without making assumptions on metabolic objectives of the cell .
Metabolic fluxes give immediate insights into the metabolism of a cell [1] , [2] . Metabolic flux analysis has proven to be useful , for example for the determination of enzyme functions [3] , for the identification of regulatory mechanisms in response to environmental perturbations [4] , or as a tool in metabolic engineering [5] . The most common method to quantify metabolic fluxes uses labeled substrates , and the measured label distribution in intracellular metabolites is interpreted together with measured uptake and production rates by means of a metabolic network model [6] . Despite successful quantification of fluxes with flux analysis in different conditions , the method has several limitations . For example , it is limited to the main branches in central carbon metabolism and fluxes cannot be resolved per compartment [7] , despite compartmentation being a highly relevant aspect of eukaryotes [8] . Moreover , today's flux analysis rests on a number of a priori assumptions , e . g . on reaction reversibilities or on relevant parts of the network [7] , [9] . To improve flux quantification , we need additional constraints on the possible flux distributions in a metabolic network . In stoichiometric network analysis , the metabolic network is modeled as a collection of biochemical reactions where all internal metabolite concentrations are assumed constant [10] . Next to the typical constraints , such as uptake and excretion rates , reaction reversibilities and maximum flux capacities , the field recently began to incorporate thermodynamic information , whereby statements on feasibility of reaction fluxes or flux distributions can be made based on calculation of changes in Gibbs energy using metabolite concentrations [11]–[13] . For example , using flux balance analysis ( FBA ) and related approaches , metabolite concentrations were used as additional constraints to predict fluxes in the non-compartmentalized organism E . coli [14] , [15] or in a model of liver metabolism [16] . Here , we develop a novel approach to integrate metabolite data into metabolic network flux analysis , to get additional insight into the compartmentalized flux physiology of Saccharomyces cerevisiae . The method combines network embedded thermodynamic ( NET ) analysis [12] , elementary flux mode ( EFM ) analysis [17]–[19] , and experimentally determined metabolome data . We employ EFM analysis instead of flux balance analysis because the collection of the generated flux modes can yield insight into all feasible flux distributions , as compared to the single thermodynamically feasible flux solution that is obtained with thermodynamically constrained FBA . Additionally , assumptions on a metabolic objective function of the cell are not required . Our new approach to analyze the compartmentalized central metabolic network of S . cerevisiae using quantitative metabolite data acquired under glucose batch growth conditions allowed us to generate novel insight into the system-level causalities underlying the intercompartmental redox metabolism . Specifically we show that the and NADH concentrations in the cytosol and the mitochondrion do not allow for the ethanol-acetaldehyde redox shuttle to be active under the investigated condition . Further , we identified a number of maximal reaction activities that could be used as constraints for flux analysis or FBA . We envision that our method becomes a useful tool to unravel system-level insights about a complex metabolic system from metabolome data .
The mass balanced flux solution space of a stoichiometric metabolic network can be described with a non-negative linear combination of its EFMs: ( 1 ) where any flux distribution is a sum of EFMs with coefficients . As we show in the proof provided in Text S1 , a thermodynamically feasible flux distribution only consists of thermodynamically feasible EFMs: ( 2 ) where the thermodynamically feasible flux distribution is only composed of EFMs from the feasible set , . The mathematical proof demonstrates that by eliminating infeasible EFMs we do not loose feasible flux distributions , because any feasible flux distribution can be composed of only feasible EFMs . Specifically , in the hypothetical case that an infeasible EFM is part of a feasible flux distribution it must involve a cancellation or directionality change of a specific reversible reaction . In this case , the flux distribution can be decomposed into one or more feasible EFMs , and a feasible combination of an infeasible EFM with another EFM . The combination of the infeasible EFM and another EFM must then be either a feasible EFM by itself , or it must be possible to achieve the feasible combination by other feasible EFMs . Hence , Eq . ( 2 ) allows us to exclude thermodynamically infeasible EFMs from the complete set of EFMs without excluding thermodynamically feasible flux distributions . It is important to note that the resulting flux solution space defined by the feasible EFMs can still contain infeasible flux solutions because it is possible that the combination of multiple feasible EFMs leads to an infeasible flux distribution . Exploiting that EFMs allow us to exclude thermodynamically infeasible EFMs , we aimed at developing an approach to generate novel insights into the complex flux physiology of the central metabolism of the yeast S . cerevisiae . Therefore , we assembled a 230 reaction stoichiometric network of its central carbon metabolism and amino acid synthesis pathways ( cf . Materials and Methods ) encompassing the cytosolic and mitochondrial compartments and many parallel pathways . With our approach we aim to obtain additional insight into the metabolic network operation , therefore we build upon current knowledge by defining the reversibilities in our model as they are defined in the original model [21] . First , we needed to calculate all EFMs . As the EFM calculation is computationally demanding , we initially applied steps to constrain the mass-balanced solution space as much as possible upfront , before we started enumerating EFMs ( Fig . 1 ) . Thus , in a first step , we performed flux variability analysis ( FVA ) [22] on the basis of the measured uptake and production rates of external compounds and biomass , to determine the reversibility of each reaction under the investigated physiological conditions , that is , for growth on glucose . From FVA , we obtain a minimum and maximum achievable flux for each reaction . A reaction is reversible if both a negative and positive flux can be achieved , else it is unidirectional . Reactions that have a minimum and maximum flux that are either positive or negative , and cannot be inactive , are reactions that are always active in the respective direction . Using this approach , we could classify 67 initially reversible reactions as unidirectional ( Fig . 2 , Dataset S1 ) . Next , we employed measured metabolite concentrations from glucose batch cultures , and NET analysis to identify additional reaction irreversibilities [12] , [23] . For the metabolite data , we assembled published and unpublished data from glucose batch experiments , and generated a consensus data set to define lower and upper concentration limits for 55 metabolites ( see Materials and Methods and Dataset S2 ) . For NET analysis , we used the reaction activities inferred from FVA . With the consensus metabolite data set , we obtained constraints on the reversibility of three additional reactions ( see Fig . 2 ) . Another iteration of FVA and NET analysis with the obtained constraints as input did not yield any further constraints . For the obtained condition-specific constrained metabolic network , we computed the EFMs and obtained 71 . 266 . 960 EFMs . Using NET analysis and the consensus metabolite data set , we classified 38 . 420 . 207 ( 54% ) EFMs as feasible and 32 . 846 . 753 ( 46% ) EFMs as infeasible . Assuming that the flux solution space in the metabolic network of an organism can be approximated by the number of EFMs , our result shows that roughly at most half of the solution space is thermodynamically feasible . With the finding that 54% of all the EFMs are thermodynamically feasible we reduced the number of EFMs that can constitute a thermodynamically feasible flux distribution considerably . In a first analysis step towards generating insights into the flux distribution , we searched the feasible EFM set for reactions that only use a subset of the possible reaction directions compared to the complete set of EFMs . For each reaction in each EFM we determined whether a backward , inactive or forward reaction activity was used . Then , the possible reaction activities for each reaction were compared between the complete set of EFMs and the feasible set of EFMs . Here , we found that the oxaloacetate transport from the mitochondrion to the cytosol is never used in an EFM of the feasible set , meaning that it has to be inactive under the investigated growth condition . Indeed , we find no contradicting evidence for the prediction when comparing with the experimental observation that a knock-out of the corresponding gene OAC1 , whose translated protein is responsible for the respective oxaloacetate transport reaction , does not have an effect on growth rate under glucose batch conditions [24] , [25] . Further , as we found that all EFMs with acetaldehyde transport out of the mitochondrion are infeasible , we conclude that during growth on glucose , acetaldehyde can only be transported into the mitochondrion . It is important to note that the EFMs with active oxaloacetate transport , or acetaldehyde transport out of the mitochondrion , are not infeasible because of the metabolite concentration constraints on the respective single reaction only , since single reaction infeasibilities are removed in the first NET analysis step before EFM generation ( see Fig . 1 ) . Instead , as we will show later , the infeasibility is the result of a system of coupled reaction activities , where all individual reactions need to be thermodynamically feasible simultaneously . Next , we wanted to test whether the feasible set of EFMs differs from the infeasible set in terms of reaction rates . Such a comparison is possible by normalizing the reaction rates in each EFM to the glucose uptake rate of the EFM ( all EFMs have glucose uptake , EFMs without glucose uptake are internal cycles and they were removed because they are physiologically meaningless [26] ) . Principal component analysis ( PCA ) of the normalized EFMs shows a clear difference between the feasible and infeasible EFMs in principal component 2 ( PC 2 in Fig . 3 ) . The reactions with the highest loadings in this component are alcohol dehydrogenase in both the cytosol and the mitochondrion ( ALCD2x , ALCD2m ) and acetaldehyde and ethanol transport to the mitochondrion ( ACALDtm , ETOHtm ) , and these reactions are likely involved in causing thermodynamic infeasibilities . Note , that although there is a separation between feasible and infeasible EFMs because there are no feasible EFMs in the center of the graph , by combination of feasible EFMs it could still be possible to obtain a feasible flux distribution that would be projected in this area of the PCA . Therefore , the loadings of PC1 are not considered . Next , we searched for the highest and lowest rate of each reaction in the complete set of EFMs and in the feasible set of EFMs to define the flux ranges that can be achieved in terms of flux per unit of glucose uptake . Any flux value in this range can in principle be achieved through a combination of EFMs . When comparing the flux ranges that can be realized by the feasible EFMs with the flux ranges of the complete set of EFMs , we find that eight reactions cannot assume the full range for thermodynamic reasons ( see Fig . 4 ) , with four of these reactions already having shown high loadings in the second principal component ( see Fig . 3 ) . These quantitative flux constraints result from metabolite concentrations and thermodynamics , and can be applied as constraints in flux balance analysis .
In this work , we demonstrated that an in vivo thermodynamically feasible metabolic flux distribution is only composed of thermodynamically feasible elementary flux modes . This EFM property allowed us to integrate EFMs and NET analysis into a novel approach to study the system-level properties of complex metabolic networks on the basis of quantitative metabolome data . As exemplified with a compartmentalized model of central metabolism in S . cerevisiae and cell-averaged metabolome data generated under glucose batch conditions , 46% of the 71 . 3 million EFMs were found thermodynamically infeasible , leading to direct insights into reaction directionalities , to constraints in several metabolic rates , and to the identification of reaction patterns that must be inactive due to a thermodynamic infeasibility . This work builds on earlier work that integrated FBA , thermodynamics and quantitative metabolite data [14] , [15] , and extends it by using EFMs , allowing us to identify the reasons underlying the infeasibilities without making a priori assumptions on the metabolic objectives of the cell , such as maximization of biomass production , as is the case with FBA . Additionally , in our study we considered a compartmented metabolic network of S . cerevisiae to analyze cell-averaged metabolite data . Notably , the results we obtain are directly related to compartmentation , as can be seen from the identified infeasibility patterns that involve both compartments . The identified infeasibility of the ethanol-acetaldehyde redox shuttle has been previously identified using NET analysis [12] and manual consideration of the system . In this work we demonstrate that by using the flux patterns that are obtained from the EFMs we systematically identify such infeasibility patterns . Although the system-level constraints and their underlying causes can be rationalized without using EFMs , we need the generated EFMs to determine the infeasible patterns that are part of a stoichiometrically balanced flux distribution . In addition , because the number of EFMs can be considered approximately proportional to the flux solution space , we find that roughly half of the flux solution space is thermodynamically infeasible due to systems of reaction activities . With the recently developed new group contribution method to estimate thermodynamic properties on a genome-scale [35] , the recently increased availability in thermodynamic properties through experimental methods [36] , the advances in quantitative metabolomics [37] and the now available methods to calculate EFMs also for large stoichiometric network [38] , [39] , we envision that the here presented approach will be helpful to shed light on metabolic flux physiology in more complex metabolic system such as higher cells simultaneously growing on multiple carbon substrates , where the applicability of classical flux analysis methods are still rather limited .
We used experimental data on metabolite concentrations for Saccharomyces cerevisiae obtained from four independent experiments with at least two replicates [40]–[42] with equal growth medium but under different cultivation conditions ( bioreactor , shake flask , 96-well ) . Based on the data from the independent experiments , we constructed a consensus data set , where for each metabolite a minimum and maximum concentration was defined . The minimum and maximum concentrations were determined from all the replicates of measurements for each metabolite . To reduce the effect of outliers on the ranges , when more than 3 replicates were available , we removed the values higher than the third quartile +1 . 5 IQR ( inter quartile range ) , and values lower than the first quartile −1 . 5 IQR . Physiological data was obtained for S . cerevisiae on glucose as carbon source from one of the four experiments . In Dataset S2 we describe the details of the experimental conditions of the data sets , the obtained concentration ranges and physiological data . The stoichiometric metabolic network model describes the core central carbon metabolism of S . cerevisiae in the cytosol and mitochondrion with 230 reactions and 218 metabolites ( see Dataset S3 ) , and was developed on the basis of the genome-scale metabolic model iND750 that contains 1149 reactions and 646 metabolites [21] . For our model , we selected the cytosolic and mitochondrial reactions belonging to glycolysis/gluconeogenesis , pentose-phosphate pathway ( PPP ) , TCA cycle , anaplerosis , pyruvate metabolism , and oxidative phosphorylation . The reversibility of each reaction was taken from Duarte et al . [21] . A cytosolic malate synthase was added to complement the glyoxylate shunt in the cytosol [43] . A citrate synthase was added to the cytosol since this is supported by localization studies [44] . To allow the model to synthesize all amino acids that are required for biomass , we added the following pathways: For L-alanine , two biosynthetic routes from pyruvate were included: cytosolic and mitochondrial alanine transaminase reactions , which were assumed to solely produce but not degrade L-alanine [45] , [46] . Furthermore , L-glutamate could be produced via three alternative pathways: cytosolic or mitochondrial NADP-dependent glutamate dehydrogenase from alpha-ketoglutarate or mitochondrial NAD-dependent glutamate synthase from alpha-ketoglutarate and glutamine [47] . For glycine synthesis , we implemented three pathways such that it could be synthesized in the mitochondria via ( i ) alanine-glyoxylate transaminase [48] , or in the cytosol by ( ii ) glycine hydroxymethyltransferase from serine [49] , or ( iii ) from L-threonine via threonine aldolase [50] . As the latter reaction was assumed to be reversible , it could also be used to produce L-threonine , and such it constitutes a second possibility to produce L-threonine next to the linear pathway from L-aspartate . For all other amino acids , the model contains only one linear cytosolic pathway consisting of consecutive enzymatic reaction steps . Here , no alternative paths exist or they are excluded based on biochemical literature as it was done also to construct models for -based flux analysis [45] , [46] . The model further includes transport reactions across the mitochondrial membrane for metabolites that participate in reactions in both the cytosol and the mitochondria . Additional transport reactions that were not contained in iND750 ( i . e . for L-glutamate , alpha-ketoglutarate , homocitrate , glyoxylate , and 2-oxobutanoate ) were added to properly connect additionally included alternative pathways for amino acid synthesis to the metabolic network . The biomass composition was adopted from iND750 besides that trehalose and glycogen were discarded since carbohydrate storage was not considered in our model . Lumped reactions for synthesis of the remaining biomass constituents , i . e . lipids , nucleotides , and cell wall components from the corresponding precursors were determined based on the biomass composition as provided in iND750 . In the model , carbon molecules that can be exchanged with the environment are glucose , glycerol , pyruvate , acetate , ethanol , succinate , and . The model is not proton balanced . The reason for this is that it is close to impossible to do the proton balancing correctly ( e . g . , for transport reactions ) . Thus , we did not want to add any potentially wrong constraints on the model and therefore did not account for proton balancing , with one exception . We only balanced the protons around the respiratory chain by replacing the cytosolic protons produced and consumed in the reactions CYOR_u6m , CYOOm and ATPS3m by a unique species “hcyt” . As a result , ATP generated in ATPS3m can only occur through the respiratory chain . With NET analysis we can determine the feasibility of a flux distribution based on ranges for the concentrations of the involved metabolites . A flux distribution is thermodynamically infeasible when one or more reaction activities conflict with the calculated Gibbs energy of reaction range ( s ) . Conversely , a flux distribution is feasible when no conflicts are found . It is important to note that a metabolite concentration is constrained in NET analysis by any reaction that has the metabolite as a reactant . Therefore , a flux distribution can be infeasible due to propagated constraints in a pathway . The NET analysis implementation constrains metabolite concentrations of metabolites that occur in multiple compartments as a sum of the compartment specific concentrations , corrected for their volume . Therefore , the compartmental distribution of such metabolites is left free . The compartmental volume fractions of the cytosol and mitochondrion are set to 0 . 35 and 0 . 1 , respectively . The NET analysis approach is similar to other thermodynamic analysis approaches [14] , [15] . A main difference from other approaches is that with NET analysis we aim at checking flux distributions for thermodynamic feasibility , and at estimating ranges of Gibbs reaction energies and metabolite concentrations . For NET analysis we used the concentration ranges defined from the experimental data . For all other metabolites in the network we assumed a default range with a minimum concentration of 0 . 0001 mM and a maximum of 120 mM , except for carbon dioxide ( “co2tot” ) , phosphate ( “pi” ) and diphosphate ( “ppi” ) that were constrained to a range of 1 mM to 100 mM [51] , [52] , and oxygen ( “o2” ) that was constrained to 0 . 001 mM to 0 . 1 mM [53] . By using such large metabolite concentration ranges we account for the noise in the metabolite concentration data and uncertainties in Gibbs energies of formation . Typical uncertainties in formation energies are in the order of 0 . 02–2 kJ/mol [35] , which are overshadowed by variations in metabolite concentration data . The compartmental pH values were set to 5 , 6 . 5 and 7 for the external , cytosolic and mitochondrial environment , respectively [54] , [55] . The ionic strength was set to 0 . 15 M for all compartments . For the correct consideration of transport thermodynamics in NET analysis , we defined the specific transported species for each transport reaction where possible , and calculated transport reaction values according to Jol et al . [56] . Computations for FVA , NET analysis and EFM generation were done using MATLAB ( The Mathworks ) . For optimization of FVA problems we used the LINDO API library ( LINDO Systems Inc . ) and for NET analysis we used anNET [23] in combination with the LINDO global solver . To generate EFMs we used the Java implementation from Terzer and Stelling [19] on a quad-core system ( 3 GHz ) with 128 GB memory . To test the thermodynamic feasibility of each EFM we used anNET , which was modified to run in an automated way on a cluster of computers encompassing on average 60 CPUs ( 3 GHz ) . Testing EFMs for thermodynamic feasibility was computationally intensive and took approximately 14 days . To find the reaction activity patterns that cause infeasibility , we considered each infeasible EFM separately and performed an iterative analysis . In NET analysis of the EFM , we removed consecutively each reaction's activity constraint from the NET analysis optimization and determined the feasibility . If the activity pattern became thermodynamically feasible , the reaction activity that was removed was identified as part of the pattern . Then we continued with removing the next reaction activity , while keeping the activities that were identified as part of the pattern . We continued this process for all the reaction activities . This process led to a set of reaction activities , which is a subset of the activities in the analyzed EFM , of which the removal of one activity leads to a feasible system . The set of reaction activities is only infeasible as a whole system , and no single reaction can be marked infeasible by itself . All possible infeasible reaction patterns may not be found when multiple patterns are present in an EFM , because the order of reaction activity removal determines which pattern is found . The obtained infeasible patterns cover all the infeasible EFMs . Flux balance analysis with maximization of biomass production , maximization of ATP production and maximization of the ratio of ATP production over the sum of squared fluxes was performed according to Schütz et al . [28] with the physiological data used for FVA as constraints . The computations were done using MATLAB ( The Mathworks ) using the LINDO API library ( LINDO Systems Inc . ) for optimization .
|
Fluxes in metabolic pathways are a highly informative aspect of an organism's phenotype . The experimental determination of such fluxes is well established and has proven very useful . To address some of the limitations of experimental flux analysis , such as when the cell is divided in multiple compartments , stoichiometric modeling provides a valuable addition . The approach that we take is based on stoichiometric modeling where we consider the thermodynamic feasibility of many different possible routes through the metabolic network of Saccharomyces cerevisiae using experimentally determined metabolite concentrations . We show that next to conclusions on single biochemical reactions in the metabolic network , we obtain system-level insights on thermodynamically infeasible flux patterns . We found that the compartmental concentrations of and NADH are the causes for the system-level infeasibilities . With the current advances in quantitative metabolomics and biochemical thermodynamics , we envision that the presented method will help gaining more insight into complex metabolic systems .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"systems",
"biology",
"model",
"organisms",
"yeast",
"and",
"fungal",
"models",
"biology",
"computational",
"biology",
"microbiology",
"genetics",
"and",
"genomics"
] |
2012
|
System-Level Insights into Yeast Metabolism by Thermodynamic Analysis of Elementary Flux Modes
|
Interferon-γ inducible factor 16 ( IFI16 ) is a multifunctional nuclear protein involved in transcriptional regulation , induction of interferon-β ( IFN-β ) , and activation of the inflammasome response . It interacts with the sugar-phosphate backbone of dsDNA and modulates viral and cellular transcription through largely undetermined mechanisms . IFI16 is a restriction factor for human cytomegalovirus ( HCMV ) and herpes simplex virus ( HSV-1 ) , though the mechanisms of HSV-1 restriction are not yet understood . Here , we show that IFI16 has a profound effect on HSV-1 replication in human foreskin fibroblasts , osteosarcoma cells , and breast epithelial cancer cells . IFI16 knockdown increased HSV-1 yield 6-fold and IFI16 overexpression reduced viral yield by over 5-fold . Importantly , HSV-1 gene expression , including the immediate early proteins , ICP0 and ICP4 , the early proteins , ICP8 and TK , and the late proteins gB and Us11 , was reduced in the presence of IFI16 . Depletion of the inflammasome adaptor protein , ASC , or the IFN-inducing transcription factor , IRF-3 , did not affect viral yield . ChIP studies demonstrated the presence of IFI16 bound to HSV-1 promoters in osteosarcoma ( U2OS ) cells and fibroblasts . Using CRISPR gene editing technology , we generated U2OS cells with permanent deletion of IFI16 protein expression . ChIP analysis of these cells and wild-type ( wt ) U2OS demonstrated increased association of RNA polymerase II , TATA binding protein ( TBP ) and Oct1 transcription factors with viral promoters in the absence of IFI16 at different times post infection . Although IFI16 did not alter the total histone occupancy at viral or cellular promoters , its absence promoted markers of active chromatin and decreased those of repressive chromatin with viral and cellular gene promoters . Collectively , these studies for the first time demonstrate that IFI16 prevents association of important transcriptional activators with wt HSV-1 promoters and suggest potential mechanisms of IFI16 restriction of wt HSV-1 replication and a direct or indirect role for IFI16 in histone modification .
Herpes simplex virus type I ( HSV-1 ) is a ubiquitous and highly contagious virus that establishes a life-long infection in host organisms . It typically enters the host through mucosal epithelia and causes a lytic , productive infection in many cell types , including fibroblast , epithelial , and endothelial cells , during which more than 80 gene products are produced from the nuclear viral genome . After primary infection , HSV-1 spreads to neuronal cells in the trigeminal ganglia where it establishes latent infection , during which only the Latency Associated Transcript ( LAT ) is produced . Periodically , HSV-1 is reactivated from latency and causes recurrent lytic infection at the site of primary infection [1] . HSV-1 typically causes oral lesions but can cause much more severe pathologies , including blindness and fatal encephalitis , due to its infection of corneal cells and the central nervous system [2]–[4] . During lytic infection , HSV-1 genes are transcribed by cellular RNA polymerase II ( RNA pol II ) , assisted by cellular transcription factors , including TATA-binding protein ( TBP ) , in a highly regulated temporal cascade . Transcription from the immediate early ( IE ) gene promoters of HSV-1 begins as soon as the viral genome enters the nucleus and is initiated by the virion tegument-associated protein , VP16 , in conjunction with the cellular transcription factors , Oct1 and HCF . Most IE genes regulate viral and cellular gene expression . The next temporal class of HSV-1 genes , early ( E ) genes , is expressed around 2–8 hours post-infection ( h p . i . ) and is largely involved in DNA replication . Expression of these genes is dependent on the viral IE regulatory proteins , ICP0 and ICP4 , and cellular RNA pol II , TBP , and other transcription factors . The final broad category of HSV-1 gene expression , late ( L ) genes , is further categorized into leaky late ( DNA replication-independent ) and true late ( DNA replication-dependent ) genes . Late genes encode predominantly structural proteins and their expression is also dependent on viral ICP4 , and host RNA pol II , and TBP proteins [1] , [5] , [6] Because of its lifelong infection of human hosts , HSV-1 has necessarily evolved a complex set of interactions with host cell factors to modulate host and viral gene expression and to evade immune detection and responses . In addition to the gene regulation outlined above , HSV-1 inhibits cellular gene expression to better co-opt the gene expression machinery for itself . The virion host shut off ( vhs ) protein inhibits protein synthesis by causing degradation of host and cellular mRNAs [7] , [8] and , conversely , enhancing expression of late viral mRNAs [9] . Similarly , ICP34 . 5 , a late protein , dephosphorylates host eIF2α and inhibits protein synthesis [10] . The virion-associated HSV-1 genomic DNA associates with the histones and nucleosome proteins , leading into its chromatinization and the epigenetic control of viral genes [11]–[16] . Histones can be shifted along DNA or modified , leading to the condensation ( heterochromatin ) or relaxation ( euchromatin ) of chromatin , resulting in the suppression or activation of gene expression , respectively . Many of these functional histone modifications occur on histone H3 , including markers for heterochromatin , including trimethylated H3 lysine 9 ( H3K9me3 ) or lysine 27 ( H3K27me3 ) , and markers for euchromatin , such as trimethylated lysine 4 ( H3K4me3 ) or acetylation of lysines K9 , K14 , or K27 [17] . Several viral gene products , including ICP0 , VP16 , and LAT have been implicated in viral chromatin remodeling during HSV-1 infection [12]–[14] , [16] , [18] . Along with these basic gene expression and replication interactions with host factors , HSV-1 regulates the host immune response by several mechanisms . ICP0 , a transcriptional regulator and E3 ubiquitin ligase , induces the degradation of antiviral factors [19]–[21] and inhibits type 1 IFN expression [20] , [22] , [23] . HSV-1 has also evolved mechanisms to inhibit IFN signaling [24]–[27] . Our recent studies have shown that HSV-1 activates and then represses the inflammasome response [19] . IFN-γ inducible protein 16 ( IFI16 ) is a DNA binding protein [28] , [29] that was first described as an IFN-inducible protein involved in the differentiation of human myeloid cells [30] , [31] . It is a transcriptional modulator [32] through mechanisms that are not yet fully defined . IFI16 interacts with p53 [33] and regulates p53 target gene expression [34] . Interestingly , both loss and gain of IFI16 induces p53 checkpoints; overexpression leads to the apoptotic p53 checkpoint and loss of IFI16 leads to cell cycle arrest [35] , [36] . IFI16 is acetylated by the histone acetyltransferase , p300 [37] , and may play a role in regulating gene expression by modulating chromatinization [38] . Several immunomodulatory roles have been described for IFI16 . It recognizes nuclear herpesviral genomes , including those of HSV-1 , Kaposi's sarcoma-associated herpesvirus ( KSHV ) , and Epstein-Barr virus ( EBV ) , and responds to human immunodeficiency virus ( HIV ) infection , leading to association with apoptosis-associated speck-like protein containing a caspase activation and recruitment ( CARD ) domain ( ASC ) through its CARD domain to induce inflammasome activity , resulting in the maturation of caspase-1 and IL-1β [19] , [39]–[42] . IFI16 is also necessary for HSV-1- , human cytomegalovirus ( HCMV ) - , and Vaccinia-virus-induced STING-mediated IFNβ expression [20] , [28] , [43] . Interestingly , HSV-1 induces the specific degradation of IFI16 at late times post-infection ( after 4 h ) , dependent , at least in part , on ICP0 [19] , [20] , [44] . Recently , IFI16 has been described as a restriction factor for herpesviral lytic replication [38] , [44] , [45] . It restricts HCMV replication by displacing transcription factors on E and L but not IE gene promoters [45] and restricts HSV-1 replication [38] , [44] , [45] , particularly replication of ICP0-deficient HSV-1 [38] , [44] . IFI16 promotes association of repressive histone modifications with the ICP4 , ICP27 , and ICP8 HSV-1 promoters during infection with this mutant virus that lacks ICP0 but had no apparent effect on histone modifications associated with viral promoters during infection during infection with a rescue virus [38] . In addition , viral gene expression was repressed , somewhat , during infection with this ICP0-null virus in IFI16-positive cells compared with that in IFI16-depleted cells , but viral gene expression of ICP0-competent rescue HSV-1 was not affected [38] . Though this study suggested that IFI16 recognizes and regulates unchromatinized DNA [38] , observations such as the recognition of chromatinized KSHV and EBV DNA by IFI16 during latency , persistence of latent KSHV and EBV gene expression in the presence of IFI16 and the IFI16-inflammasome [39] , [40] , [42] and the repressive effect of IFI16 on wild-type ( wt ) HSV-1 replication observed previously [45] or in this study suggest that the mechanisms of IFI16 restriction of HSV-1 are complex . The study showing repression of HCMV replication by IFI16 demonstrated IFI16-mediated repression of HSV-1 replication but did not pursue mechanisms , thereof [45] . There are multiple differences in cell tropism , replication kinetics , and immune evasion strategies evolved by HSV-1 and HCMV [1] , [37] . Notably , HSV-1 causes the specific degradation of IFI16 while HCMV does not [19] , [20] , [37] , [44] . Therefore , we speculated that the mechanisms of HSV-1 restriction by IFI16 may be distinct from those of HCMV restriction . Here , we show that IFI16 restricts wt HSV-1 replication and gene expression in multiple cell types . It binds to HSV-1 transcription start sites ( TSS ) of all temporal classes of HSV-1 gene expression , and prevents association of transcription factors , including RNA pol II , TBP , and Oct1 with viral promoters but not cellular promoters . We also show for the first time that IFI16 induces increased euchromatin markers and decreased heterochromatin markers associated with both wt HSV-1 and cellular DNA . These data suggest that IFI16 plays a multi-level role in the modulation of HSV-1 gene repression and that development of drugs to stabilize the function of IFI16 may potentially lead into an effective anti-HSV-1 treatment .
Gariano et al . showed that IFI16 depletion in wt HSV-1-infected cells significantly increased viral yield [45] . However , other studies showed that depletion of IFI16 increased replication of an ICP0-null virus but had no effect on the ICP0-rescue virus or wt HSV-1 strain 17 [38] , [44] . To confirm that IFI16 restricts wt HSV-1 ( KOS strain ) replication , we depleted IFI16 from HFF cells using microporated siIFI16 , which resulted in ∼91% knockdown of IFI16 compared with that of siControl ( siCtrl ) RNA ( Figure 1A ) . To determine the effect of IFI16 depletion , 48 h after microporation we infected the cells with HSV-1 at a multiplicity of infection ( moi ) of 0 . 1 or 1 . 0 plaque forming units ( pfu ) per cell . At 24 hours post infection ( h p . i . ) , cell culture supernatant was collected and viral yield was determined by plaque assay . Depletion of IFI16 resulted in a significant 5- or 6-fold increase in viral yield ( Figure 1B ) , after infection at an moi of 0 . 1 or 1 . 0 pfu/cell , respectively . This inhibition is consistent with that observed previously for wt HSV-1 , but further suggests a different mechanism of HSV-1 repression by IFI16 than that involved in repression of HCMV , which was moi-dependent [45] . Interestingly , this reduction in ICP0-positive HSV-1 yield observed here and previously [45] is inconsistent with other reports , showing IFI16-induced inhibition of an ICP0-null virus but not the ICP0-positive rescue virus or wt strain 17 virus [38] , [44] . To determine the effect of IFI16 depletion on HSV-1 gene expression , we infected HFF cells microporated with siCtrl or siIFI16 with HSV-1 at an moi of 1 pfu/cell for 2 , 4 , or 8 h and determined the relative expression of IE ( ICP0 and ICP4 ) , E ( ICP8 and TK ) and L ( gB and Us11 ) gene mRNA by qRT-PCR and normalized data to GAPDH in each sample and to siCtrl at 2 h p . i . ( Figure 1C ) or to concurrent siCtrl samples ( Figure 1D ) , using the ddCt method [46] . Expression of each gene increased over the course of infection; however , the increase was significantly further augmented in the absence of IFI16 ( Figure 1C ) . Compared to the expression in siCtrl-transfected cells , IFI16 depletion led to an increase of expression of all gene classes: at 8 h p . i . , ICP0 expression increased 1 . 5-fold , ICP4 expression was increased by about 24-fold , ICP8 increased by 8-fold , TK increased by 12-fold , gB increased 150-fold , and Us11 increased by 12-fold ( Figure 1D ) . The role of innate immune proteins can be cell-type specific [47] . To determine if our observed effects of IFI16 on HSV-1 replication and gene expression , shown in figure 1 for HFF cells , are also cell-type specific and if over-expression of IFI16 restricts HSV-1 replication and gene expression , we transduced human osteosarcoma U2OS cells with IFI16 , which was expressed 2 . 25-fold over GFP-transduced cells ( Figure 2A ) . HSV-1 yield , measured as in figure 1 , was significantly inhibited by >5-fold in U2OS cells overexpressing IFI16 ( Figure 2B ) . HSV-1 gene expression was determined and normalized as above . Consistent with the IFI16 knockdown experiments ( Figure 1 ) , though expression of each gene increased over time regardless of the presence of overexpressed IFI16 ( Figure 2C ) , IFI16 overexpression significantly inhibited expression of HSV-1 genes from all gene classes over the course of 8 h of infection: compared to GFP-transduced cells , in IFI16-transduced cells at 8 h p . i . , ICP0 was inhibited by >50% , ICP4 was inhibited by 76% , ICP8 was inhibited by 50% , TK was inhibited by 87% , gB was reduced by 74% , and Us11 was inhibited by 77% ( Figure 2D ) . Human breast epithelial cancer MCF-7 cells are naturally IFI16-deficient ( Figure 3A ) [48] . To determine the effect of IFI16 expression on HSV-1 replication and gene expression in these IFI16-negative cells , we transduced MCF-7 cells with lentiviruses expressing IFI16 or GFP ( Figure 3A ) . HSV-1 replication was inhibited ∼1 . 6-fold by IFI16 overexpression in MCF-7 cells ( Figure 3B ) . Similar to the results observed in U2OS cells , though expression of each HSV-1 gene increased over the course of infection ( Figure 3C ) , viral gene expression was relatively diminished in IFI16-expressing MCF-7 cells over the course of 8 h of infection . In comparison to expression in GFP-transduced , IFI16-negative MCF-7 cells , in IFI16-transduced cells at 8 h p . i . , ICP0 , ICP4 , ICP8 , TK , gB , and Us11 were inhibited by 50% , 70% , 62% , 87% , 46% , and 65% , respectively ( Figure 3D ) . Together , the results shown in figures 1 , 2 , and 3 confirmed that IFI16 restricts HSV-1 replication , as was first published by Gariano et al . in HFF cells [45] , and suggested that the mechanism of restriction is at the level of HSV-1 gene expression and is not cell-type dependent . Recently , we showed that HSV-1 infection induces the IFI16 and NLRP3 inflammasome , causing IFI16 and NLRP3 to associate with the inflammasome adaptor protein , ASC , and resulting in the maturation of caspase-1 and IL-1β [19] . IFI16 also has a role in the expression of HSV-1-induced type I IFN [20] . To determine if the inflammasome or IFN response play roles in the innate inhibition of HSV-1 replication and gene expression , we transduced HFF cells with shRNA-expressing lentiviruses targeting IFI16 , ASC , or IRF3 resulting in 92% , 81% , and 97% knockdown of IFI16 , ASC , and IRF3 , respectively ( Figure 4A , lanes 2 , 3 , and 4 , respectively , compared with lane 1 ) . None of the knockdown conditions tested affected STING levels ( Figure 4A ) , showing knockdown specificity . Knockdown of ASC did not affect HSV-1 viral yield , as measured by plaque assay ( Figure 4B ) . Consistent with previous reports showing no effect on wt- or ICP0-null HSV-1 replication in the absence of IRF3 [49] , knockdown of IRF3 also did not affect HSV-1 yield ( Figure 4B ) . In contrast , consistent with our previous results , knockdown of IFI16 resulted in a significant >5-fold increase of HSV-1 yield ( Figure 4B ) . To further confirm functional knockout of IFI16 and IRF3 , we assayed HSV-1-infected HFF cell culture supernatant IFNβ levels at 6 h p . i . , infected 48 h post-transduction at an moi of 1 pfu/cell . A robust IFNβ response was detected from HSV-1-infected cells that had been transduced with shCtrl or shASC but not from cells transduced with shIFI16 or shIRF3 ( Figure 4C ) . To confirm functional knockout of ASC , we determined procaspase-1 cleavage in shRNA-transduced HFF cells infected with HSV-1 at an moi of 1 pfu/cell at 6 h p . i . , 48 h post transduction . Procaspase-1 was cleaved to active caspase-1 in HSV-1-infected cells transduced with shCtrl , shIRF3 , and , to a lesser extent , with shIFI16 , likely due to the HSV-1-induced activation of the nucleotide binding and oligomerization ( NOD ) -like receptor family pyrin domain-containing 3 ( NLRP3 ) inflammasome [19] , but not in mock infected cells or in HSV-1-infected cells transduced with shASC ( Figure 4D , compare lanes 1 and 3 with lanes 2 , 4 , and 5 ) . These studies suggested that the role of IFI16 in the inhibition of HSV-1 viral replication is independent of its role in HSV-1-induced inflammasome activation and interferon induction . Transfection and transduction of cells leads to activation of innate immune responses and adversely effects HSV-1 replication [50]–[52] . In addition , exogenous DNA introduced into cells activates the absent in melanoma 2 ( AIM2 ) and/or NLRP inflammasome responses [19] , [39] , [40] , [42] , [53] . To eliminate potential artifacts from these effects and to more thoroughly investigate the effects of IFI16 on HSV-1 replication and gene expression , we used Clustered Regularly Interspaced Short Palindromic Repeat ( CRISPR ) Cas9-mediated genome editing [54] , a highly specific method for targeted eukaryotic genome editing [55]–[57] , to create a permanent IFI16-negative cell line . We designed guided RNA to target the Cas9 endonuclease to a region within the coding sequence for the IFI16 Pyrin domain ( PYD ) , the first functional domain of IFI16 ( Figure 5A ) . Wt U2OS cells were cotransfected with 3 plasmids encoding the guided RNA , Cas9 , and GFP ( a marker for transfection ) at a ratio of 4∶1∶1 , respectively . After 48 h , cells were sorted for GFP expression and grown clonally before screening for IFI16 expression by dot blot and western blot ( Figure 5B ) . The IFI16-negative U2OS clones 45 and 67 ( Figure 5B , lanes 1 and 4 , compared with lane 5 ) were further characterized . IFI16-negative U2OS cell growth was moderately slower than that of the wt U2OS parental cells; doubling time for wt cells was approximately 31 . 8 h and those of clones 45 and 67 were approximately 40 h ( Figure 5C ) . In addition , the deletion of IFI16 in U2OS cells caused a significant change in cellular morphology , leading to rounded , elongated cells when compared with the parental wt cells ( Figure 5D ) . We selected clone 67-U2OS for further experiments . To ensure that the stimulatory effects of IFI16 depletion on HSV-1 replication occurred in these newly generated IFI16-negative U2OS cells , we assayed viral yield at 24 h p . i . from wt U2OS , IFI16-negative clone 67 , and clone 67 cells transduced with IFI16 to rescue the effects of IFI16 in IFI16-negative cells . When compared with that in parental wt U2OS cells , HSV-1 yield from clone 67 cells was increased over 6-fold in IFI16-negative clone 67 U2OS cells ( Figure 6A ) . Transduction of IFI16 into clone 67 cells resulted in HSV-1 yield 30% lower than that from wt U2OS cells ( Figure 6A ) , consistent with higher levels of IFI16 expression in these cells ( Figure 6B ) , confirming that IFI16 is the factor responsible for HSV-1 restriction and further suggesting specificity of the Cas9-mediated IFI16 deletion in clone 67 cells . HSV-1 gene expression at 2 , 4 , and 8 h p . i . was also determined in these cells . Consistent with our experiments with transient knockdown of IFI16 , permanent IFI16 depletion led to increased HSV-1 gene expression from all temporal classes of genes over the course of 8 h of infection , though expression of all the genes increased over the course of infection in both conditions ( Figure 6C ) . Compared with that in IFI16-positive parental U2OS cells , in clone 67 cells at 8 h p . i . , ICP0 was increased 3-fold , ICP4 was increased 7-fold , ICP8 was increased 5-fold , TK was increased 3-fold , gB was increased 5-fold , and Us11 was increased 4-fold ( Figure 6D ) . In addition , we examined IFI16 protein levels in wt U2OS cells and clone 67-U20S cells transduced with empty vector or with IFI16 expression vector in the absence of HSV-1 infection or at 24 h p . i . ( Figure 6B ) . Though previous results showed stability of IFI16 in HSV-1-infected U2OS cells up to 12 h [44] , IFI16 protein was decreased 80% in wt-U2OS cells ( Figure 6B , lane 4 compared with lane 1 ) and 87% in IFI16-transduced clone 67 cells ( Figure 6B , lane 6 compared with lane 3 ) after a 24 h infection with HSV-1 at an moi of 1 pfu/cell . To ensure IFI16 stability in HSV-1-infected U2OS cells over the course of our gene expression analysis , we performed western blot analysis of HSV-1-infected U2OS and clone 67 cells . IFI16 was stable in U2OS cells up to 8 h p . i . and was , in fact , induced between 2–4 hours ( Figure 6E , lane 3 compared with lanes 1–3 ) , consistent with the induction seen in HSV-1-infected HFF cells [19] . ICP0 protein expression , as measured by western blot ( Figure 6E , top ) and quantified by densitometry ( Figure 6E , bottom ) , was increased in clone 67 cells when compared with that in wt U2OS cells ( Figure 6E , compare lanes 7 and 8 with lanes 3 and 4 ) , consistent with our mRNA results ( Figure 6C ) . Together , these results demonstrated that our permanent IFI16-negative cell line is an appropriate tool to further study the inhibition of HSV-1 replication and gene expression by IFI16 . Previously , we showed by FISH analysis and co-immunofluorescence that IFI16 colocalized with the HSV-1 genome at 1 h p . i . in HFF cells [19] . To determine HSV-1 genome recognition by IFI16 and colocalization in U2OS cells , we infected U2OS cells with EdU-labeled HSV-1 for 30 or 60 min , performed immunofluorescence for IFI16 , and costained for EdU-labeled HSV-1 genome ( Figure 7A ) . We also performed proximity ligation assays ( PLA ) , a fluorescence-based assay that uses DNA-oligonucleotide-linked secondary antibodies to detect closely associated proteins . If two protein epitopes are within 40 nm of each other , the antibody-linked oligonucleotides can ligate with adaptor oligonucleotides to form complete circles that are replicated via rolling-circle DNA replication and detected with fluorescent sequence-specific probes . PLA provides a method for the detection of weak or transient interactions [58] . It can also provide amplified , very distinct localization of a single protein . Here , we did PLA with two antibodies to IFI16 , and costained for EdU-labeled HSV-1 ( Figure 7B ) . In uninfected cells , IFI16 was exclusively nuclear . By 30 min p . i . , EdU-HSV-1 was detected in approximately 30% of cell nuclei and colocalized with IFI16 in small nuclear puncta ( Figure 7A and B , enlarged images of B , yellow arrows ) . By 60 min , consistent with previous reports from our laboratory and others [19] , [37] , some IFI16 was detected in the cytoplasm ( Figure 7A and B , white arrows ) , and nuclear IFI16 largely colocalized with the HSV-1 genome ( Figure 7A and B , yellow arrows ) . Pixel intensity plots for the red and green channels were generated for each of the PLA figures ( Figure 7B , bottom panels ) . These show quite clearly the yellow signals that indicate colocalization between IFI16 ( green ) and HSV-1 genomic DNA ( red ) . These data are the first to distinctly show IFI16 colocalization with the HSV-1 genome at such early times post-infection in U2OS cells and suggest that the association of IFI16 with HSV-1 occurs very shortly after viral DNA enters the nucleus . HSV-1 gene expression occurs in the nuclei of host cells [1] . To determine if the different morphologies of HFF , U2OS , and clone 67 U2OS cells affected the time post-infection that HSV-1 genomic DNA enters host cell nuclei , we mock infected or infected HFF ( Figure S1A ) , U2OS ( Figure S1B ) , or clone 67-U2OS ( Figure S1C ) cells with HSV-1 genome-labeled with 5-bromo-2-deoxyuridine ( BrdU ) at an moi of 1 pfu/cell for 15 , 30 , or 90 min and immunostained for BrdU . Cytoplasmic BrdU staining showed punctate spots , likely representing nucleocapsid-bound HSV-1 genomes [1] . BrdU-HSV-1 genomes first appeared in the nuclei of approximately 5% of all three cell types at 15 min p . i . ( Figure 8A ) . Once in the nucleus , BrdU staining was still somewhat punctate in many cells but was , overall , much more diffuse than cytoplasmic BrdU staining , likely reflecting the relative expansion of nuclear HSV-1 DNA compared with that condensed in nucleocapsids , consistent with previous studies [59] , [60] . The percent of cells with nuclear BrdU increased to approximately 25% at 30 min and to 80% at 90 min p . i . These results demonstrated that the kinetics by which the HSV-1 genome enters host cell nuclei is consistent between HFF , U2OS , and clone 67-U2OS cells in that the genome begins to enter host cell nuclei within 15 min of exposure to the virus , and nuclear HSV-1 DNA increased steadily in a consistently measurable proportion of cells at 30 min p . i . These data suggest that the kinetics of HSV-1 genome entry into cell nuclei is not affected by the morphology of these cell types . Because the number of nuclei with HSV-1 changed over time , we more quantitatively determined the relative nuclear HSV-1 levels by qPCR of the ICP4 promoter region in nuclear extracts of HSV-1-infected U2OS and clone 67 cells from 30 min to 4 h p . i . , and normalized to GAPDH and to DNA levels in U2OS cells at 30 min p . i . using the ddCt method ( Figure 8B ) . We observed that there was variation in both U2OS and clone 67-U2OS cells over time , consistent with previous studies [13] , but no difference between cell types at any time ( Figure 8B ) . Based on these results , we used 30 min p . i . as the earliest time point in all our subsequent HSV-1 gene expression studies . IFI16 binds the sugar-phosphate backbone of DNA in a purportedly sequence-independent manner in vitro [61] and has been shown to interact with oligonucleotides derived from HSV-1 [28] , [37] and colocalize with HSV-1 in infected cell nuclei ( Figure 7 ) [19] . There is some preference for cruciform or super helical DNA in vitro [29] , but otherwise , little is known about the location of IFI16 binding DNA sites on viral or cellular genomes and no studies have shown , by chromatin immunoprecipitation ( ChIP ) , association of endogenous IFI16 with HSV-1 genomic DNA during infection . Our results in Figures 1 , 2 , 3 , 4 , and 6 showed that IFI16 reduced HSV-1 gene expression . We therefore chose to examine its presence at the transcriptional start sites ( TSS ) of HSV-1 genes . To determine whether IFI16 binds HSV-1 gene TSS , we performed ChIP analysis of HSV-1 infected cells . IFI16 was immunoprecipitated from the nuclei of HFF and wt-U2OS cells infected with HSV-1 ( 1 pfu/cell ) for 30 min , 1 , 2 , or 4 h . DNA associated with IFI16 was analyzed by qPCR with primers corresponding to regions flanking the TSS of HSV-1 ICP0 , ICP4 , ICP8 , gB , and Us11 genes , as well as the cellular GAPDH and p21 genes . Because levels of nuclear HSV-1 DNA change up to 2-fold over the course of infection , we normalized our ChIP data concerning HSV-1 DNA to input HSV-1 ( ICP4 promoter region ) to avoid artifacts from the different sample inputs . Cellular gene DNA was normalized to input GAPDH . For each promoter tested data were further normalized to the 30 min time point using the ddCt method [46] . IFI16 antibodies coprecipitated each of the viral TSS tested in both HFF and wt-U2OS cells ( Figure 8C ) . The association of IFI16 with these regions increased over the course of the infection for nearly all of the genes in both cell types , most dramatically between 30 min and 1 h p . i . Interestingly , in HFF cells , IFI16 accumulated most heavily onto the Us11 promoter whereas it accumulated least on the Us11 promoter in U2OS cells ( Figure 8C ) . In an earlier study , IFI16 was not found to be bound to cellular promoters in HCMV infected HFF cells [43] . However , we found IFI16 bound to the GAPDH and p21 promoters in HFF and U2OS cells that were mock infected or infected with HSV-1 ( Figure 8D and E ) . This association did not significantly change between mock infected or HSV-1 infected cells or over the course of the 4-hour HSV-1 infection ( Figure 8D and E ) . To ensure specificity of our ChIP assays , we performed additional ChIP analysis of HSV-1-infected U2OS ( U ) or clone 67 U2OS ( 67 ) cells with IFI16 antibody or control IgG . Promoter sequences were amplified and analyzed by agarose gel electrophoresis . IFI16 antibodies , but not control IgG , precipitated viral and cellular promoters from wt U2OS cells but not IFI16-negative clone 67 U2OS cells ( Figure 8F , compare lane 1 with lanes 2–4 ) . Importantly , there was no amplification of DNA from these regions in no-antibody controls at 4 h p . i . ( Figure 8G ) . These data are the first to show endogenous IFI16 binding to HSV-1 DNA during infection and suggested that IFI16 targets all temporal classes of HSV-1 genes and does not specifically affect the immediate early and early regulatory genes . HSV-1 genes are transcribed by cellular RNA pol II [1] . Our data show that IFI16 binds HSV-1 DNA ( Figure 8 ) . To determine whether the presence of IFI16 affects the accumulation of RNA pol II on the various HSV-1 TSS , we infected wt and clone 67 U2OS cells with HSV-1 at an moi of 1 pfu/cell for 30 min , 1 , 2 , or 4 h and performed ChIP analysis with an antibody to RNA pol II and primers to HSV-1 TSS , as described above , normalizing once again to input ICP4 levels because of the differences observed in nuclear HSV-1 DNA levels ( Figure 8B ) . RNA pol II accumulation at HSV-1 TSS was increased significantly in IFI16-negative cells for all viral genes tested compared with that in parental U2OS cells ( Figure 9A ) . Interestingly , RNA pol II accumulation at the cellular GAPDH gene promoter was not affected by the presence or absence of IFI16 ( Figure 9A ) . To confirm that RNA pol II association was unchanged on cellular promoters in the presence or absence of IFI16 , we repeated the RNA pol II ChIP , in U2OS and clone 67 cells that were infected with HSV-1 for 30 min or 2 h , with primers for the GAPDH and p21 promoter regions . RNA pol II association with each of these regions was the same , regardless of infection time or the presence of IFI16 ( Figure 9B ) . HSV-1 and GAPDH amplicons were detected in samples from U2OS and clone 67 cells immunoprecipitated with RNA pol II but not with beads alone at 4 h p . i . ( Figure 9C ) . These results suggested that the presence of IFI16 restricts the association of RNA pol II with HSV-1 TSS of all temporal classes , consistent with IFI16-mediated HSV-1 gene repression ( Figures 1–3 ) . HSV-1 gene expression is dependent on many viral and cellular transcription factors [1] , [59] , [62]–[64] . Some of these factors , including the TATA binding protein ( TBP ) and Oct1 are necessary to recruit RNA pol II to all HSV-1 promoters , in the case of TBP , and only to immediate early promoters , in the case of Oct1 [65] , [66] . Because IFI16 has been shown to prevent the association of some transcription factors with the HCMV promoter [45] , we hypothesized that IFI16 prevents association of TBP and Oct1 with HSV-1 promoters , thereby inhibiting the recruitment of RNA pol II for gene expression that we observed ( Figure 9 ) . To determine whether IFI16 reduces the association of TBP with HSV-1 promoters , we infected wt and clone 67 U2OS cells with HSV-1 at an moi of 1 pfu/cell for 30 min , 1 , 2 , or 4 h , performed ChIP analysis with antibodies to TBP and primers to HSV-1 TSS ( Figure 10 ) , and normalized as described above . The absence of IFI16 expression led to significantly increased ( 3- to 10-fold ) association of TBP with HSV-1 promoters but no change in association of TBP with the GAPDH promoter ( Figure 10A ) . To confirm that TBP association with cellular promoters was unchanged by the presence of IFI16 , we repeated the TBP ChIP in both cell types infected with HSV-1 for 30 min or 2 h , and amplified the GAPDH and p21 TSS . TBP association with each of these regions was unchanged , regardless of infection time or the presence of IFI16 ( Figure 10B ) . DNA was amplified only from samples from both cell types immunoprecipitated with antibodies to TBP but not from beads alone controls ( Figure 10C ) . To determine if IFI16 reduces association of Oct1 with HSV-1 promoters , we infected wt and clone 67 U2OS cells with HSV-1 at an moi of 1 pfu/cell for 30 min , 1 , 2 , or 4 h and performed ChIP analysis with antibodies to Oct1 and primers to HSV-1 TSS as described above . The absence of IFI16 expression led to significantly increased ( 3- to 17-fold ) association of Oct1 with HSV-1 immediate early ICP0 and ICP4 promoters , a very modest increase in association of Oct1 with ICP8 and Us11 promoters , and no significant change in the association of Oct1 with gB or GAPDH promoters ( Figure 11A ) . To confirm that Oct1 association with cellular promoters was not altered by the presence or absence of IFI16 , we repeated the Oct1 ChIP in both cell types infected with HSV-1 for 30 min or 2 h , with primers for GAPDH and p21 promoters . Oct1 association with each of these regions was the same , regardless of infection time or the presence of IFI16 ( Figure 11B ) . Amplified DNA was observed only from samples immunoprecipitated with antibodies to TBP but not in the control beads alone samples at 4 h p . i . ( Figure 11C ) . Together , these results demonstrated that IFI16 does , indeed , prevent and/or reduce association of transcription factors with HSV-1 promoters but not cellular promoters and provide a potential explanation for the decrease in RNA pol II association with HSV-1 promoters in IFI16 positive cells ( Figure 9 ) . Orzalli et al . [38] suggested that IFI16 may induce changes in histone modifications associated with HSV-1 promoters in HFF cells at 6 h p . i . during infection with an ICP0-null virus , but saw no effect on histone modification with their rescue virus . Because we observed IFI16-induced differences in HSV-1 gene expression and replication during infection with wt virus , we hypothesized that there were IFI16-induced differences in the histone modifications associated with wt HSV-1 DNA . To determine if IFI16 affected the histone modifications in chromatin associated with HSV-1 TSS , we infected wt and IFI16-negative clone 67 U2OS cells with HSV-1 at an moi of 1 pfu/cell for 30 min , 1 , 2 , or 4 h and performed ChIP analysis with antibodies to total histone H3 , H3K4me3 ( a marker for active or euchromatin ) , or H3K9me3 ( a marker for repressive or heterochromatin ) , primers for HSV-1 TSS , and normalized to input ICP4 TSS levels ( for HSV-1 TSS ) or GAPDH ( for cellular TSS ) , as described above . As determined previously [38] , the presence of IFI16 had no effect on the levels of total histone H3 associated with HSV-1 promoters or the cellular GAPDH or p21 promoters over the course of infection ( Figure 12A ) . In the presence of IFI16 , H3K4me3 ( active chromatin ) was detected associated with IE , E , and L HSV-1 promoters at 30 min p . i . This association decreased 3- to 5-fold by 4 h p . i . ( Figure 12B ) . At 30 min p . i . , there was no effect of IFI16 depletion on H3K4me3 association with viral promoters ( Figure 12B ) . In contrast , IFI16 depletion led to a marked increase in the association of the active H3K4me3 with HSV-1 promoters at later times p . i . ( Figure 12B ) . Interestingly however , the absence of IFI16 also led to an increase in the levels of active H3K4me3 on the GAPDH and p21 promoters ( Figure 12B ) . IFI16 also had little to no effect on the association of H3K9me3 ( repressive chromatin ) with IE , E , and L HSV-1 promoters at the very early times post infection but at later times , the presence of IFI16 led to increased association of the repressive chromatin marker with HSV-1 promoters ( Figure 12C ) . In the absence of IFI16 , HSV-1 promoter occupancy by H3K9me9 did not change significantly over the course of the 4 hour infection . Again , interestingly , the association of H3K9Me3 with the cellular GAPDH and p21 promoters was increased in the presence of IFI16 at all times p . i . ( Figure 12C ) . These data corroborate the previous findings that IFI16 does not affect the association of total histones with HSV-1 DNA [38] but demonstrated that there is a global decrease in repressive heterochromatin markers and an increase in active euchromatin markers associated with wt HSV-1 and cellular promoters in the absence of IFI16 , which was not observed in previous studies .
Though IFI16 is known to be involved in transcriptional regulation [32]–[34] , [48] , [70]–[76] , the mechanisms of this regulation have been largely undetermined . IFI16 may modulate transcription through association with transcription factors and/or blocking their association with promoters [33] , [34] , [45] , [48] , [70] , [72]–[74] or it may promote the formation of heterochromatin on promoters , resulting in their repression [38] , [76] . Here , we show evidence for both effects during HSV-1 infection; IFI16 reduces association of RNA pol II , TBP , and Oct1 specifically with HSV-1 promoters and not with the promoters of the cellular genes , GAPDH and p21 ( Figures 9–11 ) . This suggests that IFI16 can specifically prevent association of transcription factors with HSV-1 DNA while allowing them to associate with cellular genomic DNA . We also show evidence that IFI16 directly or indirectly impacts histone modifications at the TSS of both HSV-1 and cellular genes ( Figure 12 ) . Interestingly , there was an increase in total histone association with only the TSS of ICP4 ( Figure 12A ) . This could be due to the necessity for high levels of ICP4 early during infection to stimulate E and L gene expression and the relatively little ICP4 needed at later times . Additionally , at 30 minutes p . i . , there was little difference in histone modifications between cell types , but the increase in H3K4me3 and decrease in H3K9me3 at HSV-1 TSS in IFI16-negative cells compared with that in IFI16-positive cells suggests that these modifications are being added to virally-associated histones largely between 30 min and 1 h p . i . rather than histones being modified prior to association with viral DNA . This alteration of the histone modifications at extremely early times post-infection could be due to ICP0 , VP16 , IFI16 , or other viral or cellular factors , or a combination . Previous studies have suggested that IFI16 promotes the formation of heterochromatin and reduces the formation of euchromatin on the promoters of an HSV-1 mutant genome , which does not express the viral E3 ubiquitin ligase , ICP0 [38] . However , that study did not show a change in chromatin markers associated with their wt-like rescue virus and did not examine the association of these markers with cellular genes [38] . The difference between our findings and the earlier study is likely due to their normalization of data to GAPDH promoters associated with specific chromatin markers and our normalization to input viral genomes . Normalizing our data to co-precipitated GAPDH sequences would also yield results suggesting no difference between chromatin markers on wt HSV-1 genomes because we found that association of these histone modifications with cellular DNA was also altered in the presence of IFI16 ( Figure 12 ) . We postulate that IFI16 may have a global effect on the formation of hetero- and euchromatin and not an effect strictly on viral genome-associated chromatin . This is consistent with previous studies linking IFI16 with histone modifications and histone modification machinery [37] , [76] and suggests that IFI16 may play a direct or indirect role in modulating the activity of these enzymes or their association with HSV-1 genes . The role of ICP0 in chromatin remodeling [13] and possible nuclear interaction between IFI16 and ICP0 [19] further suggest a role for IFI16 in histone modification . Further studies are required to fully understand the involvement of IFI16 in chromatin remodeling complexes in the basal state or during infection with HSV-1 or other pathogens . We found IFI16 associated with GAPDH promoters in HFF cells and U2OS cells ( Figure 8 ) . However , previously Li et al . showed that , during HCMV infection , IFI16 associated with viral genes but did not associate with cellular genes [43] . This discrepancy could be due to differences in ChIP protocol; we immunoprecipitated IFI16 from cell nuclear extracts , which reduces background . Li et al . , immunoprecipitated from total cell lysate [43] , which would lack such nuclear enrichment and , because IFI16 is exported from the nuclei to the cytoplasm of herpesvirus-infected cells [19] , [37] , [39] , [40] , [42] , [77] , immunoprecipitating from total cell lysate could capture ligands associated with cytoplasmic IFI16 , possibly reducing the threshold of detection for nuclear associations . However , given that Li et al . , show exclusively nuclear IFI16 at the time of their IFI16 ChIP experiments , differences in gene detection by ChIP may be based on the proximity of the primer-amplified DNA region with the binding site of IFI16 . It is also remarkable that IFI16 was found associated with promoters from each temporal class of HSV-1 expression ( Figure 8 ) . This suggests that the inhibition of HSV-1 by IFI16-mediated transcriptional repression is not the result of IFI16 associating with only IE promoters causing the inhibition of expression of downstream classes of genes by inhibiting their viral activators . This provides IFI16-expressing cells with redundant means of HSV-1 gene expression inhibition . However , the binding of IFI16 to each temporal class of HSV-1 gene does not exclude the possibility that it prevents other stages of HSV-1 replication . IFI16 is involved in STING-mediated type 1 IFN induction [20] , [28] , [67] , STING is associated with the ER and trans-Golgi network [78] , and HSV-1 nucleocapsids bud through the Golgi during egress [1] . IFI16 may inhibit budding of progeny HSV-1 virions . Further studies are required to investigate this possibility and also to determine the extent and localization of IFI16 binding sites to HSV-1 DNA . We show that IFI16 prevents the association of important transcription factors with HSV-1 gene promoters ( Figures 9–11 ) . However , HSV-1 gene expression is not completely abrogated in the presence of IFI16 , as shown here , and by the permissiveness of most cell types to infection with HSV-1 [1] . Therefore , the effect of IFI16 on HSV-1 gene repression cannot be absolute . This could be due to the stoichiometry of the interaction between IFI16 and HSV-1 promoters: there may not be sufficient IFI16 to promote IFN and inflammasome induction and concurrently be present on all HSV-1 promoters . The induction of IFI16 expression at early times p . i . in HFF cells [19] and U2OS cells ( Figure 6 ) could be a cellular response to simultaneously promote all of these activities . In addition , the HSV-1-induced degradation of IFI16 occurs with significantly slower kinetics than that of another nuclear foreign DNA sensor and viral restriction factor , PML [44] . This suggests that IFI16 has a nuanced role in the regulation of HSV-1 genes and may be useful for HSV-1 at early times p . i . Perhaps its viral gene repression activity facilitates HSV-1 replication , in vivo , by preventing uncontrolled viral replication and undue stress on host cells and the infection microenvironment , which may expedite immune cell recruitment . It is also possible that IFI16 may act to promote the expression of some viral genes , as it is a positive regulator of some cellular genes [48] . IFI16 is involved in cell-cycle regulation and the DNA damage response [48] , [79] , [80] . It could interfere with the HSV-1 DNA replication process , making it critical for the virus to decrease IFI16 protein levels prior to the bulk of DNA replication . Further studies are needed to fully appreciate the consequences of IFI16 degradation during the HSV-1 replication cycle . It is clear that IFI16 can discriminate between host and foreign DNA [20] , [28] , . Some studies suggest that this is due to differences in chromatinization state [38] . Because of the global differences in chromatin modifications we observed and the activation of IFI16 during EBV and KSHV latency , during which viral DNA is chromatinized [40] , [42] , we believe that there must be other factors involved . Perhaps the topography of viral DNA is different from that of the host genome , leading to increased affinity , or IFI16 may bind DNA in concert with other DNA binding proteins , thereby increasing the affinity or avidity of an interaction between IFI16 and viral DNA when compared with those of IFI16 and cellular DNA . ICP0 is a multifunctional IE alphaherpesvirus protein that is important in reactivation of HSV-1 from latency [1] . Though mechanisms of HSV-1 latency establishment are not yet well understood , ICP0 has long been considered a confounding factor . Recently a neuron-specific microRNA , miR-138 , was shown to be important in the establishment of HSV-1 latency by targeting ICP0 mRNA for degradation [81] . ICP0 is necessary for the degradation of IFI16 in non-ICP0-complementing cells [19] , [20] , [44] and here we show that IFI16 inhibits HSV-1 gene expression . If , in neurons , miR-138 effectively disposes of ICP0 , IFI16 would remain stable and able to carry out its role in gene repression . Our studies suggest that a balance between ICP0 and IFI16 may play a crucial role in determining the outcome of infection . Additionally , because the restriction of HSV-1 by IFI16 is independent of the roles IFI16 plays in the inflammasome and interferon responses ( Figure 4 ) , this offers a potential mechanism for IFI16 control of HSV-1 lytic gene expression during latency maintenance that does not lead to inflammation . IFI16 is , indeed , expressed in human neurons ( Johnson , unpublished data ) and further studies could elucidate its potential role in HSV-1 latency establishment . Like that of other innate immune factors [47] , the role of IFI16 in different innate responses may be cell type-specific and modulated by HSV-1 proteins such as ICP0 and the Us3 protein kinase [44] . We showed here that IFI16 inhibits HSV-1 replication in multiple cell types ( Figures 1–2 ) , suggesting that this effect is general and not likely to be cell-type dependent . Recent studies that point to the relative stability of IFI16 in HSV-1-infected U2OS cells [44] and more established studies suggesting that replication-defective HSV-1 ICP0 mutant viruses can successfully replicate in U2OS cells [82] caused us some concern . However , a ) similar effects of IFI16 on HSV-1 replication and gene expression in HFF and U2OS cells ( Figures 1 and 2 ) , b ) IFI16 degradation in HSV-1-infected U2OS cells ( Figure 6 ) albeit at a much later time point than in HFF cells [19] , [20] and , indeed , a later time point than was tested previously in U2OS cells [44] , and c ) enrichment of IFI16 on HSV-1 promoters in HFF cells and U2OS cells ( Figure 8 ) , demonstrate that the mechanisms of IFI16-mediated inhibition of HSV-1 gene expression are shared between cell types . The decreased association of IFI16 with the late HSV-1 Us11 promoter in U2OS cells compared with that in HFF cells may provide insights into the increased viral yield in U2OS cells compared with that in HFF cells ( Figures 1 and 2 ) . Us11 has many diverse pro-HSV-1 functions , including promotion of protein synthesis , intracellular trafficking , inhibition of RIG-like receptor signaling , and inhibition of autophagy [83]–[88] . Further studies are essential to clarify these effects . Importantly for the establishment of our CRISPR-mediated IFI16-negative cell line , a clonal population of U2OS cells can be grown from a single cell ( Figure 5 ) , which ensures identical genotypes for all cells tested . We were not able to grow such clonal populations of HFF cells . Given the consistency between U2OS and HFF cells described above , the U2OS and IFI16-negative clone 67 provide a valuable tool for further studies of the antiviral roles of IFI16 . The slower cell growth rate and change in morphology of IFI16-negative U2OS cells could be due to the interactions of IFI16 with p53 , which affect cell cycle dynamics [35] . p53 has important roles in cell morphology [89]–[92] , which may also be modulated by its association with IFI16 . Several antiviral activities have now been described for IFI16 to counter infection by a broad range of viruses , including α- and γ-herpesviruses , HIV , and Vaccinia virus [19] , [20] , [28] , [38]–[42] , [45] , [93] , [94] . Developing drugs that stabilize the foreign gene repressive functions of IFI16 ( this study and [19] , [38] , [45] ) and transiently support the IFN-inducing and inflammasome-activating functions of IFI16 without allowing for constitutive innate immune signaling could lead to effective , broad range antiviral therapeutics . To safely design such a drug , future studies need to be done to further characterize the nuances of IFI16-DNA binding , IFI16-mediated IFN induction , and HSV-1-induced IFI16 degradation .
Human osteosarcoma cells ( U2OS ) , human foreskin fibroblasts ( HFF cells ) , MCF7 ( breast epithelial cancer ) cells , human embryonic kidney ( HEK293T ) , and African green monkey ( Vero ) cells were from American Type Culture Collection ( ATCC , Manassas , VA ) . These cells were propagated in Dulbecco's modified Eagle Medium ( DMEM ) supplemented with Glutamax ( Gibco , Grand Island , NY ) , 10% fetal bovine serum ( FBS; Atlanta Biologicals , Lawrenceville , GA ) , and 1% penicillin/streptomycin ( Gibco , Grand Island , NY ) . They were routinely tested for mycoplasma using the Mycoalert kit ( Lonza , NJ ) , according to the manufacturer's instructions , and were found to be negative . KOS strain HSV-1 was propagated and titered by plaque assay on Vero cells , as described [95] . Briefly , Vero cells were infected with HSV-1 at an moi of 0 . 001 pfu/cell until cells began to round up and could be shaken from the flask ( 3–5 days ) . At 4–6 h prior to harvest , 50 µg/mL heparin was added to cell culture supernatant . Cells were removed from cell culture supernatant by centrifugation at 1 , 000 rpm for 10 min at 4°C . Virus was isolated by further centrifugation of the supernatant at 20 , 000×g for 2 h at 4°C . Pellet was resuspended in PBS-AB/15% glycerol and stored at −80°C . To generate 5-ethynyl-2′doxyuridine- ( EdU ) and 5-bromo-2-deoxyuridine ( BrdU ) -labeled infectious HSV-1 virus , a modified protocol described to produce BrdU-labeled HCMV [96] and KSHV [93] was used . Briefly , while propagating KOS strain HSV-1 , EdU Labeling Reagent ( Life Technologies , Camarillo , CA ) was added to flasks at 50 µM and BrdU Labeling Reagent ( Life Technologies ) was diluted 1∶100 and added to the culture medium at 8 , 24 , and 48 h p . i . Flasks with media containing BrdU were kept in darkness or dim light during incubation to avoid photolysis of BrdU residues . Viral purification was carried out as described [95] . Cells were incubated with HSV-1 for 2 h or until the time indicated in serum free DMEM , washed with PBS , and incubated in DMEM supplemented with 2% FBS until the times indicated . Viral yield at 24 h p . i . was determined by titering on Vero cells . Briefly , infected cell supernatants were cleared of cell debris by centrifugation . Vero cells were infected in duplicate or triplicate with serial dilutions of supernatants for 2 h in serum free DMEM , washed with PBS , overlaid with 1× DMEM/1% agarose , and incubated at 37°C until plaque formation was observed ( 48–72 h ) . Cells were fixed by overlaying the agarose layer with 4% paraformaldehyde in PBS for 20 min and then stained with 0 . 2% crystal violet in 50% methanol for 20 min . Dye was washed off and plaques counted . Figures shown are representatives of three or more experiments , each . A plasmid was constructed by cloning IFI16 guided RNA ( target sequence: GAAAAGTTCCGAGGTGATGCTGG synthesized within the guided RNA scaffold [54] ) into pGEMT , using NheI sites . Using Lipofectamine LTX and Plus reagent ( Life Technologies ) , U2OS cells were transfected with 3 plasmids encoding: guided RNA , Cas9 ( Addgene plasmid 41815 , a generous gift from Dr . George Church [54] ) , and GFP at a ratio of 4∶1∶1 , respectively . At 48 h post-transfection , GFP-positive cells were sorted individually into 96-well plates containing complete growth media . Lack of IFI16 expression in each clone was screened by dot blot and confirmed by western blot . The following antibodies were used in Western blot and immunofluorescence analysis: mouse anti-IFI16 and anti-BrdU ( Santa Cruz Biotechnology , Inc , Santa Cruz , CA ) , TBP ( Abcam , Cambridge , MA ) , ASC ( MBL Laboratories , Woods Hole , MA ) , and rabbit-anti IFI16 and anti-actin ( Sigma , St . Louis , MO ) . Antibodies used for chromatin immunoprecipitation assays ( ChIP ) were: IFI16 , RNA polymerase II , TBP , Oct1 , total histone H3 , H3K9me3 , H3K4me3 , and HSV-1 VP16 ( Abcam , Cambridge , MA ) , normal mouse IgG ( Santa Cruz Biotechnology , Inc ) . To create an IFI16-expression lentiviral vector , the IFI16 coding region ( NM_001206567 . 1 , nucleotides 291–2482 ) was cloned into pcpsppw [97] using primers ( Table 1 ) and the BamHI and ApaI sites . IFI16 lentiviral vectors were produced using a four-plasmid transfection system , as previously described [97] . Briefly , HEK293T cells were transfected with IFI16 expressing vector and packaging plasmids and the media was changed 16 h after transfection . Supernatants containing the lentiviral vectors were collected 24 h later , passed through a 0 . 44 µm filter and used to transduce cells in the presence of polybrene ( 5 µg/ml , Pierce , Rockford , IL ) . ShIFI16 , shASC , and shCtrl were obtained ( Santa Cruz Biotechnologies ) and HFF cells were transduced according to the manufacturer's instructions . No selection was done . Western blot analysis was performed to confirm the level of knockdown at 48 h post-transduction . Scrambled siRNA and siIFI16 were transfected into HFF cells using the Neon transfection system ( Life Technologies ) , according to the manufacturer's instructions and as described [93] . Briefly , subconfluent cells were harvested , washed once with PBS , and resuspended in resuspension buffer R ( provided ) at a density of 1×107 cells/ml . 10 µL of this cell suspension was mixed with 100 pmol siRNA and microporated at room temperature using a single pulse of 1350 V for 30 ms . After microporation , cells were distributed into complete medium and placed at 37°C in a humidified 5% CO2 atmosphere . 48 h post-transfection , cells were analyzed for knockdown efficiency by western blot . siRNA oligonucleotides ( siGenome SMARTpool ) for IFI16 ( catalog number MHSXX0020 ) were purchased from Thermo Fisher Scientific ( Waltham , MA ) . Cells were lysed in radioimmunoprecipitation assay ( RIPA ) buffer ( 15 mM NaCl , 1 mM MgCl2 , 1 mM MnCl2 , 2 mM phenylmethylsulfonyl fluoride ) supplemented with protease inhibitor cocktail ( Sigma ) , sonicated , and clarified by centrifugation at 16 , 000×g for 10 min . Equal amounts of protein were separated by SDS-PAGE and electrophoretically transferred to nitrocellulose membranes . Membranes were incubated with primary antibodies and secondary antibodies conjugated to horseradish peroxidase ( KPL , Gaithersburg , MD ) . Immunoreactive bands were visualized using ECL western blotting substrate ( Pierce ) . Blots were scanned using FluorChemFC2 software with the AlphaImager system ( Alpha Innotech Corporation , San Leonardo , CA ) . Figures shown are representatives of three or more experiments each . Cells in 8-chamber slides were infected as indicated before fixation for 10 min with 4% paraformaldehyde . They were washed with PBS , permeabilized with 0 . 2% Triton X-100 for 10 min , and blocked with Image-IT FX signal enhancer ( Life Technologies ) for 10 min . Cells were incubated with primary antibodies for 1 hour at room temperature in PBS , 1% BSA , washed , and incubated with Alexa fluor-conjugated secondary antibodies for 1 hour at room temperature in PBS , 1% BSA . They were washed again and mounted onto slides using Slowfade gold mounting reagent with DAPI ( Life Technologies ) . BrdU-labeled virus was detected as described [93] . Briefly , cells were fixed for 10 min with 4% paraformaldehyde then treated with 4N HCl for 10 min at room temperature to expose BrdU to antibody staining . Cells were washed with PBS , permeablized with 0 . 2% Triton X-100 for 10 min , and blocked with Image-IT FX signal enhancer ( Life Technologies ) for 10 min . Cells were stained with a rabbit-derived antibody to BrdU and Alexa Fluor-conjugated secondary antibodies ( Life Technologies ) and mounted onto slides as above . Cells were imaged using a Nikon Eclipse 80i fluorescence microscope and Metamorph software ( Molecular Devices , Silicon Valley , CA ) . Figures shown are representatives of two or more experiments each . Total RNA was extracted from cells using Trizol ( Life Technologies ) , according to the manufacturer's instructions . Briefly , cells were homogenized in Trizol and mixed with chloroform to separate the proteins and nucleic acids . RNA was precipitated from the aqueous layer using isopropanol and washed in 75% ethanol before resuspension in RNase-free water . RNA was DNase treated for 30 min at 37°C ( Promega , Madison , WI ) and reverse transcribed using Multiscribe Reverse Transcriptase ( Life Technologies ) with random primers , according to the manufacturer's instructions . To perform ChIP assays , we used a protocol modified from two previous studies [98] , [99] . Briefly , 90–95% confluent T-150 flasks of cells were cross-linked for 10 min by adding formaldehyde to a final concentration of 1% . Crosslinking was stopped by adding glycine to 125 mM for 5 min . Cells were collected and resuspended in cell lysis buffer ( 5 mM PIPES , pH 8 . 0; 1 mM EDTA; 1% SDS; protease inhibitors ) and incubated on ice for 15 min before being passed through a 27 . 5 gauge needle and a 30 gauge needle 5 times , each . Nuclei were pelleted by centrifugation and then resuspended in nuclear lysis buffer ( 50 mM Tris , pH 8 . 1; 10 mM EDTA; 1% SDS; protease inhibitors ) , incubated on ice for 15 min , and sonicated at an amplitude of 40 , 10 seconds on , 10 seconds off for 20 min . Debris was cleared by centrifugation and supernatant was flash frozen in liquid N2 and stored at −80°C overnight . Nuclear lysates were diluted in ChIP dilution buffer ( 0 . 01% SDS; 1 . 1% Triton X-100; 1 . 2 mM EDTA; 16 . 7 mM Tris , pH 8 . 1; 167 mM NaCl; protease inhibitors ) . Diluted lysates were precleared for 30 min at 4°C with salmon sperm DNA/protein G agarose slurry ( Millipore , Billerica , MA ) and then incubated overnight at 4°C with 1 . 5 µg of the indicated antibody . Immune complexes were collected with salmon sperm DNA/protein G agarose slurry for 1 hour at 4°C and washed with low salt wash ( 0 . 1% SDS; 1% Triton X-100; 2 mM EDTA; 20 mM Tris , pH 8 . 1; 150 mM NaCl ) , then high salt wash ( 0 . 1% SDS; 1% Triton X-100; 2 mM EDTA; 20 mM Tris , pH 8 . 1; 500 mM NaCl ) , and LiCl wash ( 0 . 25 M LiCl; 1% NP-40; 1% deoxycholate; 1 mM EDTA; 10 mM Tris , pH 8 . 1 ) . Complexes were eluted in elution buffer ( 1% SDS; 0 . 1 M NaHCO3 ) . The crosslinking was reversed by adding 1 µL RNase A and NaCl to a final concentration of 0 . 3 M NaCl and incubating at 65°C for 5 hours . Protein was removed by incubating lysate with proteinase K at 55°C for 1 hour . DNA was precipitated with ethanol and resuspended in nuclease-free water before real-time PCR with TSS primers ( Table 1 ) . Real-time PCR was performed using SYBR green ( Life Technologies ) and primers ( Table 1 ) according to the manufacturer's instructions . Briefly , 2 µL of cDNA ( for mRNA experiments ) or immunoprecipitated DNA ( for ChIP experiments ) was added to real-time PCR reaction mixtures containing SYBR green reaction mixture ( final concentration of 1× ) and the appropriate primers ( final concentration of 0 . 25 µM , forward and reverse ) . An ABI Prism 7500 real-time PCR system was used to amplify and detect cDNA . cDNA data were normalized to GAPDH Ct levels and ChIP data were normalized to input samples , as indicated , using the ddCt method [46] . Figures shown are representatives of three or more experiments each . PLA was performed using the DuoLink PLA Kit ( Sigma-Aldrich ) to detect protein–protein interactions using fluorescence microscopy as per manufacturer's protocol . Briefly , HFF cells were cultured and infected with EdU-labeled HSV-1 ( 1 pfu/cell ) for the indicated times in 8 chamber microscopic slides , fixed with 4% paraformaldehyde for 15 minutes at room temperature , permeabilized with 0 . 2% Triton X-100 and blocked with DuoLink blocking buffer for 30 minutes at 37°C . Cells were then incubated with primary antibodies against IFI16 ( mouse monoclonal and rabbit polyclonal ) , diluted in DuoLink antibody diluents for 1 hour , washed and then further incubated for an hour at 37°C with species-specific PLA probes under hybridization conditions and in the presence of 2 additional oligonucleotides to facilitate the hybridization only in close proximity ( <40 nm ) . Hybridized oligonucleotides were ligated to form a closed circle , which served as a template for rolling-circle amplification after adding an amplification solution to generate a concatemeric product extending from the oligonucleotide arm of the PLA probe . Fluorescently labeled oligonucleotides were hybridized to the concatemeric products and the signal was detected as distinct fluorescent dots in the Texas red channel and analyzed by fluorescence microscopy as above . IFNβ secretion was detected using the Verikine™ Human IFN Beta ELISA kit ( PBL Interferon Source , Piskataway , NJ ) according to the manufacturer's instructions . Infected cell supernatant was collected at 6 h p . i . , diluted 1∶1 with sample dilution buffer and attached to the assay wells by incubation at room temperature for 1 h . Wells were washed and incubated 1 h at room temperature with diluted antibody solution , then washed again and incubated 1 h at room temperature with diluted HRP solution . Wells were washed again and incubated 15 min at room temperature with TMB substrate solution in the dark . The reactions were stopped by the addition of stop solution and the absorbance at 450 nm was read using a Synergy2 Biotek plate reader ( Biotek , Winooski , VT ) .
|
HSV-1 , a ubiquitous human pathogen that establishes a life-long infection , has evolved several mechanisms to evade host immune detection and responses . However , it is still subject to regulation by cellular factors . Recently , a host nuclear protein , IFI16 , was shown to be involved in the innate defense response to HSV-1 infection . Here , we provide the first evidence that IFI16 inhibits wild-type HSV-1 replication by repressing viral gene expression independent of its roles in the immune response . We show that IFI16 binds the HSV-1 genome at the transcription start sites of several HSV-1 genes . Using a permanently IFI16-negative cell line that we generated , we demonstrate that IFI16 reduces the association of important transcription factors . IFI16 also promotes global histone modifications by increasing the markers of repressive chromatin and decreasing the markers for activating chromatin on viral and cellular genes . These insights into the role of IFI16 in HSV-1 biology suggest that stabilization of IFI16 is an attractive avenue for antiviral drug development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"viruses",
"herpes",
"simplex",
"virus",
"gene",
"expression",
"and",
"vector",
"techniques",
"medical",
"microbiology",
"virology",
"viral",
"pathogens",
"molecular",
"biology",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"molecular",
"biology",
"techniques",
"microbiology",
"herpesviruses",
"herpes",
"simplex",
"virus-1",
"organisms",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques"
] |
2014
|
IFI16 Restricts HSV-1 Replication by Accumulating on the HSV-1 Genome, Repressing HSV-1 Gene Expression, and Directly or Indirectly Modulating Histone Modifications
|
Campylobacter is one of the main causes of gastroenteritis worldwide . Most of the current knowledge about the epidemiology of this food-borne infection concerns two species , C . coli and C . jejuni . Recent studies conducted in developing countries and using novel diagnostic techniques have generated evidence of the increasing burden and importance of other Campylobacter species , i . e . non-C . coli/jejuni . We performed a nested case-control study to compare the prevalence of C . coli/jejuni and other Campylobacter in children with clinical dysentery and severe diarrhea as well as without diarrhea to better understand the clinical importance of infections with Campylobacter species other than C . coli/jejuni . Our nested case-control study of 439 stool samples included dysenteric stools , stools collected during severe diarrhea episodes , and asymptomatic stools which were systematically selected to be representative of clinical phenotypes from 9 , 160 stools collected during a birth cohort study of 201 children followed until two years of age . Other Campylobacter accounted for 76 . 4% of the 216 Campylobacter detections by qPCR and were more prevalent than C . coli/jejuni across all clinical groups . Other Campylobacter were also more prevalent than C . coli/jejuni across all age groups , with older children bearing a higher burden of other Campylobacter . Biomarkers of intestinal inflammation and injury ( methylene blue , fecal occult test , myeloperoxidase or MPO ) showed a strong association with dysentery , but mixed results with infection . MPO levels were generally higher among children infected with C . coli/jejuni , but Shigella-infected children suffering from dysentery recorded the highest levels ( 26 , 224 ng/mL ) ; the lowest levels ( 10 , 625 ng/mL ) were among asymptomatic children infected with other Campylobacter . Adjusting for age , sex , and Shigella infection , dysentery was significantly associated with C . coli/jejuni but not with other Campylobacter , whereas severe diarrhea was significantly associated with both C . coli/jejuni and other Campylobacter . Compared to asymptomatic children , children suffering from dysentery had a 14 . 6 odds of C . coli/jejuni infection ( p-value < 0 . 001 , 95% CI 5 . 5–38 . 7 ) but were equally likely to have other Campylobacter infections–odds ratio of 1 . 3 ( 0 . 434 , 0 . 7–2 . 4 ) . Children suffering from severe diarrhea were more likely than asymptomatic children to test positive for both C . coli/jejuni and other Campylobacter–OR of 2 . 8 ( 0 . 034 , 1 . 1–7 . 1 ) and 1 . 9 ( 0 . 018 , 1 . 1–3 . 1 ) , respectively . Compared to the Campylobacter-free group , the odds of all diarrhea given C . coli/jejuni infection and other Campylobacter infection were 8 . 8 ( <0 . 001 , 3 . 0–25 . 7 ) and 2 . 4 ( 0 . 002 , 1 . 4–4 . 2 ) , respectively . Eliminating other Campylobacter in this population would eliminate 24 . 9% of the diarrhea cases , which is almost twice the population attributable fraction of 15 . 1% due to C . coli/jejuni . Eighty-seven percent of the dysentery and 59 . 5% of the severe diarrhea samples were positive for Campylobacter , Shigella , or both , emphasizing the importance of targeting these pathogens to limit the impact of dysentery and severe diarrhea in children . Notably , the higher prevalence of other Campylobacter compared to C . coli/jejuni , their increasing burden during early childhood , and their association with severe diarrhea highlight the importance of these non-C . coli/jejuni Campylobacter species and suggest a need to clarify their importance in the etiology of clinical disease across different epidemiological contexts .
Campylobacter are curved gram-negative bacteria and a common cause of enteritis worldwide . The Campylobacter genus is composed of 26 species [1] , some of which are thermo-tolerant such as C . jejuni and C . coli , the most commonly reported Campylobacter species affecting humans ( 82 . 7% and 17 . 3% , respectively ) [2] . Campylobacteriosis , a food-borne zoonotic infection , is usually diagnosed by selective culture from stool samples , using conditions designed to enhance isolation of C . jejuni and C . coli [3 , 4] . As a result , most of our understanding of the burden and clinical manifestations of campylobacteriosis is specific to C . coli and C . jejuni . Diarrhea caused by Campylobacter species other than C . coli and C . jejuni is thought to be less severe [1] , but this common belief is poorly documented [4] . Several studies have identified a significant burden of disease associated with Campylobacter infection in children less than five years of age . This includes studies that have demonstrated that Campylobacter infection is a risk factor for poor linear growth [5–8] . In the Global Enteric Multicenter Study ( GEMS ) , Campylobacter species had the 5th greatest attributable fraction of moderate-to-severe diarrhea in children under 5 , with an attributable fraction as high as 16 . 4% in some areas [9] . The Etiology , Risk Factors and Interactions of Enteric Infections and Malnutrition and the Consequences for Child Health and Development ( MAL-ED ) is a birth cohort study conducted in eight different countries . Analyses from MAL-ED have also identified a significant burden of diarrhea associated with Campylobacter using both enzyme immunoassay ( EIA ) and polymerase chain reaction ( PCR ) for detection [6 , 10] . An analysis of a subset of Campylobacter-positive samples showed that roughly 30% of detections by EIA were due to other Campylobacter species , including C . hyointestinalis , C . troglodytis , and C . upsaliensis , which made up an important fraction of the disease burden [6] . This finding is important because other diagnostic tests , both culture and FDA-approved nucleic acid based including tests such as BD Max Enteric Bacterial Panel , BD Diagnostics ( Hunt Valley , MD ) , and BioFire Film Array Gastrointestinal Panel ( Salt Lake City , UT ) [11] , often target C . jejuni and C . coli but do not reliably detect other Campylobacter species . Prior birth cohort studies in Peru using culture-based methods have also shown that Campylobacter is a contributor to the acquisition of postnatal linear growth deficits among children in Peru [5] . Considering the prevalence of Campylobacter in diarrhea among children under five years old as identified by EIA in MAL-ED , we performed this nested case-control study using samples from the MAL-ED birth cohort study in Peru to better understand the pathogenicity of the other Campylobacter species ( i . e . non-C . coli/jejuni species ) among children living in extreme poverty .
We identified all dysentery samples in the Peru MAL-ED cohort at the time of this study ( n = 99 ) and then identified two severe diarrhea and two control/asymptomatic samples matched by the child’s age ( ± 6 months ) using a random number generator . In total , we selected 99 dysentery stool samples , 198 stools from severe diarrhea , and 198 control samples . Diarrhea was defined as maternal report of three or more loose stools in a 24-hour period or at least 1 stool with visible blood present [12] , and unique diarrheal episodes were separated by at least 2 days without diarrhea [10] . Dysentery was defined as diarrhea with visible blood reported by the child’s mother during the identified illness episode , and severe diarrhea as diarrhea samples with a modified Vesikari severity score ( MAL-ED score ) of ≥ 5 [13 , 14] . Control samples were non-diarrhea stools ( i . e . monthly MAL-ED stools ) . The severe diarrhea samples were selected from a pool of 849 samples , and the control samples from a pool of 8 , 212 samples which were both ordered and selected by random number generator . The samples used in this study were collected as part of the larger MAL-ED studies following protocols approved by the Institutional Review Board of the Johns Hopkins Bloomberg School of Public Health ( Baltimore , USA ) and of the Asociación Benéfica PRISMA ( Iquitos , Peru ) . The data analyzed were anonymized to protect the privacy of the participants ( S1 Data ) . Stools were inoculated on Campylobacter agar base supplemented with 10% defribrinated sheep’s blood with Blaser’s supplement ( Beckton Dickinson , Sparks , MD ) containing vancomycin , cephalothin , trimethoprim , polymixin , and amphotericin B . Plates were incubated for 48 hours at 42°C at 5% O2 , 10% CO2 , 85% N2 . If no growth was observed , agar plates were held at least 72 hours to confirm this finding . Colonies with typical morphology were confirmed by gram stain , and oxidase and catalase testing using standard microbiologic techniques . Campylobacter colonies were identified as C . jejuni if they hydrolyzed hippurate . Methylene blue was performed as commonly described [15] and fecal occult blood screening was done using commercial cards ( Hemoccult , BeckmanCoulter ) per manufacturer’s instructions . Samples were also analyzed for myeloperoxidase , MPO , ( Alpco , Salem , NH ) as described by Kosek et al . to assess neutrophil activity [16] . Campylobacter and Shigella used as controls for qPCR analyses in this study were American Type Culture Collection ( ATCC ) derived . The lyophilized C . coli ATCC 33559 was rehydrated in tryptic soy broth and grown under micro-aerophilic conditions ( 5% carbon dioxide , 10% oxygen and balance nitrogen ) for 48 hours at 37°C . A C . jejuni culti-loop ( ATCC 33291 ) was grown for 10 min with shaking at 37°C in Luria broth then streaked onto blood agar plates with non-selective medium . The lyophilized C . upsaliensis ATCC 49815 was rehydrated in tryptic soy broth with 5% sheep’s blood and 0 . 001% pyridoxal-HCl , and cultured on Tryptone Soy Agar ( TSA ) with 5% sheep’s blood . Both C . jejuni and C . upsaliensis were grown under the same above-mentioned micro-aerophilic conditions for 48 hours at 37°C . Given the reported similarity between some Campylobacter-associated illness and shigellosis [17 , 18] , we also tested the stools for Shigella . S . flexneri ATCC 12022 2b was grown on TSA Congo Red plates overnight at 37°C . DNA was extracted from the control bacterial cultures and stools of children using the MO-BIO DNA Power-Soil extraction kit ( Carlsbad , CA ) . Approximately 0 . 05–0 . 10 g of feces were weighted and used for DNA extraction . 16S sRNA , Campylobacter adhesin to fibronectin ( cadF ) , and invasion plasmid antigen H ( ipaH ) were used to detect all Campylobacter species , C . coli/jejuni , and Shigella by qPCR , respectively . The 25-μl 16S-cadF-ipaH triplex mixtures were prepared using 1 μl of DNA sample , 12 . 5 μl of Taq Environmental Master Mix 2 . 0 ( ThermoFisher Scientific ) , 0 . 375 μl of a primer-probe mix at final concentrations of 0 . 2 μM for each primer and 0 . 1 μM for each probe , and nuclease-free water making up the remaining volume . qPCR was performed on a StepOnePlus instrument ( Applied Biosystems , Foster City , CA ) using the following cycling conditions: 95°C for 10 min , followed by 45 cycles of 95°C for 15 s and 55°C for 1 min . C . coli and C . jejuni were used as positive controls for cadF; C . upsaliensis and S . flexneri as negative controls for cadF . C . coli , C . jejuni and C . upsaliensis were used as positive controls for 16S , and S . flexneri as negative control . S . flexneri served as the positive control for ipaH . Given the similar results obtained when we previously tested our positive and negative control samples in triplicate , the stool samples were not run in replicate; however , a subset of samples were run in duplicate for quality control purposes . C . coli , C . jejuni , C . upsaliensis and S . flexneri controls and negative controls–which included water only ( no primers or master mix ) and reactions with no DNA template–were included in every run . A cut-off cycle threshold ( Ct ) of 38 was used to determine positivity . Samples that were cadF positive were interpreted as positive for C . jejuni/C . coli ( all cadF samples were also 16S positive ) , samples positive for Campylobacter 16S but negative for cadF were designated as positive for other Campylobacter species . The primers and probes used in this study are listed in Table 1 . Data analysis was performed using STATA 13 . 0 ( College Station , Texas ) . Pearson’s chi-squared was used to test the difference between infection status , clinical outcome , age category , and the presence of blood and leukocytes . The difference in MPO levels by infection status and clinical outcome was tested using ANOVA . The odds of Campylobacter detection given clinical outcome ( control , dysentery and severe diarrhea ) were estimated using logistic regression models . We also calculated the odds of experiencing diarrhea ( combining both dysentery and severe diarrhea ) given infection status using a logistic regression model in order to compute the population attributable fraction ( PAF ) according to Blackwelder et al [21] . Age ( broken down into interval of 6 months and treated as a categorical variable ) , sex , and Shigella infections were included as potential confounders based on a priori knowledge [22] . Panel data were analyzed accounting for the clustering at the child level with robust standard errors . Hypothesis testing was done using alpha error of 0 . 05 .
Of the 99 dysentery stool samples identified in the Peru MAL-ED cohort , 198 severe diarrhea and 198 control samples age-matched to the dysentery samples , 85 , 173 , and 181 samples were available for testing and analyzed in each group , respectively . These 439 stool samples were from 201 unique children . Approximately 60% of the samples were from male children; the mean age in each clinical group was 12 months ( Table 2 ) . Almost all the children aged 0 to 8 months old ( n = 145 ) were breastfed: 96 . 2% , 98 . 1% , and 100 . 0% among the dysentery , severe diarrhea , and control group , respectively . As expected , the difference in sex , age and breastfeeding status among the three clinical groups was not statistically significant ( Table 2 ) . The highest detection of Campylobacter was reported in the dysentery group –60 of the 85 samples tested positive–followed by the severe diarrhea ( 93 of 173 ) and asymptomatic groups ( 63 of 181 ) . The prevalence of C . coli/jejuni was lower than that of other Campylobacter across all clinical groups . Among the Campylobacter-positive samples , 55 . 0% , 81 . 7% , and 88 . 9% were classified as other Campylobacter among the dysentery , severe diarrhea , and control groups , respectively . Overall , 76 . 4% of all the Campylobacter infections ( n = 216 ) were other Campylobacter ( n = 165 ) . Shigella infection was also more predominant in the dysentery group where 36 . 5% of the samples tested positive , compared to 17 . 9% in the severe diarrhea group and 11 . 1% in the asymptomatic group . Of the 439 stool samples , 43 . 1% ( n = 189 ) were negative for infection with Shigella and Campylobacter while 10 . 9% ( n = 48 ) were co-infected with both bacteria , most of which were Shigella and other Campylobacter co-infections . The infection status by clinical group is presented in Table 3 . The relatively higher prevalence of other Campylobacter was corroborated by the culture-based results performed by our collaborators in Iquitos , Peru ( Fig 1 ) . qPCR results showed a higher prevalence of other Campylobacter ( non-C . coli/jejuni ) than C . coli/jejuni across all clinical groups . It is important to highlight that the culture assay allowed the distinction between C . jejuni and all other hipppurate negative species of Campylobacter , with the latter including C . coli . The results obtained via culture showed a lower prevalence of all three bacteria ( C . jejuni , other Campylobacter ( non-C . jejuni ) , and Shigella ) than the qPCR results , notably for the other Campylobacter , which is most likely inherent to the lower sensitivity of culture compared to qPCR and the low discriminatory power that culture allows . Children aged 0–6 months were the least burdened by Campylobacter or Shigella as 29 . 7% of them tested positive for either bacteria , a lower percentage than the 76 . 9% observed among children over 18 months old . Across all age groups , the prevalence of other Campylobacter was higher than that of C . coli/jejuni and Shigella and followed an overall increasing pattern over time –19 . 8% among children aged 0–6 months and 52 . 7% among those over 18 months . C . coli/jejuni were least common among those over 18 months ( 2 . 6% ) and most common in the >6-12-month group ( 19 . 4% ) . The pattern of Shigella prevalence across age groups mirrored that of other Campylobacter with the highest prevalence among those over 18 months ( 44 . 9% ) and lowest among those 0–6 months ( 2 . 0% ) ( Fig 2 ) . The prevalence of all Campylobacter , other Campylobacter ( non-C . coli/jejuni ) and Shigella increased over time while that of C . coli/jejuni decreased overall . The isolation frequency for all the bacteria groups increased after 6 months and remained higher than the prevalence observed in the 0–6 month group , except for that of C . coli/jejuni among children over 18 months old which reached levels less than those among the infants . To study the association between infection status , clinical outcome , and injury , we performed the fecal occult blood test in conjunction with the methylene blue test ( Table 4 ) . Children suffering from dysentery and those infected with C . coli/jejuni were more likely to have blood present in their stool . The difference across clinical groups for the fecal occult test was only significant among children infected with other Campylobacter as 10 . 7% , 51 . 5% , and 23 . 7% of these children tested positive in the control , dysentery , and severe diarrhea group , respectively . A similar trend was observed with the methylene blue test used to detect leukocytes in the stools . Fecal leukocytes were detected primarily in the dysentery group and among children infected with C . coli/jejuni . Infection with C . coli/jejuni was associated with an equal likelihood of children having leukocytes present in their stools regardless of their disease status , which contrasts with other Campylobacter infections . Among children who were infected with other Campylobacter , 39 . 4% of those suffering from dysentery had leukocytes present in their stool compared to 10 . 7% among the asymptomatic children and 21 . 1% in the severe diarrhea group . As expected , most of the asymptomatic children tested negative for methylene blue whether they were free of any Campylobacter and Shigella , infected with other Campylobacter , or infected with Shigella ( approximately 10% ) . This percentage was more than quadrupled when asymptomatic children harbored C . coli/jejuni infections ( 42 . 9% ) ; this difference in percentage was not significant across clinical groups ( 51 . 9% and 41 . 2% in the dysentery and severe diarrhea group , respectively ) . A biomarker of environmental enteropathy of interest was myeloperoxidase ( MPO ) . MPO data were available for 415 samples and were missing completely at random ( 169/181 controls , 80/85 dysentery samples , and 166/173 severe diarrhea samples ) . As expected , high neutrophil activity correlated with dysentery and Campylobacter or Shigella infections ( Table 4 ) . We aggregated the data presented in Table 4 to compare mean MPO levels across clinical outcome without further breakdown based on infectious status , and mean MPO levels based on infectious status without considering the clinical groups . Dysentery was still associated with the highest MPO levels ( 23 , 163 ng/mL ) , which were significantly different from MPO levels in the control and severe diarrhea groups ( 12 , 333 and 11 , 970 ng/mL , respectively ) . When we compared the MPO levels between children infected with any Campylobacter species and those who were not , the difference was not statistically significant ( 15 , 099 and 13 , 510 ng/mL , respectively ) . More specifically , C . coli/jejuni were associated with significantly more inflammation ( MPO levels of 19 , 909 ng/mL ) than children who tested negative for Campylobacter ( 13 , 510 ng/mL ) while the difference in MPO levels between children infected with other Campylobacter ( 13 , 538 ng/mL ) and Campylobacter-free children was not significant . As expected , the highest MPO levels were noted in children who tested positive for Shigella ( 19 , 738 ng/mL ) and were significantly different from the MPO levels of 13 , 071 ng/mL measured among those who tested negative . The odds of detecting Campylobacter in a stool sample given the clinical outcome ( dysentery , severe diarrhea , control ) were modelled separately for each group of Campylobacter species ( C . coli/jejuni , other Campylobacter , and all Campylobacter ) adjusting for age , sex , and Shigella infection . Compared to the asymptomatic group , the odds of detecting C . coli/jejuni were 14 . 6 among the dysentery samples ( p-value <0 . 001 , 95% CI: 5 . 5–38 . 7 ) and 2 . 8 ( p-value = 0 . 034 , 95% CI: 1 . 1–7 . 1 ) among the severe diarrhea samples . Using the same reference group , the odds of being infected with other Campylobacter were lower than those of C . coli/jejuni: children in the dysentery group were equally likely to have other Campylobacter in their stools compared to the controls ( odds of 1 . 3 , p-value = 0 . 434 , 95% CI: 0 . 7–2 . 4 ) while children in the severe diarrhea group were 1 . 9 times more likely to have other Campylobacter in their stools ( p-value = 0 . 018 , 95% CI: 1 . 1–3 . 1 ) . The previous models allowed the comparison of the odds of infection with C . coli/jejuni ( or other Campylobacter ) in the entire study population ( n = 439 ) , i . e . the odds of being infected with C . coli/jejuni ( or other Campylobacter ) versus not being infected with C . coli/jejuni ( or other Campylobacter ) , with the latter also including stools that were negative for both Campylobacter and Shigella . To compare the odds of infection with C . coli/jejuni to infection with other Campylobacter , another logistic regression model was used , restricting the sample size to the stools that were positive for Campylobacter spp . ( n = 216 ) . Based on this new model , the odds of C . coli/jejuni infection were 6 . 6 among the dysentery samples ( p-value = 0 . 006 , 95% CI: 1 . 7–24 . 9 ) , and 1 . 8 among the severe diarrhea samples ( p-value = 0 . 298 , 95% CI: 0 . 6–5 . 7 ) compared to the controls . This demonstrates that C . coli/jejuni infections were more likely than other Campylobacter infections to be detected among the dysentery samples , but C . coli/jejuni and other Campylobacter were equally likely to be found among severe diarrhea cases . The overall odds of Campylobacter infection ( both C . coli/jejuni and other Campylobacter infections ) in the study population were significantly increased among the dysentery samples ( OR = 4 . 9 , p-value < 0 . 001 , 95% CI: 2 . 5–9 . 8 ) and severe diarrhea samples ( OR = 2 . 4 , p-value = 0 . 001 , 95% CI: 1 . 4–4 . 0 ) compared to controls . The odds of Shigella infection were also increased in the dysentery samples ( OR = 6 . 4 , p-value < 0 . 001 , 95% CI: 2 . 4–17 . 3 ) , but the increased odds of 0 . 9 were not significant in the severe diarrhea samples . There was no significant association between C . coli/jejuni infection and age , but the increased odds of other Campylobacter and Shigella infections were statistically significant across age groups , with infections being more prevalent among older children , supporting the results from the descriptive analysis . To calculate the PAF , the association between diarrhea and infection status was investigated , combining and categorizing the dysentery and severe diarrhea groups as diarrhea ( Table 5 ) . Adjusting for sex , age , and Shigella infection , children infected with C . coli/jejuni were 8 . 8 times more likely to have diarrhea compared to those who tested negative for Campylobacter ( p-value <0 . 001 , 95% CI: 3 . 0–25 . 7 ) . Using the same reference group , infection with other Campylobacter increased the odds of diarrhea to 2 . 4 ( p-value = 0 . 002 , 95% CI: 1 . 4–4 . 2 ) . In other words , children with C . coli/jejuni-positive stools were approximately three times more likely to have diarrhea than children infected with other Campylobacter species . Adjusting for sex , age , and Campylobacter infection , children infected with Shigella were approximately three times more likely to have diarrhea ( OR = 3 . 3 , p-value = 0 . 003 , 95% CI: 1 . 5–7 . 3 ) . With other Campylobacter detected in 42 . 3% of the diarrhea samples and odds of diarrhea of 2 . 4 , approximately a quarter of all diarrhea cases in this study ( 24 . 9% , 95% CI: 12 . 1–32 . 2% ) was attributed to other Campylobacter; this PAF was higher than that of C . coli/jejuni . Indeed , eliminating C . coli/jejuni would only eliminate 15 . 1% of the diarrhea cases ( 95% CI: 11 . 4–16 . 4% ) considering their low prevalence among the diarrhea samples ( 0 . 2 ) , despite the odds of diarrhea of 8 . 8 . Overall , 40 . 8% of the diarrhea cases ( 95% CI: 27 . 1–48 . 6% ) could be attributed to Campylobacter ( prevalence of 0 . 6 ) , which is more than double the PAF of 16 . 7% ( 95% CI: 7 . 8–20 . 7% ) due to Shigella ( prevalence of 0 . 2 ) .
Our study of the epidemiology of Campylobacter among children in peri-urban communities in the Peruvian Amazon underscores the importance of other Campylobacter species ( i . e . non-C . coli/jejuni ) in the etiology of diarrhea . Other Campylobacter had the highest attributable fraction for childhood diarrhea ( 24 . 9% ) and accounted for more infections than C . coli/jejuni in all three clinical groups ( asymptomatic , dysentery , and severe diarrhea ) , comprising more than half of all Campylobacter infections ( 165 of the 216 Campylobacter detections ) . Although C . coli/jejuni were more strongly associated with dysentery , other Campylobacter were equally likely as C . coli/jejuni to be detected in severe diarrhea cases . The lower percentage of other Campylobacter detected using culture methods is most likely due to the lower sensitivity of this diagnostic tool optimized for the detection of C . jejuni , and the higher sensitivity of qPCR compared to culture [4 , 23] . The higher prevalence of other Campylobacter compared to that of C . coli/jejuni is at odds with some of the published literature which has shown that C . coli/jejuni are responsible for more than 80% of all Campylobacter infections and are the only species of major public health importance [2 , 3] . However , it is important to highlight that the prevalence of C . coli/jejuni does fall within the reported range in developing countries [4 , 24] , suggesting that the discrepancy between our results and published literature stems from the fact that we were detecting more of the non-C . coli/jejuni species [3] . Additionally , others have also reported instances where more than 50% of the Campylobacter isolated were non-C . coli/jejuni species [25] . As more research is being conducted in developing countries using diagnostic techniques that detect other Campylobacter species more reliably , the importance of non-C . coli/jejuni species has been raised [3 , 4 , 25 , 26] . Discordance analysis of BioFire rapid diagnostic system also reported a high amount of culture-negative but probe-positive specimens in a study conducted in different clinics across the USA ( 58 positive using BioFire compared to 35 positive using culture ) [11] , suggesting the other Campylobacter may not be entirely restricted to impoverished settings . Our study also found evidence that the detection of other Campylobacter was more frequent than C . coli/jejuni among children across all age groups and increased with age while that of C . coli/jejuni decreased in the second year of life . Despite the overall increase in diarrhea and infection with all Campylobacter , other Campylobacter , and Shigella from 0 to 18 months , the illness-to-infection ratio decreased over time during the first 18 months . This trend , not observed with C . coli/jejuni , suggests a shift towards more asymptomatic infections which may be explained by the natural immunity developed after early exposure to these bacteria [3 , 17 , 24] . Why such trend is not observed with C . coli/jejuni remains unclear and requires additional investigation . The surge in the illness-to-infection ratio observed among the children aged 18–32 months might be due to the relatively small number of children in that age group , and requires additional investigation with a larger study population to validate these findings . Some of our results about the association between MPO and clinical outcome were unexpected . Since Shigella and Campylobacter are known pathogens that cause inflammatory diarrhea [18] , a positive association was expected between the level of MPO and infection with any of these pathogens . Our finding of no significant association between MPO and infection with any Campylobacter species is at odds with McCormick et al . ’s study that reported a consistent and positive association between MPO and Campylobacter which had data from all eight MAL-ED sites and greater power to detect associations [7 , 27] . However , it is important to highlight that we did observe a significant positive association between MPO and C . coli/jejuni compared to the negative-Campylobacter group . Despite this positive association , our overall results of no association between MPO and infection with any Campylobacter species seem to be driven by the lack of significant association between MPO and other Campylobacter , which accounted for most of the Campylobacter infections . A limitation of our study was a relatively low number of dysentery and severe diarrhea cases , as this study was a community-based study , and most episodes in such cases are mild to moderate . Another limitation was the focus on only two pathogens , Campylobacter and Shigella . Other microbes most likely contributed to symptomatic infections , especially in the severe diarrhea group . Other diarrhea-causing agents that could be added to the assay include enterohemorrhagic E . coli and Salmonella , as well as Yersinia and E . histolytica as alternate causes of bloody and severe diarrhea [28] . These pathogens would not be expected to be very prevalent relative to Shigella and Campylobacter and indeed have been evaluated in this population ( 10 ) . However , this more inclusive assay would be useful for the evaluation and treatment of individual patients in other epidemiologic settings . Of note , enteroinvasive E . coli would have been classified as Shigella with our PCR assay and is implicitly contained in the data presented . Despite these limitations , our study showed a clear association between clinical outcome and specific Campylobacter species . This finding is consistent with other studies that reported that the highest burden of diarrhea in Loreto , Peru and Venda , South Africa was associated with Campylobacter in the first year of life , and that Campylobacter is the most frequently isolated pathogen in Loreto , Peru [4] . The very few C . coli/jejuni infections detected among children aged 18–32 months ( n = 2 ) might be an artifact of the rather small number of samples from children in that age group . Most importantly , the burden of other Campylobacter was most likely underestimated in our study which defined other Campylobacter as positive for 16S rRNA but negative for cadF . Since infections with C . coli/jejuni and other Campylobacter are not mutually exclusive , it is possible that some of the C . coli/jejuni-positive samples were also positive for other Campylobacter . A comparison of the cadF and 16S rRNA Ct values in the C . coli/jejuni group revealed that 50 out of the 51 samples had a Ct difference greater than 8 . Given the average three copies of 16S rRNA per C . coli and C . jejuni genome [29–31] , the burden and impact of other Campylobacter is most likely greater than reported . A more refined assessment of the burden of these other Campylobacter species will require sequencing or the development of assays that detect gene sequences specific to these species , which would strengthen our findings . Lastly , the importance of co-infections and how to attribute disease burden in this and other studies of enteric disease in the setting of comprehensive diagnostics remains a challenge , but the random selection of samples from the collection makes alternative co-infections unlikely to undermine study findings . Overall , our study highlights the importance of other Campylobacter in the etiology of diarrheal disease and contributes to filling the knowledge gap about the epidemiology of Campylobacter infection in developing countries . Although C . coli/jejuni are important causes of dysentery , the severity of disease caused by non-C . coli/jejuni Campylobacter species and the frequency with which they are detected combined with a greater than expected association with moderate to severe diarrhea strongly suggest that they are the cause of an important fraction of the Campylobacter burden in children in the developing world . Further work is underway to determine which of these other Campylobacter species are most prevalent and strongly associated with both clinical disease , enteropathy and the acquisition of linear growth deficits in this population .
|
Campylobacter is a major public health concern in developed and developing countries . C . coli and C . jejuni have long been considered to be the major disease-causing species , and clinical microbiologic approaches target these two species . However , less selective diagnostic approaches have shown the increasing importance of other Campylobacter species ( i . e . non-C . coli/jejuni ) . Our case-control study investigated the association between diarrhea , C . coli/jejuni , and other Campylobacter among 439 stool samples from 201 children in peri-urban communities in Loreto , Peru . Three quarters of the 216 Campylobacter detections were associated with other Campylobacter , whose prevalence increased with age and was greater than that of C . coli/jejuni in all age and clinical groups ( dysentery , severe diarrhea , and asymptomatic ) . Despite their lower prevalence , C . coli/jejuni were more strongly associated with higher levels of myeloperoxidase , clinical dysentery , and the presence of leukocytes and blood in the stool compared to other Campylobacter . Other Campylobacter were equally likely as C . coli/jejuni to be detected in severe diarrhea cases–odds ratio of 1 . 9 ( p-value = 0 . 018 , 95% CI 1 . 1–3 . 1 ) and 2 . 8 ( 0 . 034 , 1 . 1–7 . 1 ) , respectively . Removing C . coli/jejuni in this population would eliminate 15 . 1% of diarrhea compared to 24 . 9% if other Campylobacter were eliminated .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"dysentery",
"pathogens",
"tropical",
"diseases",
"microbiology",
"shigella",
"pediatrics",
"diarrhea",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"gastroenterology",
"and",
"hepatology",
"neglected",
"tropical",
"diseases",
"pediatric",
"infections",
"bacteria",
"campylobacter",
"bacterial",
"pathogens",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"medical",
"microbiology",
"gastroenteritis",
"microbial",
"pathogens",
"gastrointestinal",
"infections",
"shigellosis",
"bacterial",
"gastroenteritis",
"diagnostic",
"medicine",
"blood",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2018
|
The other Campylobacters: Not innocent bystanders in endemic diarrhea and dysentery in children in low-income settings
|
Few data on dengue epidemiology are available for Lao PDR . Here , we provide information on the complexity of dengue epidemiology in the country , demonstrating dynamic circulation that varies over space and time , according to serotype . We recruited 1 , 912 consenting patients presenting with WHO dengue criteria at Mahosot Hospital , Vientiane ( central Laos ) , between 2006 and 2010 . Between 2008 and 2010 , 1 , 413 patients with undifferentiated fever were also recruited at Luang Namtha ( LNT ) Provincial Hospital ( northern Laos ) and 555 at Salavan ( SV ) Provincial Hospital ( southern Laos ) . We report significant variations in Dengue virus ( DENV ) circulation between the three sites . Peaks of DENV infection were observed in the rainy seasons , although 11% of confirmed cases in the provinces and 4 . 6% in the capital were detected during the dry and cool seasons ( between December and February ) . Four DENV serotypes were detected among the 867 RT-PCR positive patients: 76 . 9% DENV-1 , 9 . 6% DENV-2 , 7 . 7% DENV-4 and 5 . 3% DENV-3 . DENV-1 was the predominant serotype throughout the study except in LNT in 2008 and 2009 when it was DENV-2 . Before July 2009 , DENV-2 was not detected in SV and only rarely detected in Vientiane . DENV-3 and DENV-4 were commonly detected in Vientiane , before 2008 for DENV-4 and after 2009 for DENV-3 . The phylogenetic analyses of DENV envelope sequences suggest concurrent multiple introductions of new strains as well as active DENV circulation throughout Laos and with neighboring countries . It is therefore of great importance to develop and strengthen a year-round nation-wide surveillance network in order to collect data that would allow anticipation of public health issues caused by the occurrence of large dengue outbreaks .
Dengue is an arboviral disease transmitted to humans by Aedes mosquitoes . Infections are caused by single-stranded positive-sense RNA Dengue virus ( DENV ) from the Flavivirus genus , Flaviviridae family . Any of the four virus serotypes ( DENV-1 to DENV-4 ) can cause dengue fever , dengue hemorrhagic fever and dengue shock syndrome [1]; now regrouped under dengue with or without warning signs and severe dengue [2] . Infection by one of the 4 DENV serotypes confers lifelong immunity to that serotype only , as each is antigenically distinct [3] . Several factors such as prior immunity , viral load and infecting genotype or strain are believed to contribute to the severity of DENV infections [4 , 5] . The risk of severe dengue occurrence also has to be viewed through spatial and temporal distribution of concurrent or sequential circulation of DENV serotypes [6] . DENV is widespread in tropical and subtropical areas and is endemic in more than 100 countries [7] . There are an estimated 390 million DENV infections per year , with only 96 million being symptomatic [7] . According to WHO latest estimates , 500 , 000 people are requiring hospitalization every year , with a ~2 . 5% mortality [8] . Also worrying is that dengue’s apparent global burden has increased four-fold in the past 30 years [9] . Now , 3 . 9 billion people are considered at risk of contracting dengue , 70% of those live in the Western Pacific and in South-East Asia [8] . Lao PDR ( Laos ) is a low-middle income country of ~6 . 5 million people [10] bordered by China , Vietnam , Cambodia , Thailand and Myanmar . In Laos , DENV infection is a major cause of morbidity with a rising fatality rate [11] , and approximately 3 . 9 million people are thought to be at risk of contracting a DENV infection [12] . Although hospitalized dengue cases have been reported since 1979 and recorded in a national database since 2008 [13 , 14] , very few have been laboratory-confirmed [11] . Because only a limited number of studies have been conducted so far [12 , 14–22] , little is known about the epidemiology of dengue or the DENV serotypes circulating within the country . Here , we present DENV molecular epidemiological data from patients at 3 different hospitals in Laos: the provincial hospital of Luang Namtha in the north and the provincial hospital of Salavan in the south for 2008 to 2010 , and Mahosot Hospital in the capital city of Vientiane ( central Laos ) for 2006 to 2010 .
Written informed consent was obtained from all recruited patients or responsible guardians . Ethics approval was obtained from the Lao National Ethics Committee for Health Research and the Oxford Tropical Research Ethics Committee . Venous blood was collected on admission from all patients and during convalescence when possible . Venous blood non-anticoagulated specimens were transported to Vientiane at ambient temperature within 48h such as described [19] . Serum samples were centrifuged at Mahosot Hospital upon reception and then stored at -80°C until use . The following Panbio Ltd . ( now Alere Inc , Waltham , Massachusetts , USA ) ELISA kits were used to investigate DENV infections according to manufacturer’s instructions and as described [19 , 23]: Dengue Early ELISA ( Cat no . E-DEN01P ) , Japanese encephalitis-dengue IgM Combo ELISA ( Cat no . E-JED01C ) and Dengue IgG capture ELISA ( Cat no . E-DEN02G ) . All admission sera from SV and LNT patients and admission sera from dengue ELISA positive patients from Mahosot were extracted with the QIAamp Viral RNA Mini kit ( Qiagen , AG , Hombrechtikon , Switzerland ) according to manufacturer’s instructions . The starting volume was 140μL while the final elution volume was 80μL . Internal phage control was added to all samples in order to monitor the extraction process and to check for PCR inhibitors [24] . The detection of serotypes 1–4 DENV RNA was performed in a single step TaqMan real-time reverse transcription PCR with the SuperScript III Platinum One-Step qRT-PCR kit ( ThermoFisher Scientific , Waltham , Massachusetts , USA ) . Primers and probes followed Leparc-Goffart et al . [25] and 5μL of RNA extract was used as a template in a 25μL reaction volume . Positive samples were further characterised by using serotype-specific primers and probes [25] . Results were classified according to the USA CDC’s definition [26] . Confirmed dengue patients were those either with positive RT-PCR , positive NS1 ELISA or when a negative admission serum was paired with a convalescent serum positive for anti-dengue IgM or IgG . Presumptive dengue patients were those with anti-dengue antibody detection alone and no seroconversion . Patients with confirmed and/or presumptive dengue were classified as dengue patients . Isolation of dengue viruses were performed in a biosafety level 3 laboratory at the Infectious Disease Centre in Mahosot Hospital , from dengue patient sera as described [23] . Following one passage in a 25cm2 flask , RNA was extracted from 140μL of cell culture supernatant with the QIAamp Viral RNA kit according to manufacturer’s instruction ( Qiagen , AG , Hombrechtikon , Switzerland ) . DENV RT-PCR was then performed as described above . Specific amplifications of DENV genomes were performed from RNA extracted from cell culture or , when culture was not available , from patient serum samples . Amplicons were then sequenced by next-generation sequencing using Ion Torrent Personal Genome Machine ( ThermoFisher Scientific , Waltham , Massachusetts , USA ) as described by Baronti et al . [27] in Marseille , France , at the Faculty of Medicine , Emerging Viruses Unit .
In LNT , a total of 1 , 413 patients were recruited between May 2008 and December 2010 ( Table 1 ) . Dengue was confirmed in 90 patients ( 6 . 4% ) and a further 133 ( 9 . 4% ) were classified as having a presumptive dengue infection . During the period of this study , monthly mean temperatures ranged from 18°C to 28 . 8°C ( median 25 . 8°C ) , significantly colder than in Vientiane ( p<0 . 001 using Mann Whitney U test ) and Salavan ( p<0 . 001 using Mann Whitney U test ) . As observed in Vientiane , significant decreases in temperature were recorded between December and February with peaks of rainfall between April and September ( Fig 2 ) . As in the capital , more patients recruited in the rainy season were diagnosed with dengue although 10/90 ( 11 . 1% ) confirmed dengue cases were detected during the coldest months of December to February . A total of 34 samples gave positive results with the DENV real-time RT-PCR ( Table 1 ) . DENV-1 and DENV-2 were the only serotypes detected in LNT , in 55 . 9% ( 19/34 ) and 41 . 2% ( 14/34 ) , respectively , of all patients . DENV-2 was the predominant serotype in 2008 and 2009 ( Fig 3 ) . The year 2010 marked a change and DENV-1 then became significantly more frequent . One patient was found positive for both DENV-1 and DENV-2 in 2008 . The genomes of 75 DENV strains were sequenced: 51 DENV-1 ( 20 from patients in Vientiane , 2 from LNT ( one was from 2011 ) and 29 from SV ) , 15 DENV-2 ( 7 from Vientiane , 6 from LNT and 2 from SV ) , 8 DENV-3 ( all from Vientiane ) and 1 DENV-4 ( from SV in 2009 ) ( S1 Table , S2 Table and S3 Table ) . Unfortunately , no samples from 2006 and 2007 were available for sequencing . The dataset includes 43 additional Lao sequences previously published , 33 DENV-1 , 3 DENV-2 , and 7 DENV-3 [23 , 36–38] ( S4 Table ) .
A total of 3 , 880 patients were recruited in this study and 1 , 159 ( 29 . 9% ) were laboratory confirmed as having dengue . Most of the dengue patients were recruited each year from June to November . This reflects typical infection peaks seen during rainy seasons . Of all confirmed dengue cases , 4 . 6% in Vientiane and 11% in SV and LNT were however detected between December and February . This is in line with previous findings of active DENV circulation during dry seasons [22 , 23] . Numbers for Vientiane are however likely to be underestimated since dengue testing was not systematically done during these drier season periods . Interestingly , the number of dengue patients admitted at Mahosot Hospital was significantly higher in 2007 and 2010 , in the context of a regional epidemic in 2010 [11 , 21 , 39] . The proportion of dengue cases , among recruited patients , was higher in Vientiane ( 59 . 8% ) than in the other sites ( 15 . 8% in Luang Namtha and 37 . 8% in Salavan ) . This may be explained by differences in recruitment criteria or physician suspicions , by potentially sub-optimal transport conditions from SV or LNT , but also by environmental criteria since the incidence of DENV is influenced by climate and weather patterns ( high temperature and relative humidity have been associated with increased dengue occurrence [40] ) . The situation in the colder and mountainous Luang Namtha region ( northern Laos ) is particularly reminiscent of that of the Chinese Yunnan region ( adjacent to Laos ) where dengue has low frequency [41–43] . DENV-1 and DENV-2 were first reported in 1943 and 1944 , respectively , DENV-1 in Japan [44] , DENV-2 in Papua New Guinea [45] . DENV-3 and DENV-4 were reported simultaneously in 1956 in the Philippines [46] . DENV-1 has been the most frequently reported serotype in the world since its isolation , followed by DENV-2 , DENV-3 and DENV-4 . By the end of the 60s , the four DENV serotypes were co-circulating in South-East Asia , where DENV is now hyperendemic in most constituent countries . In Laos , due to the limited dengue data available , little is known about the dynamics of serotypes circulation . The first serological evidence of the four DENV serotypes circulating in Laos were reported in 1987 [14] . All four serotypes were then detected by RT-PCR among patients in Vientiane in 2004 and 2005 [15] . The National Center for Laboratory and Epidemiology recently reported the results of the Lao National dengue surveillance between 2006 and 2012 [11] . A total of 361 RT-PCR confirmed dengue cases were reported between 2007 and 2012 . The four serotypes were detected with DENV-1 as the main serotype until 2011 , replaced by DENV-3 in 2012 , but it was not stated which provinces the samples came from . Then , in 2013 , a large epidemic of DENV-3 was reported in Vientiane [22] . In our study , all four serotypes were detected: DENV-1 from genotype I Asia 3 clade , the predominant clade , DENV-2 from genotype Asian I , DENV-3 from genotype II and DENV-4 from genotype I . Differing dynamics of serotype circulation over time were however observed at the three sites . DENV-1 was the main serotype detected , in 667 over the 867 DENV RT-PCR positive patients ( 76 . 9% ) . DENV-1 was predominant in Vientiane and SV over the study period , whereas it was rarely detected in LNT before 2010 . In contrast , DENV-2 was the main serotype in LNT in 2008 and 2009 whereas it was rarely detected in Vientiane and Salavan before July 2009 . Interestingly , similar profile to Vientiane and SV was observed in Thailand with predominant DENV1 and an increase of DENV2 after 2009 [47] . DENV-3 was only detected in five patients before May 2010 ( three in Vientiane and two in SV ) and was afterwards detected in 41 patients in Vientiane . Large DENV-3 epidemics were then reported in Vientiane in 2012 and 2013 [22] . The eight DENV-3 strains isolated in our study , from Vientiane in 2010 , are closely related to the Lao 2012–2013 epidemic strains . This shows there was a local circulation of DENV-3 at least 3 years before the 2013 outbreak . Lao et al suggested that this genotype was introduced in Laos in 2011 or before [22] . Our results therefore confirm that the introduction happened prior to 2010 . That we did not find any DENV-3 genotype III also supports the suggestion by Lao et al [22] that this genotype was only recently introduced into Laos . DENV-4 was not detected in LNT and was limited to five patients in SV although it was the second most common serotype circulating in Vientiane in 2006 ( 39 . 7% ) and in 2007 ( 18 . 5% ) , then was not detected in the following years in Laos . DENV-4 probably came from Thailand where peaks of cases were reported in 2005 and 2006 . DENV-4 infections have rarely been detected in the surrounding countries [48 , 49] , perhaps due to the serotype being less prevalent or the infections being subclinical and less severe . Active DENV circulation is indeed likely to occur between Laos and neighboring countries , probably more intensively with Thailand . This country has important commercial ties with Laos , facilitated by a long border with multiple ports of entry and similarities in language , which all allow multiple concurrent introductions of new strains . This is supported by our phylogenic analyses showing a distribution of Lao strains from the same periods in different clusters . In addition , DENV circulation throughout Laos may also be important . Indeed , we observed some DENV-1 and DENV-2 strains grouping in clusters independently of locations and years of isolation . The 83 available Lao envelope sequences for DENV-1 ( when excluding the 1996 strain from Asia 2 clade ) suggest that some strains are maintained over long periods of time after their introduction , whereas others are only sporadically detected . DENV-1 strains were distributed in 10 clusters , with 2 additional strains not fitting in any clusters . Clusters 1 and 7 were predominant since they were maintained in SV and in Vientiane for several years . Cluster 3 seems to have been predominant in Vientiane before the establishment of Cluster 7 . Results from remaining clusters suggest frequent introduction events from neighboring countries ( Fig 5 ) , with no sustainable maintenance over time or with a “silent” circulation inducing undetected mild or asymptomatic infections . Further investigation of clinical data would prove very useful in understanding differences in possible strain-associated pathogenicity or to confirm the hypothesis that propagation of more virulent strains are more visible to public health surveillance [50] . The study of asymptomatic infections would provide evidence as to whether there are specific populations of strains that are actively circulating “silently” . Our study provides a picture of the complexity of dengue epidemiology in Laos , with circulation dynamics varying over time according to the serotype and location . Indeed , important variations were observed between the capital of Vientiane , located on the Thai border and connected by flights and road to an increasing number of countries , and the remoter rural areas of Salavan and Luang Namtha . Dengue epidemiology will become more complex with time as tourism and commercial traffic with neighboring countries increase . It therefore is of great importance to develop and strengthen a year-round nation-wide surveillance network , particularly because DENV circulation during inter-epidemic periods plays a crucial part in the onset and the course of the subsequent epidemics . Long term studies are needed in order to determine if periodic multi-annual cycles of dengue epidemic exist in Laos such as observed in Vietnam and Thailand [48 , 51] . Switches in phylogenetic lineages within one serotype could have important implications as it is often associated with changes in disease severity and incidence . Obtaining more epidemiological as well as virological data will also be crucial for the country for when dengue vaccines becomes available; funding for such a vaccine would need to be justified [52] . This prove to be challenging in developing countries such as Laos , because laboratory facilities capable of confirming dengue cases only exist in the capital city . Innovation to improve simple field collection techniques , such as filter paper or Rapid Diagnosis Tests , to enable surveillance in remote rural Asia is hence needed [53–59] .
|
Dengue is a mosquito-borne disease that can be caused by 4 viruses . It is a flu-like disease but can sometime be more severe and cause hemorrhage or death . An estimated 390 million people are infected every year , mainly in the Western Pacific and in South-East Asia . In Laos , where our study was conducted from 2006 to 2010 , little was known on the circulation of the different dengue viruses . A total of 1 , 912 patients were recruited at Mahosot Hospital , Vientiane ( central Laos ) , 1 , 413 patients at Luang Namtha Provincial Hospital ( northern Laos ) and 555 at Salavan Provincial Hospital ( southern Laos ) . Although most Dengue virus infections were detected during rainy seasons ( where mosquitoes are the most active ) in all three provinces , some patients were also infected during the dryer months . All 4 dengue viruses were detected with different distributions: mostly type 1 in the capital and in the South , and type 1 and 2 in equal proportion in the north . Type 3 and 4 were not detected in Luang Namtha and rarely in Salavan . Comparison of Dengue virus sequences from Laos with sequences collected worldwide showed an active year-round circulation of dengue within Laos and with neighboring countries . It is hence of great importance to develop and strengthen a year-round nation-wide surveillance network in order to collect data that would allow anticipation of public health issues caused by the occurrence of large dengue outbreaks .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"reverse",
"transcriptase-polymerase",
"chain",
"reaction",
"taxonomy",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"lao",
"people",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"tropical",
"diseases",
"microbiology",
"rna",
"extraction",
"ethnicities",
"viruses",
"phylogenetics",
"data",
"management",
"rna",
"viruses",
"phylogenetic",
"analysis",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"extraction",
"techniques",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"sequence",
"alignment",
"bioinformatics",
"artificial",
"gene",
"amplification",
"and",
"extension",
"medical",
"microbiology",
"dengue",
"fever",
"microbial",
"pathogens",
"biological",
"databases",
"evolutionary",
"systematics",
"molecular",
"biology",
"people",
"and",
"places",
"sequence",
"databases",
"flaviviruses",
"polymerase",
"chain",
"reaction",
"database",
"and",
"informatics",
"methods",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"viral",
"diseases",
"evolutionary",
"biology",
"organisms"
] |
2018
|
Molecular epidemiology of dengue viruses in three provinces of Lao PDR, 2006-2010
|
In natural systems , selection acts on both protein sequence and expression level , but it is unclear how selection integrates over these two dimensions . We recently developed the EMPIRIC approach to systematically determine the fitness effects of all possible point mutants for important regions of essential genes in yeast . Here , we systematically investigated the fitness effects of point mutations in a putative substrate binding loop of yeast Hsp90 ( Hsp82 ) over a broad range of expression strengths . Negative epistasis between reduced expression strength and amino acid substitutions was common , and the endogenous expression strength frequently obscured mutant defects . By analyzing fitness effects at varied expression strengths , we were able to uncover all mutant effects on function . The majority of mutants caused partial functional defects , consistent with this region of Hsp90 contributing to a mutation sensitive and critical process . These results demonstrate that important functional regions of proteins can tolerate mutational defects without experimentally observable impacts on fitness .
Genetic changes that alter protein sequence or expression level can lead to adaptation , suggesting these protein properties are central to evolutionary processes . Many studies have individually investigated the effects of changes to either protein sequence or expression level . For example , protein sequences have been optimized under selective pressure using in vitro evolution [1] . In addition , changes in protein sequence relative to synonymous substitutions are a hallmark of positive selection in natural populations [2] , [3] . The influence of protein expression level on fitness has also been well documented [4] . For example , changes to the expression level of the Agouti protein ( but not its sequence ) have been shown to affect fitness in wild mice by modulating coat coloration [5] . In addition , experiments in E . coli demonstrate that expression from the lac operon is rapidly tuned for optimal growth over a wide range of lactose concentrations [6] . While most studies to date have focused individually on either expression level or protein sequence , in principle the fitness effects of these two protein properties are interdependent [7] , [8] . Here , we systematically investigate selection on the sequence and expression level of yeast Hsp90 ( Hsp82 ) . We recently developed an approach termed EMPIRIC [9] , which is a genetic screen that provides fitness measurements of all possible amino acid substitutions in short regions of important genes in yeast . By sampling across the variety of different amino acid substitutions , EMPIRIC provides detailed information about the physical constraints on protein function . We previously reported a bimodal distribution of fitness effects ( DFE ) for an evolutionarily conserved region of the yeast Hsp90 gene[9] , an essential chaperone required for the maturation of many kinases [10]–[12] . Bimodal DFEs , where most mutants have fitness effects close to either null or wild type ( WT ) , appear common in nature as they have been observed in many other fitness studies [13]–[17] . Bi-modal DFEs are consistent with a recently proposed model where the impacts of mutations on protein stability have a dominant impact on fitness [18] . This model is founded on two concepts: positions that contribute directly to rate-limiting steps in protein function are rare; and the natively folded structure is required for function . Under these conditions , selection results in stably-folded proteins [19] , [20] , such that modestly destabilizing mutations can be tolerated without dramatic changes to the fraction of natively folded protein molecules and hence function . Because protein folding is cooperative there is a narrow range of stability where both the folded and unfolded state are highly populated , consistent with relatively few mutations having intermediate function . In this stability-dominated model , mutations to critical functional positions ( e . g . catalytic sites in enzymes ) destroy activity , but are presumed rare and so do not contribute greatly to the DFE . Of note , the prevalence of positions in proteins that directly contribute to rate-limiting steps in protein function and the fragileness of these positions to mutation have not been thoroughly investigated . The effects of mutations on protein function can be investigated based on fitness effects; however , fitness effects need not correspond directly to functional effects . For example , many essential proteins can be dramatically reduced in net function ( defined here as the product of expression level and function per molecule ) without dramatic reductions of fitness [13] , [21]–[25] . Heterozygotes with one null allele are often highly fit , indicating that 50% reductions in net function can be tolerated [26] . The relationship between fitness and the net function of a protein is formally an elasticity function [21] . Around the wild type net function , the elasticity function often has a slope less than one indicating that reductions in net function have dampened impacts on fitness [27] , [28] . Experimental analyses of fitness effects are also constrained by experimental measurement precision , which is currently on the order of 1% [29] . In natural systems , the resolution of selection depends upon the inverse of effective population size and is on the order of 10−7 for yeast [30] , [31] . Thus , the effects of mutations on function that are important in natural selection can be hidden to experimental fitness analyses . For example , the net function of lysozyme in phage T4 must be reduced about 30-fold before experimentally measurable impacts on growth are observed [13] . At the endogenous expression level in this system , large defects in per molecule function are hidden to experimental fitness analyses . We searched for hidden fitness effects in Hsp90 by examining the Hsp90 elasticity function . We varied the expression level of the native protein sequence and monitored effects on yeast growth rate . Determining theHsp90 elasticity function enabled us to estimate mutant effects on per molecule function from fitness measurements . The elasticity function was non-linear such that at the endogenous expression level , mutant defects up to 79% in per molecule function were hidden to experimental fitness analyses . To reveal potentially hidden functional defects of mutants , we repeated EMPIRIC analyses at reduced expression strengths , which systematically varied fitness sensitivity to amino acid substitutions in Hsp90 . Using this approach , we were able to construct a full distribution of mutant effects on function for a region of Hsp90 . Structural analyses suggest that the region we chose to analyze is a putative substrate binding loop [32] . Our experimental fitness analyses at the wild type expression level resulted in a bimodal DFE , which is a hallmark of a scaffolding region with stability dominated effects on fitness [18] . By analyzing fitness at varied expression strengths , we found that the majority of Hsp90 point mutants had intermediate ( 10–90% ) defects in per molecule function that were hidden to our analyses at wild type expression level . These observations indicate the region of Hsp90 we analyzed is involved in a rate-limiting step in function , and supports its putative role in binding to substrates [32] . Because many mutant defects may be hidden to experimental measurement at the wild type expression level , our results suggest that rate-limiting functional sites in proteins may be more prevalent than previously appreciated , and provides a useful guide for interpreting the growing field of systematic mutant analyses [9] , [33]–[40] .
While our initial EMPIRIC study [9] was performed with a temperature sensitive allele of Hsp90 co-expressed with all mutants; here , we report results in an Hsp90 shutoff strain where mutants were analyzed without potential co-expression artifacts . We developed a yeast shutoff strain ( DBY288 ) where the only chromosomal copy of Hsp90 is regulated by a strictly galactose-dependent promoter [41] . In galactose media , the DBY288 strain expressed Hsp90 at endogenous levels and grew robustly . When switched to dextrose media , the DBY288 strain stalled in growth with Hsp90 levels rapidly dropping below detection ( Supplementary Figure S1 ) . This strain enabled plasmid encoded Hsp90 variants to be maintained and amplified under non-selective conditions ( galactose media ) . Switching to dextrose media then applied selective pressure on the plasmid encoded Hsp90 variants . We analyzed the fitness effects of Hsp90 point mutants by performing a bulk competition in the DBY288 strain . A library of plasmids containing all possible single codon substitutions at amino acid positions 582–590 ( Figure 1A ) was transformed into a single batch of yeast . These experiments used a plasmid and promoter construction previously shown to match the endogenous expression level of Hsp90 [42] . Transformed yeast cells were preferentially amplified in galactose media that allowed all mutations including null alleles to propagate . The bulk culture was transferred to shutoff conditions to initiate selection on the mutant library . The beginning of strong selection on the mutant library was estimated from the growth plateau of control cells harboring a null rescue plasmid ( Supplementary Figure S1 ) . After the initiation of selection on the mutant libraries , samples were harvested over the following 36 hours and the relative abundance of each mutant quantified using focused deep sequencing . By comparing the trajectory of each mutant relative to wild type , we directly determined competitive advantage or disadvantage of each amino-acid substitution as an effective selection coefficient ( s ) that represents the competitive asexual growth advantage/disadvantage of each mutant in a defined environment [29] . We have previously demonstrated that the EMPIRIC approach provides highly reproducible measures of fitness effects that strongly correlates with the growth rate of individual mutants grown in monoculture [43] . Consistent with our previous work , effective selection coefficients were highly reproducible ( R2 = 0 . 96 ) in a full experimental repeat ( Figure S2 ) . At the endogenous expression strength , the distribution of fitness effects for this region of Hsp90 was bi-modal ( Figure 1B , Supplementary Table S1 ) , with peaks near wild type and null . Bi-modal fitness distributions are predicted based on a model where fitness effects are dominated by the impact of mutations on protein stability [18] . Thus , our fitness analyses at wild type expression level are consistent with this region of Hsp90 serving a primarily scaffolding purpose . To further probe the relationship between the net function of Hsp90 and fitness , we varied expression level of the WT sequence and analyzed impacts on growth rate ( Figure 2 ) . To vary expression level , we swapped both promoter and terminator ( 3′ untranslated ) sequences . Closely following the start of strong shutoff selection ( 19 hours in dextrose ) , we observed a 2-fold range in growth rate with these constructs ( Figure 2A ) and a 100-fold range in expression level ( Figure 2B ) . We quantified expression level using a Western blot assay directed against an 6×His epitope tag only present on the rescue copy of Hsp90 that we had previously optimized to yield a linear response [44] . These expression level measurements were performed after 19 hours in dextrose , where the second copy of Hsp90 driven by the galactose regulated promoter was undetectable ( Supplementary Figure S1 ) . To further investigate expression level , we developed an Hsp90-GFP fusion construct that we monitored by flow cytometry . Across all promoter constructs , the Hsp90-GFP fusion supported similar yeast growth rates to non-GFP tagged versions ( Supplementary Figure S3 ) . These findings indicate that the GFP fusion has minimal impacts on Hsp90 function . The expression levels determined by GFP and flow cytometry were in close agreement with those measured by Western blotting and the average of both measures was used to estimate expression levels ( Supplementary Table S1 ) . Both the Western and GFP experiments demonstrate that the expression level of Hsp90 can be reduced dramatically ( 15-fold ) without major impacts on growth rate , which is consistent with previous reports [22] , [45] . The growth rate to Hsp90 expression level profile that we determined has the shape of a binding curve ( Figure 2C ) , and can be fit to a binding equation that represents the elasticity function for Hsp90 . This elasticity function defines how yeast growth rate varies with the net Hsp90 function and enabled us to calculate per molecule function of mutants from fitness measurements . The non-linear elasticity function for Hsp90 describes the coupling of mutant effects on function and fitness . For example , when expressed at endogenous levels , an Hsp90 amino acid substitution would need to reduce per molecule function by 79% in order to result in a readily measureable growth defect of 5% . Thus the bimodal DFE that we observe for Hsp90 ( Figure 1B ) does not necessarily imply a bimodal distribution of mutant effects on function . In particular , the fitness analyses do not provide detailed information on mutants with up to 79% defects in function . Due to the shape of the Hsp90 elasticity curve , the bimodal DFE is consistent with either a bimodal distribution of function as predicted by the stability dominated fitness model [18] , or a primarily unimodal distribution of functional effects ( Figure 3 ) . To distinguish between these possibilities we sought to reveal effects on function that could be hidden at wild type expression strength . To reveal the latent function of Hsp90 mutants , we analyze fitness effects at reduced expression strengths ( Figure 4 , Supplementary Table S2 ) . The population in all bulk competitions was managed such that the population size at constriction points was always in gross excess to library diversity ( Supplementary Figure S4 ) . Because there is selection pressure to increase expression in these experiments , we examined the expression level of the wild type Hsp90 sequence over time in shutoff conditions using Hsp90-GFP fusions ( Supplementary Figure S5 ) . Cells respond to selection by increasing expression from weak promoters over time . As predicted by the elasticity function ( Figure 2 ) , the increased expression from weak promoters results in an increase in growth rate ( Supplementary Figure S6 ) . The observed increase in growth rate closely matches predictions based on the expression increase we observed by flow cytometry and the elasticity function , indicating that the underlying model is sound . To minimize the impact of time dependent changes in expression on fitness analyses of coding sequence mutations , we performed bulk competition of Hsp90 mutants over a short time window , 12–48 hours in dextrose ( Supplementary Figure S4 ) . We performed simulations to investigate how the observed increase in expression level over time in shutoff conditions would impact competition trajectories ( Supplementary Figure S7 ) . The impact of increasing expression level has a minor impact on competition trajectories and indicates that constant expression models provide estimates of sufficient quality to interpret general features of the distribution of mutant effects on fitness and function , which is the focus of this study . The DFEs that we observed exhibited a consistent trend as expression strength was reduced . At high expression strength , the majority of mutants had WT-like growth rates , with very few mutants of intermediate effect . As expression strength was reduced , the WT-like peak decreased and the prevalence of mutants with intermediate effects increased . In terms of epistasis , the fitness effects of amino acid substitutions displayed pervasive negative epistasis with expression strength ( Supplementary Figure S8 ) . In terms of function , these results strongly indicate that the DFE at endogenous expression strength ( Figure 1B ) does not mirror the underlying effects of point mutations on Hsp90 function . We estimated mutant effects on Hsp90 function ( Figure 5 , Supplementary Table S3 ) based on fitness measurements at distinct expression strengths and the elasticity function . As described in the methods section , we employed the elasticity function to calculate per molecule function from fitness taking into account bounds on measurement and calculation precision . For example , at the endogenous expression strength , mutants with activity defects of up to 79% were obscured to fitness analyses and were demarcated as such ( functional efficiency >0 . 21 ) . Because a distinct range of function is revealed to selection at each expression strength ( Table 1 ) , our integrated analyses provided estimates of the functional effects of all mutants . Estimates of mutant effects on function based on fitness measurements at different expression strengths exhibit a reasonable correlation ( R2 = 0 . 75 ) ( Supplementary Figure S9 ) . The strength of this correlation , despite simplifying assumptions ( further discussed in the methods section ) , indicates that the calculated mutant effects on function are fair estimates . The distribution of functional effects for a region of a protein provides information about the contributions of that region to biochemical activity . For example , scaffolding regions that are not directly involved in a critical or rate-limiting step in protein function should be hard to break by mutation ( due to selection for stability in the wild type protein ) , but once broken destroy activity [19] , [20] . In contrast , regions that contribute to a rate-limiting step should be easy to injure by mutation , with the severity of mutant defects mediated by the rigidness of chemical and physical requirements ( e . g . catalytic sites in enzymes being ultimately rigid with any mutation destroying activity ) . The distribution of functional effects ( Figure 5A ) for the region of Hsp90 we analyzed had one main peak with most mutations exhibiting partial defects relative to wild type . Our finding is consistent with this region of Hsp90 contributing to a critical and rate-limiting step in function . The intermediate functional defect of most mutants indicates that the chemical and physical requirements are flexible , consistent with this region of Hsp90 providing a hydrophobic docking site for binding to substrates , as was inferred from structure [32] . Taking a closer look at the aromatic amino acids at position 583 ( Phe ) and 585 ( Trp ) located on the surface of the Hsp90 structure , most amino acid substitutions are tolerated when expressed at endogenous levels , but a clear functional preference for hydrophobic amino acids is revealed at reduced expression strengths ( Figure 5B ) . Hydrophobic interactions [46] are malleable to slight alterations in geometry and physical composition compared to other physical interactions ( e . g . hydrogen bonds ) . Thus , it is reasonable that some substitutions that maintain hydrophobicity would be well tolerated , but that most non-conservative substitutions would result in strong defects . Our fitness-based estimates of mutant effects on function integrate over all properties that contribute to cell growth including catalysis , binding affinity , as well as the thermodynamic stability of folding to the native state [18]–[20] , [27] , [47] . In terms of stability , the prevalence of intermediate functional defects that we observe is inconsistent with this region of Hsp90 serving a purely scaffolding function , which theory predicts should exhibit a bi-modal distribution [18] . Furthermore , we observed a similar distribution of functional effects for positions located on the protein surface , which should have relatively small impacts on stability [48] , as those that orient towards the protein interior ( Supplementary Figure S9 ) . This finding suggests that the functional effects of mutants at solvent shielded positions are caused primarily by local structural changes that impact the organization of solvent exposed positions ( e . g . as required for efficient binding to substrate ) . We have observed a similar surface-core relationship in ubiquitin [43] , and at a lower resolution this type of surface-core association has been postulated based on the slow evolutionary divergence of sites in proteins located proximal to binding sites [49] . Of note , Hsp90 is a dimeric protein and subunit folding and association are coupled [44] . Thus , decreased expression strength could increase sensitivity to destabilizing mutations . In this case , destabilizing mutations would exhibit larger activity defects at lower expression strength . Across the dataset our functional estimates are largely independent of expression strength ( Supplementary Figure S9 , Panel A ) . Thus , the effects of mutations on dimer stability appear to have at most a minor impact on our activity estimates , consistent with the location of this region of Hsp90 far from the dimer interface [50] . To further examine the effect of mutations on stability , we simulated the stability effects of each possible point mutation based on the structure of Hsp90 [50] using Rosetta [51] , which accurately predicts the experimental effects of mutations on stability . The simulated stability effects for Hsp90 correlate extremely weakly with activity ( Figure 5C ) , consistent with our conclusion that stability is not a dominant contributor to activity for this region of Hsp90 . Of note , substitutions of amino acids with similar physical and chemical properties ( as estimated by BLOSUM similarity ) to the wild type residue tend to be compatible with function ( Figure 5D ) . The stronger correlation of function with amino acid similarity compared to stability suggests that the stability simulations do not fully capture all biologically relevant structures . For example , high resolution structures of Hsp90 bound to substrate are not available; but if they were available , might provide a stronger structural explanation for the observed functional effects of mutations . To further test our model and conclusions , we experimentally investigated the biochemical properties of five non-conservative amino acid substitutions . We chose mutations that dramatically change the hydrophobic binding surface and largely destroy function ( F583D and W585D ) , mutations that disrupt intra-molecular interactions and severely impair function ( S586H disrupts a buried hydrogen bond , and A587D introduces a buried charge at a solvent shielded location ) , and a charge reversal mutation ( E590K ) on the surface that causes a moderate functional defect . The growth rate of these mutants in monoculture closely matched the fitness effects observed in the bulk competitions ( Supplementary Figure S10 ) . As discussed above , our estimates of function integrates over multiple protein properties . For example , a mutation that increases the degradation rate ( with the synthesis rate unchanged ) should exhibit reduced steady state levels leading to a defect in net function . All of the disruptive individual mutations that we investigated accumulated at similar steady state levels ( Figure 6A ) , suggesting that individual mutations do not commonly disrupt Hsp90 protein levels . We examined the biophysical properties of these non-conservative Hsp90 mutant proteins in purified form . To maximize the sensitivity of these analyses for potential alterations to structure and stability , we generated C-domain constructs . All of the mutations we analyzed are located in the C-domain and do not contact other domains in the Hsp90 structure . The circular dichroism ( CD ) spectra of all five mutant proteins overlay closely with WT ( Figure 6B ) indicating that all of the mutants fold into native conformations with similar secondary structure content to WT . We investigated the stability of each mutant protein to urea-induced unfolding ( Figure 6C ) . Similar concentrations of urea were required to unfold all mutants and WT indicating that none of the mutants compromises folding under native conditions . These findings demonstrate that non-conservative mutations in this region of Hsp90 are generally capable of folding to stable native states , and strengthen our conclusions that the 582–590 region of Hsp90 that we analyzed is not critical for folding stability , and is instead a structurally malleable region that forms a critical hydrophobic docking site . Our studies as well as those of others [21] , [24] , [25] , [27] , [52] , [53] demonstrate that biochemical flux models and the elasticity function in particular provide a fundamental link between molecular and cellular/organismal properties . Non-linear elasticity functions of the identical form to those described here for Hsp90 have also been observed in E . coli for β-galactosidase[53] , isopropylmalate dehydrogenase [24] , and dihydrofolate reductase ( DHFR ) [25] . In E . coli , DHFR point mutations were commonly observed to impact protein degradation rates leading to fitness effects that were strongly dependent on the level of protein quality control [25] . In addition , flux models can provide a mechanistic explanation for many common fitness features including pleiotropy and epistasis [54] . This study clearly demonstrates that functional defects of mutants can be hidden to experimental fitness measurements due to a non-linear elasticity function . Uncovering these latent effects revealed that the region of Hsp90 we analyzed contributes to a rate-limiting step in Hsp90 function . These findings indicate that critical functional regions in proteins are more prevalent than considered based on fitness analyses performed without consideration of the elasticity function . The elasticity function relating net function and fitness is critical for a thorough understanding of mutant fitness effects .
For expression analysis , the yeast Hsp90 gene was cloned into the pRS414 plasmid with different promoters and 3′ untranslated region ( UTR ) . We used constitutive promoters previously demonstrated to generate a wide variation in expression level [55] including GPD , TEF , ADH , and CYC . Constructs were generated with or without the 3′ UTR from the CYC gene , which allowed further variation in expression level [56] . In constructs lacking the CYC terminator , the 3′UTR was composed of sequence from the plasmid vector . All Hsp90 plasmids contained a 6X-His sequence ( GGHHHHHHGGH ) at the N-terminus to facilitate detection by Western blotting . Point mutant libraries previously generated in p417 plasmids [9] were transferred to the pRS414 promoter variant plasmids using SLIC cloning [57] . Briefly , for each promoter strength construct , we prepared a destination vector with the first and last 30 bases of Hsp90 bracketing a unique SphI restriction site . We excised the Hsp90 library from the original 417GPD plasmid using restriction enzymes that cut immediately upstream and downstream of the Hsp90 gene . We cut destination vectors with SphI . We generated ∼30 base complementary overhangs using T4 DNA polymerase in both the destination vectors and the Hsp90 library , annealed the complementary DNA , transformed into competent bacteria , grew in bulk selective ( Amp ) cultures and prepared plasmid . A small portion of the transformation was plated and the number of independent transformants ( ∼30 , 000 ) was in gross excess to the library diversity . In addition , all replication is performed in bacteria where multiple systems ensure high fidelity reducing the probability of undesired secondary mutations . The DBY288 Hsp90 shutoff strain ( can1-100 ade2-1 his3-11 , 15 leu2-3 , 12 trp1-1 ura3-1 hsp82::leu2 hsc82::leu2 ho::pgals-hsp82-his3 ) was generated from the Ecu Hsp90 plasmid swap strain [42] by integration of Hsp90 driven by a GalS [41] promoter together with a HIS3 marker into the HO genomic locus . DBY288 cells were transformed with pRS414 plasmids and selected on synthetic raffinose and galactose ( SRGal ) plates lacking tryptophan ( -W ) . Single colonies were then grown in liquid SRGal-W on a rotator at 30°C to late-log phase ( OD600∼0 . 8 ) . Cells were collected by centrifugation , washed with synthetic dextrose ( SD ) –W media , and then grown in SD-W medium at 30°C in an orbital shaker . Culture density was maintained in log phase ( OD600 between 0 . 1 and 0 . 8 ) by periodic dilution . Culture growth was monitored based on increases in OD600 taking into account cumulative dilution . The log of OD600 versus time was fit to a linear equation to determine growth rate . Analyses were performed on time points in dextrose where control cells lacking a rescue Hsp90 had depleted Hsp90 by Western analyses ( Figure 2B & Supplementary Figure S1 ) and had stalled in growth ( Figure 2A and Supplementary Figure S1 ) . To analyze expression levels of different promoter constructs , cells were grown for 19 hours in SD -W media , and 108 yeast cells were collected by centrifugation , and frozen as pellets at −80°C . Cell lysates were prepared by vortexing thawed pellets with glass beads in lysis buffer ( 50 mM Tris-HCl pH 7 . 5 , 5 mM EDTA and 10 mM PMSF ) , followed by addition of SDS to 2% . Lysed cells were centrifuged at 18 , 000 g for 1 minute to remove debris , and the protein concentration of the supernatants was determined using a BCA assay ( Pierce Inc . ) . Lysates with 15 µg of cell protein were resolved by SDS-PAGE , transferred to a PVDF membrane , and Hsp90 probed using α-HisG antibody ( Invitrogen Inc . ) . Importantly , we have previously shown that detection of this 6×His Hsp90 construct in yeast can be detected with a broad linear range using this antibody and Western blot approach [44] . Flow cytometry was used as an alternative approach to measure the expression level of Hsp90 at the single cell level in yeast cells . A gene encoding EGFP was inserted into the unstructured tail of Hsp90 after amino acid position 684 . This Hsp90-GFP fusion construct was cloned into the variable strength promoter constructs used with non-GFP tagged Hsp90 . These plasmids were transformed into DBY288 yeast competent cells and grown on SRGal-W plates . A single colony of each strain was grown for two days at 30°C in SRGal-W media to near saturation . These cultures were diluted 1∶50 into SRGal-W media and grown to late log phase ( ∼106 cells/ml ) . Each strain was then further diluted 1∶50 in SD-W media for 48 hours at 30°C with dilution every 12 hrs in order to maintain cells in log phase growth . Samples of cells were collected after 19 , 36 , and 48 hours in dextrose . Collected cells were washed twice in wash buffer ( 50 mMTris , 150 mMNaCl , pH 7 . 6 , 0 . 1% w/v BSA ) , diluted to 107cells/ml in wash buffer and analyzed on a Becton-Dickinson FACSCalibur flow cytometer equipped with a 15 mW air cooled 488 nm argon-ion laser using a 530 nm high-pass filter . Greater than 100 , 000 cells were analyzed for each sample . Data were processed and analyzed using FlowJo software . Debris including clumped cells was excluded by gating on the forward and side scatter ( excluded less than 5% of points ) . To compare with bulk Western measurements , mean fluorescence was calculated using cells without GFP in order to subtract out background due to autofluorescence . C-domain constructs of Hsp90 bearing an N-terminal 6×His tag were generated in a bacterial over-expression plasmid , expressed , purified , and analyzed by circular dichroism ( CD ) as previously described [44] . Briefly , CD spectra were obtained using a 1 mm path length cuvette at a protein concentration of 20 µM in 20 mM potassium phosphate at pH 7 and 25°C . Urea titrations were performed under the same conditions using samples that were equilibrated for 30 minutes . Urea concentrations were determined based on their refractive index . CD ellipticity at 222 nm was used to follow urea induced unfolding and the resulting data was fit to a two-state unfolding model as previously described [44] . The effect of point mutants on yeast growth was analyzed as previously described [36] . Time points in dextrose were selected for analysis where control cells lacking a rescue Hsp90 began to stall in growth in order to observe the rapid decrease in relative abundance of deleterious mutants ( e . g . premature stop codons ) . The growth rate of cells harboring the WT coding sequence in bulk competitions was estimated from monoculture growth of WT constructs performed in parallel to the bulk competitions . For the GPD , TEF and TEFΔter constructs we analyzed time points in dextrose of 12 , 16 , 20 , 24 , 32 , 40 , and 48 hours ( Supplementary Table S4 ) . For the CYC , ADH , CYCΔter , and ADHΔter constructs where the relative decrease of deleterious mutants was less severe ( due to slower growth rate of fit mutants ) we analyzed time points in dextrose of 16 , 20 , 24 , 32 , 40 , and 48 hours . To process these time point samples , yeast pellets were lysed with zymolyase and total DNA was extracted and purified through a silica column . The DNA encoding amino acids 582–590 was PCR amplified , and prepared for 36 base single-read Illumina sequencing . 3 . 4×107 high quality reads ( >99% confidence across all 36 bases ) were obtained and analyzed . The relative abundance of each point mutant at each time point for each promoter was tabulated . Effective selection coefficients for yeast growth were determined by linear fits to the change in mutant abundance relative to wild type for each possible codon substitution . To account for the rapid depletion of null-like mutants to noise levels , only the first three timepoints in selection were used to determine effective selection coefficients for stop codons and all other mutants with effective selection coefficients within two standard deviations of stop codons ( corresponding to s = −0 . 28 for GPD , s = −0 . 37 for TEF , s = −0 . 4 for TEFΔter , s = − . 0 . 35 for CYC , s = −0 . 46 for ADH , s = 0 . 44 for CYCΔter , and s = −0 . 43 for ADHΔter ) . Because these null and near-null mutants rapidly deplete from the culture it is challenging to precisely measure their relative growth effects and they were binned as “null-like” ( Supplementary Table S2 ) . Potential noise was analyzed by calculating normalized residuals ( residuals/time points fit ) . Codon substitutions with residuals per time point greater than 0 . 25 or low initial mutant abundance ( mutant/wt less than 0 . 004 ) were omitted ( ∼7% of codons ) . For mutants that persist in the bulk competition ( s>−0 . 1 ) synonymous codons exhibit a narrow distribution ( Supplemental Figure S11 ) indicating that the amino acid sequence is a dominant determinant of fitness . The effective selection coefficient for each amino acid substitution was estimated as the average of the effective selection coefficients of all synonymous codons . Epistasis between expression strength and amino acid substitutions was calculated as the difference in effective selection coefficient at reduced expression strengths relative to endogenous strength . For the epistasis calculations , null-like mutants were considered as true nulls . Thus , a mutant with wild type fitness at endogenous expression strength , and null-like fitness at the reduced expression strength would have an epistasis of −1 . Function per molecule was calculated based on observed selection coefficients , the elasticity function , and the expression level for each different promoter construct using the following equations . ( 1 ) ( 2 ) Where G is growth rate , Gmax is the maximal growth rate , Em is the relative expression level that results in half maximal growth , E is the expression level relative to the endogenous level , F is the per molecule functionof a mutant relative to WT , Wmut is the growth rate of a mutant relative to WT , and s is the effective selection coefficient . Equation 1 is an extension of the elasticity equation ( Figure 2 ) , where the expression of functional molecules or net function ( EF ) is explicitly modeled . With the WT coding sequence ( F = 1 by definition ) , equation 1 simplifies to the elasticity function in Figure 2 . These equations can be combined and rearranged to define F as follows . ( 3 ) Equation 3 was used to estimate mutant effects on function ( Supplementary Table S3 ) using the observed selection coefficients ( Supplementary Table S2 ) , Em = 0 . 014 ( Figure 2 ) , E for each promoter construct based on experimental measurements ( EGPD = 1 , ETEF = 0 . 32 , ETEFΔter = 0 . 094 ) , or estimated from the observed growth rate and the elasticity function for weak promoter constructs where experimental measures of expression were noisier ( ECYC = 0 . 028 , ECYCΔter = 0 . 015 , EADH = 0 . 014 , EADHΔter = 0 . 010 ) . Where growth rates prohibited accurate estimation of fitness ( null-like mutants , or absolute growth rates within 5% of Gmax ) , bounds on relative per molecule function were calculated ( Table 1 ) . For each amino acid substitution , a final per molecule function estimate was generated by averaging across all promoter constructs that yielded a numerical estimate ( and not a bound ) . For all pair-wise numerical function estimates ( e . g . at two different expression strengths ) , we compared function effects between all constructs with adjacent expression levels ( Figure S9 ) . To facilitate biophysical comparisons , we used the Blosum62 matrix [58] to calculate the amino acid similarity to wild type for each possible point mutation , and Rosetta [51] to simulate effects on thermodynamic folding stability . We make the simplifying assumption that expression level is independent of mutations to the coding sequence . Steady state expression level is determined by the rates of both synthesis and degradation . Because degradation occurs after protein synthesis , it should depend primarily on the protein sequence such that synonymous substitutions minimally impact degradation rates . Across our data set we noted that synonymous substitutions did not have dramatic impacts on fitness , suggesting that synthesis rates were relatively independent of mutation . Protein degradation rates vary depending on protein sequence , but all of the mutants that we analyze are single amino acid substitution , and hence minimally differ in overall sequence . In the event that a point mutant impacts degradation rate , it should be consistent across each promoter construct . Thus , mutant impacts on degradation should be rare ( see Figure 6 ) , but would be incorporated into our estimates of function . In analyzing the effect of mutations relative to wild type , we make the simplifying assumption that function is independent of expression level . We examined the validity of this assumption by analyzing the standard deviation in function for each amino acid substitution determined at different expression levels . The average standard deviation was 0 . 1 , indicating that this assumption is valid on a rough scale ( on the order of 0 . 1 ) and is appropriate for interpreting the main features of the distribution of mutant effects on function . Of note , the mutations that we observe to improve function at reduced Hsp90 expression levels ( Figure 5 , Supplementary Table S3 ) may be an artifact of this assumption . The elasticity function does not include a cost of expression and as such has a maximum fitness at infinite expression level . Thus , we assume that expression cost is negligible relative to expression benefit over the range of our analyses . As the expression cost of native proteins is below experimental detection in yeast [59] , this assumption appears reasonable . We infer differences in cellular growth rates from measurements of DNA abundance . This inference is valid if DNA and cellular abundance are coupled . In previous work , we demonstrated that EMPIRIC measurements of fitness based on measures of plasmid abundance correlate strongly with cellular growth rates for a large set of mutants [43] , indicating that plasmid abundance and cellular abundance are coupled . In addition , the copy number of the CEN plasmids utilized in this study is regulated , as cells maintaining multiple CEN plasmids grow slowly [60] . In addition , the low copy number of CEN plasmids is dominant to the addition of high copy genetic elements [61] and genetic alterations that increase CEN abundance are rare [62] . Nonetheless , CEN plasmids are not as stable as chromosomally encoded DNA , which may lead to a small amount of noise in our measurements .
|
Changes in protein sequence or expression strength can both lead to adaptation in natural systems . While many studies have focused individually on either expression strength or protein sequence , in principle the fitness effects of these two protein properties are interdependent . We systematically investigated the fitness effects of both expression strength and protein sequence for the yeast Hsp90 gene ( Hsp82 ) . We analyzed the fitness effects of all possible point mutations in a putative substrate binding loop under seven different expression strengths . The fitness effects of amino acid substitutions were strongly dependent on expression strength . Many point mutations exhibited fitness defects at reduced expression strength that were hidden at the natural expression strength . Revealing these hidden mutant defects suggested that this region of Hsp90 contributes to a rate-limiting step in function , consistent with its putative role in substrate binding . This study is important because it indicates that critical regions in proteins are more prevalent than would be estimated based on experimental fitness analyses performed at natural expression strengths . As hidden fitness effects are likely to occur in other systems , these findings have broad implications for the field of experimental evolution .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"mutagenesis",
"mutation",
"model",
"organisms",
"genetic",
"mutation",
"gene",
"expression",
"genetics",
"population",
"genetics",
"yeast",
"and",
"fungal",
"models",
"biology",
"saccharomyces",
"cerevisiae"
] |
2013
|
Latent Effects of Hsp90 Mutants Revealed at Reduced Expression Levels
|
Rural populations in low-income countries commonly suffer from the co-morbidity of neglected tropical diseases ( NTDs ) . Podoconiosis , trachomatous trichiasis ( both NTDs ) and cataract are common causes of morbidity among subsistence farmers in the highlands of northern Ethiopia . We explored whether podoconiosis was associated with cataract or trachomatous trichiasis ( TT ) among this population . A comparative cross-sectional study was conducted in East Gojam region , Amhara , Ethiopia in May 2016 . Data were collected from patients previously identified as having podoconiosis and from matched healthy neighbourhood controls . Information on socio-demographic factors , clinical factors and past medical history were collected by an interview-administered questionnaire . Clinical examination involved grading of podoconiosis by examination of both legs , measurement of visual acuity , direct ophthalmoscopy of dilated pupils to grade cataract , and eyelid and corneal examination to grade trachoma . Multiple logistic regression was conducted to estimate independent association and correlates of podoconiosis , TT and cataract . A total of 700 participants were included in this study; 350 podoconiosis patients and 350 healthy neighbourhood controls . The prevalence of TT was higher among podoconiosis patients than controls ( 65 ( 18 . 6% ) vs 43 ( 12 . 3% ) ) with an adjusted odds ratio OR 1 . 57 ( 95% CI 1 . 02–2 . 40 ) , p = 0 . 04 . There was no significant difference in prevalence of cataract between the two populations with an adjusted OR 0 . 83 ( 95% CI 0 . 55–1 . 25 ) , p = 0 . 36 . Mean best visual acuity was 0 . 59 ( SD 0 . 06 ) in podoconiosis cases compared to 0 . 44 ( SD 0 . 04 ) in controls , p<0 . 001 . The proportion of patients classified as blind was higher in podoconiosis cases compared with healthy controls; 5 . 6% vs 2 . 0%; adjusted OR 2 . 63 ( 1 . 08–6 . 39 ) , P = 0 . 03 . Individuals with podoconiosis have a higher burden of TT and worse visual acuity than their matched healthy neighbourhood controls . Further research into the environmental and biological reasons for this co-morbidity is required . A shared approach to managing these two NTDs within the same population could be beneficial .
Neglected tropical diseases ( NTDs ) do not occur in isolation but have substantial geographical overlap . This results in an increased burden of co-morbidity within a population , commonly leading to individuals suffering from one or more NTDs[1] . These conditions share common risk factors including lack of access to clean water , sanitation and hygiene practices[2] . We describe here a study in which we aimed to explore the association between the NTD podoconiosis and two common eye diseases; one an NTD ( trachomatous trichiasis ) and another a common age-related disabling eye disease ( cataract ) , within a rural population in northern Ethiopia . Podoconiosis is a non-filarial elephantiasis that predominantly affects subsistent farmers in areas of red clay soil covered highlands of tropical Africa , Northern India and South and Central America [3 , 4 , 5 , 6] . It causes painful swelling and deformity of the lower legs with acute , painful inflammatory events known as acute adenolymphangioadenitis ( ALA ) [7] . Although the aetiology is not fully understood , current evidence suggests it occurs as a result of both a genetic susceptibility and exposure to irritant mineral particles in volcanic soils [7 , 8 , 9 , 10] . The disease carries a high socio-economic burden and is highly stigmatizing [11 , 12 , 13] . Nationwide mapping in 2015 found podoconiosis to be endemic in 345 districts in Ethiopia with a prevalence of 4% [14] . East Gojam zone in Amhara region , where this study took place , has a podoconiosis prevalence of 3 . 3% [15] . The recommended management of podoconiosis is inexpensive and simple involving foot hygiene , emollient , bandaging , exercise and wearing socks and shoes [16] . Anecdotally clinicians and researchers working with populations affected by podoconiosis report a high prevalence of cataract and trachomatous trichiasis ( TT ) among affected individuals . The 2006 National Blindness Survey of Ethiopia found the prevalence of blindness and low vision in Amhara to be 1 . 4% and 4 . 9% , respectively [19]Nationwide , cataract and trachomatous corneal opacity were found to be the leading causes of blindness with cataract accounting for 49 . 9% and trachomatous corneal opacity 11 . 5% [19] . Cataract is a clouding of the lens , which results in decreased vision . The leading cause is age . Cataract develops at a younger age in tropical and poor countries . The precise reasons for this are unclear but it is likely to be due to a combination of factors including episodes of dehydration in early life , diet , and solar and heat radiation [18] . Trachoma is the leading infectious cause of blindness worldwide [17] . The disease starts in childhood with recurrent infection of the tarsal conjunctiva by Chlamydia trachomatis producing chronic inflammation . This leads to tarsal scarring followed by entropion ( inward rotation of the eyelid ) and trichiasis , a painful condition where eyelashes rub on the cornea causing corneal scarring . Ethiopia is one of the most trachoma-affected countries in the world; in 2016 nearly 50% of people at risk of trachoma globally live in Ethiopia , Malawi and Nigeria [20] . The number of people with TT awaiting surgery in 2016 is 693 , 000 , again the largest in the world [20] . The 2006 survey found Amhara Regional State bears an estimated 45% of the national trichiasis burden with approximately one in twenty of all adults suffering from the condition . Since 2001 TT surgery has been provided by the Amhara Regional Trachoma Control Programme with health workers throughout Amhara Region trained to perform TT surgery , including 40 in East Gojam Zone . Ethiopia carries a high burden of Neglected Tropical Diseases , and is estimated to have the highest burden of podoconiosis and trachoma in Sub-Saharan Africa [21] . Both podoconiosis and trachoma are part of the Ethiopian Government’s National Neglected Tropical Disease ( NTD ) Master Plan [22] . Launched in 2013 , the NTD Master Plan pledged to achieve WHO NTD elimination and control targets by 2020 . Evidence of the presence of the double burden of podoconiosis and TT will facilitate decisions on the integration of policy and treatment programmes for these two NTDs .
A comparative cross-sectional study was carried out in East Gojam Zone , Amhara Region , Ethiopia during four weeks in May 2016 . Amhara is one of 9 regions and 2 city administrations of Ethiopia and is divided into 10 zones . East Gojam zone is divided into 20 woredas ( equivalent of districts ) , which are divided into kebeles ( the lowest governmental administrative unit ) . Within East Gojam Zone the Woreda Enarj Enawga was selected for this study as it was known to have a high burden of podoconiosis and access to recent complete data from the Podoconiosis Burden Assessment 2015 [23] . Enarj Enawga has a population of 167 , 402 , of which 92% are rural inhabitants [24] . A non-random convenience sampling method was used to select 12 kebeles within Enarj Enwarga based on geographical location and accessibility . Assuming the prevalence of cataract in healthy controls is 8% and the effect estimate is an odds ratio of 2 , we calculated the need to recruit 336 cases and 336 cases for 80% power and 5% significance [19] . Cases were located using the Podoconiosis Burden Assessment 2015 which provided participant name , household name , and village for all individuals over the age of 40 with podoconiosis within each kebele [23] . Every podoconiosis case over age 40 within each of the 12 kebeles was invited to take part in the study . In total , 460 podoconiosis cases were identified . They were invited to participate in the study through both verbal information via the Kebele Leader and a written letter of invitation asking them to attend a health centre on a particular date . Neighbourhood controls matched by age ( +/- 5 years ) , sex and village were selected randomly from a residents’ register detailing each individual living within each village . Village leaders , Health Extension Workers and kebele leaders were able to locate each control selected . All were provided with information about the study and invited by letter to attend a health centre within their kebele on a particular date . Controls were excluded if they were under 40 years old , had podoconiosis on clinical examination , or had a first-degree relative with podoconiosis . The primary outcomes for participants with podoconiosis and healthy neighbourhood controls were: 1 ) presence of cataract 2 ) presence of TT . Secondary outcomes were: grade of cataract , severity of TT , visual acuity , socio-demographic variables ( age , sex , occupation ) , socio-economic status , history of hypertension , history of diabetes , previous eye surgery , previous eye trauma , previous eye diagnosis , number of acute podoconiosis attacks and stage of podoconiosis . Data were collected through an interview-administered questionnaire followed by clinical examination . Questionnaires were administered by trained Amharic-speaking data collectors . Visual acuity was tested using PEEK Visual Acuity on smartphones recorded on a LogMar scale [25] . LogMAR values were categorised according to the ICD-10 classification of visual impairment; normal vision , ≤ 0 . 4; mild impairment , 0 . 4–1 . 0; severe impairment , 1 . 0–1 . 3; blindness ≥1 . 3 [26] . In this study we present only best unaided vision as no patients had spectacles for distance correction . Podoconiosis grading involved examination of both legs using a validated 5 stage grading system carried out by a local Podoconiosis Nurse Specialist [27] . An ophthalmic officer and a medical doctor carried out the clinical eye examinations . Tropicamide 1 . 0% mydriatic drops were administered to both eyes . Cataract was examined using a direct ophthalmoscope at 30cm from dilated pupils as slit lamp examination was unavailable . Cataract was classified into five grades according to Mehra and Minassian’s method of grading in eye surveys using degree of opacity in the red reflex to define the grade [28] . Grades 4 and 5 were regarded as severe matured cataract requiring surgery . Both eyes were then examined using a torch and x2 . 5 magnifying binocular loop for signs of trachoma . Each eye was examined for in-turned lashes ( TT ) , the cornea inspected for central corneal opacities ( opacities within central 4mm ) and the upper conjunctiva everted and examined for inflammation ( Trachomatous Inflammation—Follicular ( TF ) and Trachomatous Inflammation—Intense ( TI ) ) and scarring ( Trachomatous Scarring ( TS ) ) . The WHO simplified trachoma grading system was used to define each of these stages [29] . TT was defined as one or more lashes touching the globe or evidence of eyelash epilation [17] . Severity of TT was recorded by counting the number of lashes touching the globe when looking straight ahead and subdivided into corneal lashes ( touching cornea ) or peripheral lashes ( touching medial or lateral conjunctiva ) . TT severity was classified into two groups: major as >5 peripheral or corneal lashes and minor as <6 peripheral or corneal lashes . Patients were told the findings of the eye examination at the end of the study . Individuals with signs of active trachoma ( TF and/or TI ) were offered treatment with 1% tetracycline eye ointment . TT patients were referred to health centres where free TT surgery was available . Patients with grade 4 or 5 cataract were referred to Debre Markos Hospital for cataract surgery . All podoconiosis patients were counseled by a Podoconiosis Specialist Nurse for podoconiosis management and enrolled into existing podoconiosis clinics . Data were coded , entered , cleaned and analysed using IBM SPSS Statistics version 22 . Descriptive analysis of the socio-demographic and clinical characteristics of cases and controls was performed . When comparing simple frequencies , the χ2 test was used to establish significance . Means were compared using the independent t-test . Principal Components Analysis was used to reduce 15 wealth index factors ( gained from the interview-administered questionnaire ) into three household index values taking the first component as a measure of economic status divided into three categories: poor , middle class and wealthy [30 , 31] . A logistic regression model was used to determine the clinical and socio-demographic correlates of trachomatous trichiasis and cataract . The model was used to measure the association between podoconiosis and these two eye diseases ( cataract and TT ) adjusting for sex , age , occupation and socio-economic status . These confounding factors were chosen prior to data collection . They were chosen as factors likely to influence the association between podoconiosis and these two eye diseases based on previous literature [32 , 33 , 34 , 35 , 42] . Ethical approval was gained from Amhara Regional Health Bureau and the Research Governance & Ethics Committee of Brighton & Sussex Medical School . An Amharic-speaking study supervisor gave , to each of the participants , an introduction to the study and the reasons why it was being conducted . Then , all participants were given written information in Amharic outlining the reason for the study and what would be involved if they chose to participate . If participants were unable to read or write , the information sheet was read to them individually . Informed consent was gained by signature and a thumbprint was used if the participant was unable to write [36] . The consent was then countersigned by an independent witness . Study participants identified with ocular disease or podoconiosis were managed as per local protocol . A podoconiosis nurse was present throughout the study to provide education regarding podoconiosis treatment and integration into existing clinics .
A total of 700 participants were included in this study: 350 podoconiosis cases and 350 healthy neighbourhood controls . The socio-demographic characteristics of cases and controls are described in Table 1 . More cases ( 60 . 3% ) and controls ( 62 . 6% ) were male than female . The mean age distribution between cases and controls was similar at 57 and 56 years respectively . The great majority of both cases and controls were rural farmers and married . However , compared to their neighbourhood controls , significantly larger numbers of podoconiosis cases were either divorced or widowed; 102 ( 29 . 2% ) vs 56 ( 16 . 0% ) , p = 0 . 001 , and lived in poorer households; 146 ( 41 . 7% ) vs 88 ( 25 . 1% ) ; p<0 . 001 . Over half of all podoconiosis patients had a first-degree relative with podoconiosis and had experienced an acute attack in the past 30 days . The median stage of podoconiosis for both legs was stage 2 ( defined as persistent below knee swelling ) [27] . The clinical characteristics of cases and controls are described in Table 2 . Few patients in either group had a history of hypertension and diabetes . Compared to the controls , a significantly higher proportion of podoconiosis cases had an ocular problem; 52 ( 14 . 9% ) vs 86 ( 24 . 6% ) , p = 0 . 001 , and had had eye surgery; 49 ( 14 . 0% ) vs 75 ( 21 . 4% ) ; p = 0 . 01 . In particular , compared to controls , a higher proportion of podoconiosis cases had had TT surgery ( 61 ( 17 . 4% ) in cases vs 42 ( 12 . 0% ) in controls , p = 0 . 04 ) , a diagnosis of TT ( 62 ( 17 . 7% ) in cases vs 40 ( 11 . 4% ) in controls , p = 0 . 03 ) or a diagnosis of cataract ( 16 ( 4 . 6% ) in cases vs 6 ( 1 . 7% ) in controls , p = 0 . 02 ) . Podoconiosis patients were found to have significantly lower visual acuity than healthy controls . ( Table 3 ) . Mean best visual acuity was 0 . 59 ( SD 0 . 06 ) in podoconiosis cases compared to 0 . 44 ( SD 0 . 04 ) in controls , p<0 . 001 . The proportion of patients classified as blind was significantly higher in the podoconiosis group; 5 . 6% vs 2 . 0%; OR 2 . 97 ( 95% CI 1 . 24–7 . 11 ) , p = 0 . 02 . When adjusted for age , sex and socioeconomic status the association remained significant; adjusted OR 2 . 63 ( 1 . 08–6 . 39 ) , p = 0 . 03 . The prevalence of TT was higher in cases ( 65 , 18 . 6% ) than controls ( 43 , 12 . 3% ) ; OR 1 . 63; ( 95% CI 1 . 07–2 . 47 ) , p = 0 . 02 . 13 out of 65 ( 20% ) cases had major trichiasis compared to 5 out of 43 ( 11% ) of healthy controls ( p = 0 . 16 ) . The odds of having TT remain significantly greater for individuals with podoconiosis after adjustment for age , sex , occupation , socio-economic status and distance from water; adjusted OR 1 . 57 ( 95% CI 1 . 02–2 . 40 ) , p = 0 . 04 . Female patients had greater odds of TT compared to male patients OR 1 . 58 ( 95% CI 1 . 38–1 . 77 ) , p<0 . 001 ( Table 4 ) . No significant difference in cataract prevalence was found between the two groups; 272 ( 77 . 7% ) vs 286 , ( 81 . 7% ) , OR , 0 . 87; ( 95% CI , 0 . 59–1 . 26 ) , p = 0 . 47 . However , podoconiosis cases were shown to have more severe cataract . The number of patients with grade 4 or 5 cataract in either eye or both was 37 ( 10 . 6% ) for cases vs 29 ( 8 . 6% ) for controls ( p = 0 . 01 ) . Many more eyes in the podoconiosis case group could not be examined due to severe corneal opacity or phthisis; 43 out 700 ( 6 . 1% ) compared with 14 out of 700 ( 2% ) . The odds of having cataract were not affected by the presence of podoconiosis; adjusted OR 0 . 83 ( 95% CI 0 . 55–1 . 25 ) , p = 0 . 36 , and only significantly associated with age with an adjusted odds ratio of 1 . 07 ( 95% CI 1 . 05–1 . 09 ) , p = 0 . 001 ( Table 5 ) . Podoconiosis patients were found to have a significantly higher prevalence of trachomatous inflammation , trachomatous scarring and corneal opacity than healthy controls ( p < 0 . 001 for each ) .
The study found that podoconiosis patients have worse visual acuity than healthy neighbourhood controls , with many more podoconiosis patients classified as blind . The prevalence of TT causing low vision through corneal opacification is higher in podoconiosis patients , creating a double burden of neglected tropical disease in this population . No significant difference in the prevalence of cataract was observed between podoconiosis patients and controls , however a higher number of podoconiosis cases had dense cataract ( grade 4 or 5 ) and previous cataract surgery when compared to their neighbourhood controls . NTDs commonly overlap within a population . Indeed , 80 million people in Ethiopia live in areas where one or more NTDs co-exist . Mapping of NTDs in Ethiopia has shown much geographical overlap between podoconiosis and other NTDs . For example Nationwide mapping has shown that 29 districts in Ethiopia are co-endemic for LF and podoconiosis , 116 co-endemic for onchocerciasis and podoconiosis , 302 co-endemic for trachoma and podoconiosis , 342 co-endemic for SHT and podoconiosis [37] . However , little is known regarding the overlap of NTDs within individuals . One study showed an overlap between Soil Transmitted Helminth ( STH ) infection and podoconiosis [38] . The authors concluded that this was likely to be a result of barefoot practices predisposing the individual to both diseases , rather than a shared biological mechanism . The association between the two NTDs podoconiosis and TT can be hypothesised to be the result of both shared environmental risk factors and a common biological pathology . It is known that NTDs often co-exist within a population due to shared environmental risk factors such as sanitation , hygiene , poverty and access to health care [2 , 39] . The management of both trachoma and podoconiosis share common hygiene messages; for example , trachoma elimination programmes have focused on promoting facial cleanliness , while podoconiosis programmes educate patients about regular foot washing practices [41 , 16] . Both diseases are associated with reduced availability of water , sanitation and hygiene ( WASH ) facilities . 40 Similarly , both diseases have been linked to poverty [42 , 14] . In this study podoconiosis patients were found to be significantly poorer than their neighbourhood controls , as we would expect from previous studies [41] . Likewise , trachoma is widely considered a disease of poverty [42] . Podoconiosis has a social impact and this in turn may lead to an increased burden of eye disease , independent of an association with trachoma [11] . Alongside economic poverty , individuals with podoconiosis are marginalised and stigmatised within their societies , leading to reduced living standards in comparison to their healthy neighbours [11 , 14] . Marginalisation within society alongside poor living standards among individuals living with podoconiosis may predispose to eye disease and reduced visual acuity through affecting health seeking behaviour [43] , attendance for surgical procedures and disease prevention awareness . Alongside an environmental and socio-economic hypothesis , it is possible that the association between podoconiosis and TT is also a result of shared biological pathology in these two chronic , scarring inflammatory diseases [44 , 45] . Previous research investigating pro-inflammatory and pro-fibrotic markers in serum and HLA associations in these two diseases could provide some insight into a potential shared pathology [8] . , To date , very few studies have investigated the systemic effects of podoconiosis . Bilateral leg lymphoedema is thought to be caused by elements common in irritant volcanic soils ( eg . aluminium , silicon , magnesium and iron ) being absorbed through the foot and entering lower limb lymph nodes [46 , 7 , 10 , 47 , 48] . While it is the foot which facilitates dermal absorption , it is possible that these irritant elements cause wider systemic inflammatory effects . Addisu et al [44] compared levels of oxidative stress biomarkers in podoconiosis patients and healthy controls and found higher levels of these biomarkers in the serum of podoconiosis patients , suggesting inflammation in the early stage of the disease . Burton et al studied progressive trachomatous conjunctival scarring in an Ethiopian and Tanzanian cohort . They found that scarring progressed over a two year period in the absence of C . trachomatis [45] . Progressive scarring was associated with mucosal inflammation with an increased association in individuals with more frequent inflammatory episodes . This chronic conjunctival inflammation was associated with increased expression of pro-inflammatory factors and extracellular matrix regulators systemically including the pro-fibrotic factor CTGF , closely associated with TGFβ . Both podoconiosis and progressive trachomatous scarring have been associated with altered levels of pro-inflammatory factors in the serum . It is possible that systemic inflammation in podoconiosis patients could provide an initial hypothesis to explain the increased levels of trachomatous inflammation , trachomatous scarring and trachomatous trichiasis among podoconiosis cases . Further studies viewing podoconiosis as a systemic disease are required to investigate further potential effects of chronic systemic inflammation in these patients . Many individuals who come into contact with irritant volcanic soils do not go on to develop podoconiosis [46 , 49] . Similarly , the natural history of trachoma varies significantly in prevalence and severity between families and communities with shared environmental risk factors [50] . The gene-environment interaction for both these diseases has been investigated , including the frequency of HLA antigens in both podoconiosis and blinding trachoma . Both podoconiosis and trachoma are associated with HLA Class II suggesting both are T cell mediated inflammatory diseases [8 , 51] . A shared pathogenesis associated with HLA Class II may account for some of the association between these two NTDs . Both podoconiosis and TT have been shown to worsen poverty , reduce quality of life and increase the burden of disability in predominately poor rural communities [52 , 53 , 54 , 55 , 56] . Understanding that individuals can be burdened with both diseases could reduce the disease-specific approach to management and lead to an integrated management of these two NTDs at different levels . Firstly , the screening and diagnosis of these diseases could combine using the same community based health extension worker led household screening that is seen in trichiasis screening in Amhara region , Ethiopia . Both are highly visible diseases making community based screening very effective . Secondly , integrated treatment approaches could focus on common hygiene messages among the two groups and a focus on the improvement of WASH facilities , a key element of NTD management projects globally [40] . A key limitation of this study is the absence of a detailed slit lamp ophthalmic examination of the anterior segment and retina of the eye . While we found that TT is more common in this population , we were unable to study the prevalence of other eye diseases that could also account for the reduced vision in podoconiosis patients . Examining for cataract using direct ophthalmoscopy rather than slit lamp risks missing certain types of cataract , in particular nuclear cataract . A further limitation is the possibility of selection bias . While 460 podoconiosis patients were invited to take part , there was a high attrition rate before enrolment in the study with only 350 enrolling . Those with more disabling podoconiosis or worse vision may have been less likely to attend the local health facilities due to difficulties of travel . Lastly , the study only took place in one small rural population of northern Ethiopia reducing the generalizability of results to other populations affected by podoconiosis , trachoma and cataract .
We conclude that podoconiosis patients have a greater burden of visual impairment than individuals living in the same neighbourhood without the disease . They are more likely to suffer from TT , and other stages of trachoma . Podoconiosis patients are poorer than their neighbours without the disease , but this alone may not be enough to account for this association and the significant difference in their burden of poor vision and blindness . Alongside shared environmental risk factors , shared biological mechanisms between these two NTDs , podoconiosis and TT , may contribute to the association that has been found and warrants further research to gain a better understanding of their co-endemicity . In particular , a focus in the future of studying and managing these two diseases together may help to reduce their burden in this northern Ethiopian population and farther afield .
|
Podoconiosis is an NTD causing chronic leg swelling in subsistence farming communities in the tropics . There is no research on the association between podoconiosis and two common causes of blindness and visual impairment; trachomatous trichiasis ( TT ) and cataract . TT is the blinding consequence of conjunctival scarring in trachoma , the leading infectious cause of blindness globally . Cataract is an age-related disease of the lens and remains the leading cause of visual impairment worldwide . Both podoconiosis and TT are NTDs endemic to Ethiopia and promote poverty through many factors such as disability , reduced economic productivity and stigma . This comparative cross-sectional study explored the association between podoconiosis and these two eye diseases . We found that podoconiosis patients were burdened with higher levels of blindness and low vision , had higher prevalence of TT and more severe cataract than their matched neighbourhood controls . These findings can help to direct an integrated approach to managing these two NTDs ( podoconiosis and TT ) and trigger further research in to the wider context of the double burden of eye disease and NTDs .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusions"
] |
[
"medicine",
"and",
"health",
"sciences",
"lens",
"disorders",
"tropical",
"diseases",
"geographical",
"locations",
"ocular",
"anatomy",
"parasitic",
"diseases",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"bacterial",
"diseases",
"eye",
"diseases",
"neglected",
"tropical",
"diseases",
"ethiopia",
"eyes",
"africa",
"infectious",
"diseases",
"elephantiasis",
"podoconiosis",
"head",
"people",
"and",
"places",
"anatomy",
"cornea",
"cataracts",
"ophthalmology",
"biology",
"and",
"life",
"sciences",
"ocular",
"system",
"cataract",
"surgery",
"trachoma",
"ophthalmic",
"procedures"
] |
2017
|
Podoconiosis, trachomatous trichiasis and cataract in northern Ethiopia: A comparative cross sectional study
|
Chikungunya virus is a mosquito-borne alphavirus which causes an acute febrile illness associated with polyarthralgia . Beginning in August 2013 , clinicians from the Yap State Department of Health in the Federated States of Micronesia ( FSM ) identified an unusual cluster of illness which was subsequently confirmed to be chikungunya virus disease . Chikungunya virus disease previously had not been recognized in FSM . Information from patients presenting to healthcare facilities was collected and analyzed . During August 11 , 2013 , to August 10 , 2014 , a total of 1 , 761 clinical cases were reported for an attack rate of 155 clinical cases per 1 , 000 population . Among residents of Yap Main Island , 3% were hospitalized . There were no deaths . The outbreak began on Yap Main Island and rapidly spread throughout Yap Main Island and to three neighboring islands . Chikungunya virus can cause explosive outbreaks with substantial morbidity . Given the increasing globalization of chikungunya virus , strong surveillance systems and access to laboratory testing are essential to detect outbreaks .
Chikungunya virus is a mosquito-borne alphavirus capable of causing large outbreaks of acute febrile illness with severe polyarthralgia [1 , 2] . First identified in East Africa in 1952 , as of mid-2013 chikungunya virus had caused outbreaks in Africa , Europe , Asia , and islands in the Indian and Western Pacific Oceans [1 , 2] . Chikungunya virus transmission previously had not been identified in the Federated States of Micronesia ( FSM ) . During August to October 2013 , clinicians from the Yap State Department of Health in FSM identified an unusual cluster of illness characterized by acute fever and arthralgia . Among 51 patients with samples initially sent to the U . S . Centers for Disease Control and Prevention ( CDC ) , 38 ( 75% ) had evidence of recent chikungunya virus infection by detection of ribonucleic acid ( RNA ) or anti-chikungunya virus immunoglobulin ( Ig ) M and neutralizing antibodies . Testing for other arboviruses , including dengue , Zika , Ross River , and O’nyong-nyong viruses by reverse transcription-polymerase chain reaction ( RT-PCR ) or IgM tests was negative . We describe the epidemiologic features of FSM’s first known chikungunya virus disease outbreak .
Yap State is a collection of islands in the western part of FSM in the western Pacific Ocean . It consists of Yap Main Island , which is divided into 10 municipalities , and 10 inhabited neighboring islands and atolls ( Fig 1 ) . Yap State’s 2010 census population was 11 , 376 persons , with 7 , 370 ( 65% ) on Yap Main Island [3] . Yap State has a public health care system with no private providers . We defined a clinical case of chikungunya virus disease as a Yap State resident with acute onset of fever and new onset of arthralgia or arthritis on or after August 11 , 2013 ( the onset date for the first laboratory-confirmed case ) . A laboratory-confirmed case was a clinical case with chikungunya virus RNA or IgM antibodies detected in serum . Clinical cases who resided on an island with no laboratory-confirmed cases were excluded from the analysis . A clinical case without detectable IgM antibodies in serum collected ≥8 days after illness onset was considered not to have chikungunya virus infection . Cases were identified among patients presenting to any health care facility in Yap State , including Yap Memorial Hospital on Yap Main Island or any outpatient clinic located throughout Yap State . All patients presenting with fever or new onset of arthralgia or arthritis were evaluated by a healthcare provider . For patients who met the clinical case definition , health care providers on Yap Main Island completed a case report form and on neighboring islands maintained a line list with patient demographic and symptom information . After identification of chikungunya virus as the causative agent of the outbreak , only a subset of subsequent clinical cases had serum samples tested for chikungunya virus infection . Testing of all samples was not possible for logistical reasons . Clinical cases from new geographic areas , who were hospitalized , and those aged <5 years were prioritized . In most weeks a convenience sample of clinical cases from Yap Main Island also had samples tested . Samples collected from patients beginning January 1 , 2013 , as part of routine dengue fever surveillance on Yap Main Island also were available . There samples were retrospectively tested to establish the onset date of the chikungunya virus disease outbreak as accurately as possible . Serum samples were tested for chikungunya virus RNA byRT-PCR and/or anti-chikungunya virus IgM antibodies by enzyme-linked immunosorbent assay ( ELISA ) . The test used depended on the timing of sample collection , with RT-PCR typically being used for samples collected ≤5 days after illness onset and IgM ELISA for samples collected ≥5 days after illness onset [4] . Nearly all chikungunya virus testing was performed at the CDC Arboviral Diseases Laboratory . One batch of samples was tested at the Institut Louis Malarde Laboratory in French Polynesia . A subset of clinical cases also had serum samples tested for dengue virus non-structural protein 1 ( NS1 ) antigen and IgM antibodies by the Standard Diagnostics BIOLINE Dengue Duo kit . Data were entered into a Microsoft Access database and analyzed using Microsoft Excel 2010 and Epi Info 7 . 1 . 4 . For persons who presented on two or more occasions , only data from the first visit were included . Chikungunya virus disease attack rates per 1 , 000 population were calculated using 2010 census data [3] . Hospitalization rates were calculated for cases on Yap Main Island where the only hospital was located . This assessment was judged to be non-research public health practice , and therefore it was not subject to institutional review board review requirements . For epidemiological analysis , the data were anonymized .
From August 11 , 2013 , to August 10 , 2014 , a total of 1 , 761 clinical cases were reported to the Yap State Department of Health . This represents an attack rate of 155 clinical cases per 1 , 000 persons i . e . , 15% of the population sought health care for a clinically-compatible illness . Overall , 904 ( 51% ) clinical cases were female , and the median age was 30 years ( range: 3 weeks–92 years ) . All age groups had attack rates greater than 110 per 1 , 000 population , with the highest attack rates in persons aged ≥30 years ( Table 1 ) . Among the 1 , 761 clinical cases , 1 , 412 ( 80% ) resided on Yap Main Island ( Table 2 ) . The highest attack rate of 463 per 1 , 000 population was on the neighboring island of Fais . Among residents of Yap Main Island where the hospital was located , 3% ( 45/1345 ) were hospitalized . There were no deaths . The outbreak began on Yap Main Island in the northeastern municipality of Tomil ( Fig 1 ) . The first reported case had no travel history , and investigations could not identify any case with an international travel history . From mid-August through the end of September 2013 , 34 clinical cases had been reported from six of the 10 municipalities . The outbreak peaked in late 2013 , with 1 , 514 ( 86% ) of the 1 , 761 clinical cases occurring from October through December and clinical cases reported from all Yap Main Island municipalities and the islands of Fais and Ulithi ( Fig 2 ) . Clinical case numbers declined considerably in early 2014 . However , in February 2014 cases were reported from the island of Ifalik , where transmission previously had not been documented . As of August 2014 , sporadic clinical and laboratory-confirmed cases were still being reported , but after that no further laboratory testing was conducted . Overall , 171 ( 10% ) clinical cases across Yap State had serum samples tested and 119 ( 70% ) had laboratory-confirmed chikungunya virus infection . Four ( 2% ) had equivocal or indeterminate results . Forty-four ( 26% ) had negative chikungunya virus laboratory results , but their serum samples were collected too early ( <8 days ) after illness onset to exclude infection . Only four ( 2% ) clinical cases had no detectable anti-chikungunya IgM antibodies in serum collected ≥8 days after illness onset and were thus considered not to have recent chikungunya virus infection . Two-hundred nine ( 12% ) clinical cases had dengue testing performed by NS1 antigen or IgM antibody testing . One patient with illness onset in September had dengue virus IgM antibody detected; this patient also had chikungunya virus infection confirmed by IgM antibody testing . Sixty-nine patients on Yap Main Island with dengue-like illness during January 1 through August 10 , 2013 , had samples retrospectively tested for evidence of recent chikungunya virus infection . None were positive for chikungunya virus RNA or IgM antibodies .
Beginning in August 2013 , Yap State in FSM experienced an explosive chikungunya virus disease outbreak . Initially limited to Yap Main Island , the outbreak spread to two neighboring islands within 3 months , and to a third within 6 months . By August 2014 , 15% of Yap State’s population had sought healthcare for symptoms compatible with chikungunya virus disease . We believe this was FSM’s first chikungunya virus disease outbreak . Previous dengue and Zika virus disease outbreaks in FSM were investigated without finding evidence of chikungunya virus infections in either humans or mosquitoes [5–8] . Furthermore , all age groups had relatively high clinical disease attack rates , suggesting no prior immunity to the virus among Yap State residents . Finally , there was no laboratory evidence of chikungunya virus infection in patients with a similar clinical syndrome in the 7 months preceding this outbreak . Given these findings , it is unlikely that chikungunya virus previously was circulating in FSM . It is unknown how chikungunya virus was introduced into Yap State , but it was likely imported by an infected person , as has occurred in previous chikungunya virus disease outbreaks [9 , 10] . Genetic analyses indicated the strain circulating in Yap was within the Asian genotype and closely related to strains recently isolated in China and the Philippines [11] . Travel between Yap and Southeast Asia for commerce and tourism is common , and a traveler might have been the route of introduction , with Aedes hensilli and Ae . aegypti mosquitoes in Yap State serving as disease vectors [12] . Previous studies have suggested considerable morbidity among patients affected by chikungunya virus disease . During an outbreak on Grand Comore Island , 52% ( 79/152 ) of ill persons missed work or school for a mean of 7 days [13] . Furthermore , some patients have persistent rheumatologic symptoms after chikungunya virus disease [1 , 2] . While these issues were not investigated in this outbreak , given Yap State population’s high disease attack rate , this outbreak’s impact was likely substantial . There were two main limitations to our analysis . First , cases were limited to patients who presented to medical care and were reported to the Yap State Department of Health by a healthcare provider . Because we were unable to identify sick persons who did not present for medical care , the reported attack rate is likely an underestimate of the true rate . Second , some clinical cases might not actually have had chikungunya virus infection as not all clinical cases were tested . However , there was minimal circulation of dengue virus and no circulation of other arboviruses that cause similar clinical syndromes , and our clinical case definition was similar to one shown to have 84% sensitivity and 89% specificity during a chikungunya outbreak in Mayotte [14] . Therefore , we believe most of our clinical cases were infected with chikungunya virus . Chikungunya virus is capable of causing explosive outbreaks with substantial morbidity , especially in places with certain Aedes species of mosquitos and immunologically-naïve populations . Given the increasing globalization of chikungunya virus , strong surveillance systems and access to laboratory testing are essential to detect outbreaks [15] . When outbreaks occur , swift vector control response and aggressive prevention measures ( e . g . , removing trash and water-collecting containers from yards , using air conditioning or ensuring window screens are intact , and applying mosquito repellant or wearing long sleeves and pants where feasible ) are important to limit disease spread .
|
Chikungunya virus can cause large outbreaks with substantial morbidity , especially in places with certain Aedes species of mosquitos and immunologically-naïve populations . In August 2013 , chikungunya virus disease was identified for the first time in the Federated States of Micronesia . The explosive outbreak that followed is described in this report . Given the increasing globalization of chikungunya virus , strong surveillance systems and access to laboratory testing are essential to detect outbreaks . When outbreaks occur , a swift vector control response and implementation of prevention measures are important to limit disease spread .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"chikungunya",
"infection",
"federated",
"states",
"of",
"micronesia",
"pathogens",
"immunology",
"tropical",
"diseases",
"microbiology",
"geographical",
"locations",
"alphaviruses",
"health",
"care",
"viruses",
"chikungunya",
"virus",
"rna",
"viruses",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"antibodies",
"immunologic",
"techniques",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"immunoassays",
"people",
"and",
"places",
"biochemistry",
"health",
"care",
"facilities",
"flaviviruses",
"oceania",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms"
] |
2017
|
Chikungunya virus disease outbreak in Yap State, Federated States of Micronesia
|
Synaptic structure and activity are sensitive to environmental alterations . Modulation of synaptic morphology and function is often induced by signals from glia . However , the process by which glia mediate synaptic responses to environmental perturbations such as hypoxia remains unknown . Here , we report that , in the mutant for Trachealess ( Trh ) , the Drosophila homolog for NPAS1 and NPAS3 , smaller synaptic boutons form clusters named satellite boutons appear at larval neuromuscular junctions ( NMJs ) , which is induced by the reduction of internal oxygen levels due to defective tracheal branches . Thus , the satellite bouton phenotype in the trh mutant is suppressed by hyperoxia , and recapitulated in wild-type larvae raised under hypoxia . We further show that hypoxia-inducible factor ( HIF ) -1α/Similar ( Sima ) is critical in mediating hypoxia-induced satellite bouton formation . Sima upregulates the level of the Wnt/Wingless ( Wg ) signal in glia , leading to reorganized microtubule structures within presynaptic sites . Finally , hypoxia-induced satellite boutons maintain normal synaptic transmission at the NMJs , which is crucial for coordinated larval locomotion .
Animals need oxygen and food , not only to sustain life , but also for motility . In vertebrates , oxygen and nutrients are delivered through the vascular systems to organs and tissues throughout the body . To maintain proper nutrient and oxygen supply , and thus physiological functions , the vascular system is also highly coordinated with the nervous system during development . Indeed , the vascular and nervous systems resemble each other in terms of their anatomical structures and developmental processes [1] . In the brain , nerves and vessels , form close associations and are in physical contact through the third player astrocytes to form neurovascular units ( NVU ) [2] . Such organization is essential for controlling oxygen and glucose delivery through the blood vessels by neuronal activity , and this regulatory process is mediated through the coupled astrocytes [3] . However , some invertebrates lack the complex vascular systems [4] . In nematodes , oxygen is supplied simply by ambient diffusion to inner cells [5] . Insects such as Drosophila have evolved a prototype of the tracheal system to deliver oxygen and a primitive vascular system , the dorsal vessel , to facilitate nutrient delivery [6] . However , the physical association of nerves , trachea , and glial processes has also been demonstrated at the NMJs of adult Drosophila flight muscles [7] . Animals respond to changing oxygen levels by altering their oxygen delivery system . Insufficient oxygen levels ( hypoxia ) activate a broad range of genes to re-establish body homeostasis . One crucial regulator of these hypoxia-responsive genes is the sequence-specific DNA-binding transcription factor hypoxia inducible factor 1 ( HIF-1 ) [8] . HIF-1 consists of α and β subunits that form heterodimers [9] . Whereas HIF-1β is expressed constitutively , HIF-1α protein levels are modulated by oxygen levels [10] . Under normal oxygen conditions ( normoxia ) , oxygen-dependent prolyl hydroxylases ( PHDs ) catalyze hydroxylation of a conserved prolyl residue in the central oxygen-dependent degradation ( ODD ) domain of HIF-1α [11] . Hydroxylation of HIF-1α promotes interaction with Von Hippel Lindau ( VHL ) , which is the substrate recognition subunit of the cullin2-based E3 ubiquitin ligase , leading to HIF-1α ubiquitination and proteasomal degradation . Under hypoxia , prolyl hydroxylation does not occur , HIF-1α proteins are stabilized and are translocated from the cytoplasm to the nucleus where they form heterodimers with HIF-1β to activate transcription of target genes . One major class of target genes encoding the Fibroblast Growth Factor ( FGF ) is involved in inducing angiogenesis in mammals . In Drosophila , the FGF member encoded by Branchless ( Bnl ) induces tracheal branching [12] . When oxygen levels are reduced , oxygen-starved cells express Bnl as a chemo-attractant to guide the growth tracheal terminal branches toward them [13] . In addition to adaptations of the respiratory system , the nervous system also responds to hypoxia . Oxygen levels modulate the survival , proliferation , and differentiation of radial glial cells ( RGCs ) in the human cerebral cortex . Interestingly , physiological hypoxia ( 3% O2 ) induces neurogenesis and differentiation of RGCs into glutamatergic neurons [14] . Hypoxia induces neurite outgrowth in PC12 cells through activation of A2A receptor [15] . Brief exposure to anoxia and hypoglycemia caused axonal remodeling in hippocampal neurons , including presynaptic protrusion of filopodia and formation of multi-innervated spines [16] . Under hypoxia or upon depletion of PHD2 , upregulation of the actin cross-linker Filamin A ( FLNA ) induces generation of more immature spines [17] . Astrocytes have been shown to play a crucial role in ischemic tolerance via the activation of P2X7 receptors , which trigger upregulation of HIF-1α [18] . Neuronal PAS ( NPAS ) proteins containing a DNA-binding Per-Arnt-Sim domain function in vascular and nervous system development . In mice , NPAS1 is responsible for cortical interneuron generation [19] , whereas NPAS3 is required for adult neurogenesis [20] . NPAS1 and NPAS3 also play key roles in lung development [21 , 22] . The homolog of NPAS1/3 in Drosophila , Trachealess ( Trh ) , has been well studied for its involvement in formation of the respiratory tracheal system . Trh is a master regulator of tracheal cell fates , activating gene expression to induce tracheal development [23 , 24] . However , the role of Trh in the development of other tissues , particularly the nervous system , is unknown . In this study , we found altered synaptic bouton morphology at the NMJs of trh1/trh2 mutants . By performing trh-RNAi knockdown and UAS-trh transgene rescue experiments , we show that trh is required in tracheal cells for normal bouton formation . Defective tracheal branching in the trh1/trh2 mutant mimics the effect of hypoxic conditions during larval development , and supplying higher than normal oxygen levels restored normal bouton morphology . We further show that glial cells respond to hypoxia by elevating Wnt/Wg expression to mediate synaptic bouton reorganization through HIF1-α/Sima in Drosophila . Finally , we reveal that this morphological change may be linked to normal synaptic transmission and locomotion of larvae .
To understand the possible role of Trh in synapse formation , we examined NMJ morphology in the trh mutant . Since both trh1 and trh2 loss-of-function alleles are homozygous lethal [23 , 25 , 26] , we examined the trans-heterozygous trh1/trh2 mutant that survived to adult stages and compared it to wild-type ( w1118 ) and heterozygous trh1/+ controls . Synaptic boutons of w1118 and trh1/+ NMJs were evenly spaced along the axonal terminals , displaying the typical “beads-on-a-string” pattern ( Fig 1A , upper and middle panels , enlarged images at right ) . Strikingly , the trh1/trh2 mutant larvae exhibited aberrant NMJ morphology . Small synaptic boutons formed clusters or surrounded a normal large bouton ( Fig 1A , bottom panel ) ; a phenotype described as “satellite boutons” [27] . This satellite bouton phenotype in the trh1/trh2 mutant was detected at a high frequency; 20 . 7 ± 4 . 1% ( n = 10 ) of total boutons were satellite ones , more than three-fold increases compared to 4 . 3 ± 1 . 7% ( n = 10 ) for trh1/+ and 6 . 3 ± 1 . 7% ( n = 10 ) for w1118 ( Fig 1B ) . Although the percentage of satellite boutons in the trh1/trh2 mutant was greatly increased , the total bouton number was slightly higher than that observed in controls ( 74 . 1 ± 4 . 7 , n = 10 in w1118; 89 . 6 ± 5 . 2 , n = 10 in trh1/+; and 95 . 5 ± 9 . 1 , n = 11 in trh1/trh2 , bottom panel in Fig 1B ) . Also , the muscle areas were not significantly different from each other ( S1A Fig ) . Given the small size and clustering of satellite boutons in the trh1/trh2 mutant , we examined whether these boutons express synaptic proteins normally . We found that the synaptic vesicle protein Synapsin ( Syn in Fig 1A ) was normally distributed relative to control , but the active zone protein Bruchpilot ( Brp ) was expressed at higher levels in the trh1/trh2 mutant ( S1B Fig ) . The postsynaptic glutamate receptor , as revealed by GluRIII ( S1B Fig ) and GluRIIA ( S1C Fig ) signals , as well as dPAK ( S1C Fig ) were also localized in satellite boutons , which were surrounded by the subsynaptic reticulum protein Dlg ( S1D Fig ) . Thus , although the Brp signal intensity in the trh1/trh2 mutant was stronger than in controls , the composition of synaptic proteins in satellite boutons was largely similar to that of normal-sized boutons . As trh is expressed in both tracheal and nervous systems in embryonic stages [28] , altered bouton morphology in the trh1/trh2 mutant could be due to a lack of trh in neurons , tracheal cells or other cells/tissues . Therefore , we performed trh-RNAi knockdown by using tissue-specific GAL4 drivers for trachea ( btl-GAL4 ) , neurons ( elav-GAL4 ) , glia ( repo-GAL4 ) , and muscles ( MHC-GAL4 ) . We observed a dramatic increase in satellite boutons upon tracheal trh-RNAi knockdown using btl-GAL4 ( 10 . 2 ± 1 . 7% , n = 10 ) compared to those in the trh-RNAi ( 2 . 7 ± 1 . 3% , n = 10 ) or btl-GAL4 ( 3 . 3 ± 0 . 9% , n = 12 ) control ( Fig 1C and 1D . In contrast , trh-RNAi knockdown by other tissue-specific GAL4 drivers failed to significantly increase satellite boutons , displaying similar amounts of satellite boutons to respective GAL4 drivers ( 2 . 0 ± 0 . 7% , n = 10 in elav-GAL4 v . s . 1 . 4 ± 0 . 6% , n = 10 in elav-GAL4/trh-RNAi; 2 . 8 ± 1 . 0% , n = 10 in repo-GAL4 v . s . 3 . 4 ± 0 . 9% , n = 10 in repo-GAL4/trh-RNAi; and 1 . 3 ± 0 . 8% , n = 10 in MHC-GAL4 v . s . 2 . 1 ± 1 . 0% , n = 10 in MHC-GAL4/trh-RNAi , Fig 1D ) . To further confirm the necessity of tracheal trh for normal bouton formation , we performed a rescue experiment for the trh1/trh2 phenotype . When UAS-trh expression was driven by tracheal btl-GAL4 in the trh1/trh2 mutant , the satellite bouton phenotype was suppressed ( 1 . 8 ± 1 . 4% , n = 13 , Fig 1E and 1F ) . Controls bearing only btl-GAL4 ( 14 . 3 ± 3 . 3% , n = 10 ) or UAS-trh ( 21 . 1 ± 3 . 3% , n = 10 ) still contained large amounts of satellite boutons . These results indicate that trh is required in the trachea for normal bouton formation . Apart from specifying the tracheal cell fate , Trh is also involved in the branching of tubular structures during post-embryonic stages [24] . Therefore , we examined the tracheal phenotypes in the trh1/trh2 larvae ( S2A Fig ) and observed an increase in the number of terminal branches in the dorsal branch of the third segment ( S2B Fig , 5 . 7 ± 0 . 15 , n = 10 for trh1/+ , and 7 . 5 ± 0 . 28 , n = 11 for trh1/trh2 ) . Furthermore , we identified morphological defects such as tracheal breaks and tangles , suggesting structural defects in the trh1/trh2 larvae ( arrows in S2A Fig ) . Tracheal branching activity is enhanced under hypoxia [12] . Thus , the increased terminal branches in trh1/trh2 could be a compensatory mechanism for defective trachea formation . To understand whether trh1/trh2 mutant cells are under hypoxia , we used the hypoxia biosensor GFP-ODD , in which the GFP is fused to the ODD domain of Sima , under the control of the ubiquitin-69E ( ubi ) promoter [29] . We first confirmed that GFP-ODD signal was low under normoxia ( 21% O2 ) and enhanced under hypoxia ( 5% O2 ) in wild-type late-stage embryos when tracheal tubules are already formed and functioning [29] . Indeed , enhanced GFP signal was ubiquitous under hypoxia in wild-type embryos with some pronounced focal GFP signals ( Fig 2A , upper row , and S2C Fig ) . The signal of mRFP-nls , also under the control of the ubi promoter as an internal control , remained constant under hypoxia ( Fig 2A , bottom row , and S2D Fig ) . Quantification of the GFP/RFP ratio revealed a significant difference between normoxia and hypoxia ( 0 . 18 ± 0 . 05 , n = 6 at 21% O2 and 0 . 84 ± 0 . 15 , n = 6 at 5% O2 , Fig 2B ) . We then examined whether oxygen supply is deficient in the trh1/trh2 mutant by measuring the GFP/RFP ratios . We detected a higher GFP/RFP ratio ( 0 . 82 ± 0 . 15 , n = 6 ) in the mutant compared to that in heterozygous trh1/+ ( 0 . 09 ± 0 . 02 , n = 6 ) in the 21% O2 condition , supporting that the trh1/trh2 mutant senses reduced oxygen levels internally ( Fig 2A and 2B ) . Thus , formation of satellite boutons in the trh1/trh2 mutant could be caused by hypoxia . To test this hypothesis , we reared wild-type larvae under hypoxia ( 5% ) and assessed synaptic bouton morphology at the third instar stage ( Fig 2C ) . Consistently , small clustered boutons increased , mirroring the satellite bouton phenotype ( Fig 2D , 4 . 6 ± 1 . 5% , n = 10 in 21% O2 v . s . 19 . 4 ± 1 . 2% , n = 11 in 5% O2 ) . Furthermore , when the trh1/trh2 mutant was raised in a high oxygen level ( 50% O2 ) , normal bouton morphology was restored ( Fig 2C ) ; the satellite bouton phenotype was almost completely suppressed ( Fig 2D , 15 . 2 ± 2 . 5% , n = 10 in 21% O2 v . s . 0 . 74 ± 0 . 49% , n = 10 in 50% O2 ) . These results suggest that hypoxia due to the defective tracheal system in the trh1/trh2 mutant induces the satellite bouton phenotype , and that this phenotype can be suppressed by extra oxygen supply . HIF-1α/Sima mediates the response to low oxygen supply [30 , 31] . The protein levels of Sima are increased in wild-type Drosophila embryos subjected to hypoxia [32] , leading to transcriptional activation of downstream target genes and the induction of tracheal branching [12] . We overexpressed Sima in tracheal cells , neurons , glia , or muscle cells by tissue-specific drivers to investigate which types of cells may play a role in modulating synaptic bouton formation . Overexpressing Sima in trachea caused embryonic lethality , preventing us from observing NMJ phenotypes . Larvae in which Sima was overexpressed in muscles , neurons , or glia could survive to the third instar stage , allowing us to examine the bouton phenotypes . We found that Sima overexpression in glia gave the highest number of satellite boutons ( Fig 3A , and 11 . 9 ± 2 . 3% , n = 10 for repo-GAL4/UAS-sima in Fig 3B , v . s . 2 . 8 ± 1 . 0% , n = 10 for repo-GAL4 in Fig 1D ) . However , Sima overexpression in neurons by elav-GAL4 failed to induce satellite boutons ( 2 . 6 ± 0 . 8% , n = 8 in elav-GAL4/UAS-sima in Fig 3B v . s . 2 . 0 ± 0 . 7% , n = 10 for elav-GAL4 in Fig 1D ) . Also , muscle expression of Sima also maintained a basal level of satellite boutons ( 1 . 7 ± 1 . 0% , n = 10 for MHC-GAL4/UAS-sima in Fig 3B v . s . 1 . 3 ± 0 . 8% , n = 10 for MHC-GAL4 in Fig 1D ) . This result shows that the hypoxia-responding factor Sima is capable of inducing satellite bouton formation when overexpressed in glia . The Drosophila PHD protein Fatiga ( Fga ) promotes prolyl hydroxylation and degradation of Sima in normaxia [33] . Consistently , glial knockdown by expressing fga-RNAi led to a significant increase of satellite boutons ( S3A and S3B Fig , 8 . 3 ± 1 . 3% , n = 10 for repo-GAL4/fga-RNAi ) from the control ( 4 . 3 ± 0 . 7% , n = 10 for fga-RNAi ) . Instead , tracheal knockdown by btl-GAL4 ( 3 . 8 ± 0 . 7% , n = 9 for btl-GAL4/fga-RNAi ) showed no apparent difference to the control . This result is consistent with that Sima upregulation in glia induces satellite bouton formation . If glial Sima is the factor responding to hypoxia in the trh1/trh2 mutant , reducing Sima level in glia would suppress satellite bouton formation . Accordingly , we expressed the sima-RNAi transgene , which could reduce sima expression in both transcript and protein levels ( S3C and S3D Fig ) , using repo-GAL4 in the trh1/trh2 mutant . As our prediction , the satellite bouton phenotype was suppressed upon glial sima knockdown ( 3 . 6 ± 1 . 0% , n = 10 in repo-GAL4/sima-RNAi trh1/trh2 ) , as compared to controls of the trh1/trh2 mutant carrying either the UAS-sima-RNAi transgene ( 9 . 2 ± 1 . 5% , n = 10 in sima-RNAi trh1/trh2 ) or the repo-GAL4 driver ( 12 . 2 ± 2 . 2% , n = 10 in repo-GAL4 trh1/trh2 ) that displayed high percentages of satellite boutons ( Fig 3C and 3D ) . We also tested whether low oxygen level-induced satellite bouton formation is mediated through Sima in glia . Satellite bouton phenotypes were detected in controls carrying either UAS-sima-RNAi ( 12 . 5 ± 2 . 7% , n = 11 ) or repo-GAL4 ( 16 . 4 ± 2 . 2% , n = 10 ) when raised in 5% O2 . However , almost no satellite boutons were detected in larvae carrying both repo-GAL4 and UAS-sima-RNAi ( 0 . 45 ± 0 . 45% , n = 11 in repo-GAL4/sima-RNAi ) when raised in the same condition ( Fig 3E and 3F ) . We also examined whether the Sima protein levels are changed in hypoxia or in the trh mutant . We found ubiquitous increases in the Sima levels in the wild-type control under the 5% O2 condition or in the trh1/trh2 mutant ( S3E and S3F Fig ) . The increases could be identified in glial processes along the peripheral nerves and in different subtypes of glia . Thus , glial Sima could play the role to mediate hypoxia in the trh1/trh2 mutant and in the low O2 condition to modulate synaptic bouton formation . Next , we explored possible signals transduced from glia to neurons in response to hypoxia . The glia-secreted Wingless ( Wg ) signaling molecule regulates synaptic growth at Drosophila NMJs [34 , 35] . Therefore , we examined whether Wg can be induced under hypoxia in the trh1/trh2 mutant . Wg signals were enriched around the synaptic boutons of wild-type NMJs ( Fig 4A ) . Whereas the pattern of Wg signals at trh1/+ NMJs was similar to that of w1118 , we detected much higher levels of Wg signals at the trh1/trh2 NMJ ( Fig 4A ) . Quantification of Wg immunofluorescence intensities normalized to co-stained HRP in trh1/trh2 ( Wg/HRP: 0 . 49 ± 0 . 07 , n = 9 , Fig 4B ) revealed ~3-fold increases relative to w1118 ( 0 . 18 ± 0 . 03 , n = 8 ) and trh1/+ ( 0 . 15 ± 0 . 02 , n = 9 ) . We then examined whether glial Sima is required for the enhanced Wg expression in the trh1/trh2 mutant ( Fig 4C ) . The Wg level relative to the HRP level in the trh1/trh2 mutant carrying repo-GAL4 ( 0 . 68 ± 0 . 16 , n = 8 for repo-GAL4 trh1/trh2 ) was also about 3 folds to the repo-GAL4 control ( 0 . 23 ± 0 . 04 , n = 11 , Fig 4D ) . When we reduced sima levels in the trh1/trh2 mutant by repo-GAL4-driven UAS-sima-RNAi , Wg signals were suppressed to a level equivalent to that in the repo-GAL4 control ( 0 . 25 ± 0 . 05 , n = 9 for repo-GAL4/sima-RNAi trh1/trh2 ) . Interestingly , sima-RNAi knockdown in glia of the repo-GAL4 control had no effect on the Wg level ( 0 . 25 ± 0 . 02 , n = 10 for repo-GAL4/sima-RNAi ) , suggesting that Sima is induced in the trh1/trh2 mutant to upregulate Wg expression but has no role in basal Wg expression in the wild-type . Taken together , we suggest that glial Sima is required for Wg upregulation at the NMJs of the trh1/trh2 mutant . If glia-secreted Wg is responsible for satellite bouton induction in the trh1/trh2 mutant , then glia-specific wg knockdown should phenocopy sima knockdown in glia to suppress satellite bouton formation . In control , satellite boutons were still prominent in the trh1/trh2 mutant bearing only UAS-wg-RNAi ( Fig 4E and 4F , 17 . 0 ± 2 . 8% , n = 8 for wg-RNAi trh1/trh2 ) . The satellite bouton phenotype was suppressed in the trh1/trh2 mutant bearing both repo-GAL4 and UAS-wg-RNAi ( 4 . 4 ± 1 . 4% , n = 10 for repo-GAL4/wg-RNAi trh1/trh2 ) . Also , overexpression of Sima by repo-GAL4 induced satellite bouton formation ( Fig 3B , 11 . 9 ± 2 . 3% , n = 10 for repo-GAL4/UAS-sima ) , which was suppressed by co-expressing the wg-RNAi transgene ( Fig 3B , 2 . 3 ± 1 . 3% , n = 7 for repo-GAL4/UAS-sima UAS-wg-RNAi ) . Taken together , these results strongly suggest that glia-secreted Wg mediates Sima activity in promoting satellite bouton formation in the trh mutant . As Wg signals are secreted from both glia and presynaptic neurons [34 , 36] , we found that reduction of Wg signals from neurons also suppressed satellite bouton formation in the trh1/trh2 mutant ( Fig 5A and 5B , 10 . 1 ± 1 . 5% , n = 10 for elav-GAL4 trh1/trh2; 6 . 0 ± 1 . 0% , n = 8 for elav-GAL4/wg-RNAi trh1/trh2 ) . Also , neuronal and glial but not muscle overexpression of Wg induced satellite bouton formation ( S4A and S4B Fig ) . Glial and neuronal knockdown of Wg reduced total Wg levels at the trh1/trh2 NMJ , although the reduction by neuronal knockdown was not significant ( S4C and S4D Fig ) . Thus , neuronal Wg also contributes to satellite bouton formation . However , only glial overexpression of Sima induced higher levels of Wg , but not neuronal or muscle overexpression of Sima ( S4E and S4F Fig ) . Thus , while neuronal expression of Wg contributes to the overall level at NMJs in the trh mutant , Sima-induced Wg expression is likely glial-specific . At larval NMJs , Wg signaling in pre-synaptic and post-synaptic sites through distinct signaling pathways [35–37] . However , the receptor Frizzled2 ( Fz2 ) is involved in transducing the signaling pathway activities in both sites . We tested whether pre- or post-synaptic Wg signaling is involved in satellite bouton formation by fz2 knockdown in the trh mutant . Neuronal expression of fz2-RNAi by elav-GAL4 suppressed satellite bouton formation in the trh1/trh2 mutant ( Fig 5A and 5B , 11 . 8 ± 2 . 7% , n = 9 for fz2-RNAi trh1/trh2; 10 . 1 ± 1 . 5% , n = 10 for elav-GAL4 trh1/trh2; and 2 . 5 ± 1 . 1% , n = 9 for elav-GAL4/fz2-RNAi trh1/trh2 ) . In contrast , muscle expression of fz2-RNAi by MHC-GAL4 had no significant difference to the controls ( 6 . 4 ± 1 . 6% , n = 10 for MHC-GAL4 trh1/trh2 and 7 . 3 ± 1 . 6% , n = 8 for MHC-GAL4/fz2-RNAi trh1/trh2 ) . Thus , the elevated Wg level in the trh mutant is mainly transduced in presynaptic sites to modulate satellite bouton formation . Inactivation of Wg signaling leads to a reduction of the more stabilized microtubule loops within synaptic boutons , which could be visualized by immunostaining for the microtubule-binding protein Futsch [35 , 36] . We examined whether elevated Wg signaling alters microtubule loops in presynaptic boutons of the trh1/trh2 mutant by Futsch immunostaining and found significantly more Futsch-positive loops within the boutons of the trh mutant ( Fig 5C and 5D , 2 . 92 ± 0 . 29 , n = 12 for trh1/+; and 9 . 56 ± 1 . 11 , n = 9 for trh1/trh2 ) , supporting the elevation of presynaptic Wg signaling . Glial processes invade synaptic boutons to match the growth of NMJs [38] , which intrigued us to assess whether glia in the trh1/trh2 mutant exhibits morphological change . In a live fillet preparation for imaging NMJs , we found that glial processes labeled by GFP invaded the area of synaptic boutons in the trh1/trh2 mutant , whereas glial processes were relatively restrained from the bouton areas in the control ( Fig 5E ) . Quantification of the glial process overlaying the synaptic bouton area revealed significantly greater area of overlap in the trh1/trh2 mutant relative to control ( Fig 5F , 6 . 2 ± 1 . 0% , n = 10 for trh1/+; and 16 . 7 ± 3 . 7% , n = 10 for trh1/trh2 ) . This increased extent of glial processes in the synaptic area may facilitate signal transduction from glia to synaptic boutons for structural reorganization . Taken together , these results suggest that Wg plays a prominent role in the trh1/trh2 mutant to transduce the hypoxia signal from glia to modify presynaptic bouton structure . Given the evident morphological changes at trh1/trh2 NMJs , we wondered if locomotion is affected in mutant larvae . We observed larvae crawling under free-movement conditions and found that wild-type control and trh1/+ heterozygous larvae presented smooth crawling paths ( Fig 6A ) , with an average speed of 0 . 64 ± 0 . 09 mm/s ( n = 13 ) in w1118 and 0 . 45 ± 0 . 08 mm/s ( n = 12 ) in trh1/+ ( Fig 6B ) . However , the trh1/trh2 larvae had much shorter paths and a slower speed of 0 . 14 ± 0 . 02 mm/s ( n = 14 ) . The head turning angle of the trh1/trh2 mutant was comparable to both controls , not contributing to the slow movement ( Fig 6C , 7 . 8 ± 0 . 6 degree/s , n = 13 for w1118; 9 . 7 ± 1 . 2 degree/s , n = 12 for trh1/+; and 11 . 3 ± 1 . 9 degree/s , n = 14 for trh1/trh2 ) . Larval crawling is a rhythmic behavior involving a series of periodic strides ( S1 Movie ) [39] . We noticed uncoordinated crawling in the trh1/trh2 larvae , with their posterior body segments failing to follow the rhythmic movement ( S2 Movie ) . We recorded larval forward crawling and constructed kymographs to represent the stride cycle . In wild-type larvae , normal and consistent periodic strides were apparent with regular head and tail displacements ( Fig 6D , left panel ) . Similar to the wild-type , head movements of trh1/trh2 larvae were smooth and periodic , albeit slower . However , tail movements of trh1/trh2 larvae were abrupt ( Fig 6D , right panel ) . While wild-type larvae crawled completely normal ( n = 10 ) , 69 . 2 ± 9 . 6% ( n = 10 ) of the strides of trh1/trh2 larvae were uncoordinated ( Fig 6E ) , which might contribute to the slower crawling of the trh1/trh2 mutant . This uncoordinated stride cycle prompted us to examine the bouton morphology in A2-A6 segments of the trh1/trh2 larvae ( Fig 7A and 7B ) . Strikingly , large numbers of satellite boutons were detected in segments A2 ( 18 . 0 ± 4 . 3% , n = 10 ) and A3 ( 20 . 7 ± 4 . 1% , n = 10 ) and an intermediate level of satellite boutons was detected in segment A4 ( 9 . 2 ± 2 . 6% , n = 9 ) . The numbers of satellite boutons were low in more posterior segments A5 ( 4 . 9 ± 2 . 3% , n = 10 ) and A6 ( 3 . 2 ± 1 . 7% , n = 9 ) . In wild-type larvae , all segments had relatively normal bouton morphology except the A3 segment ( 0 . 6 ± 0 . 4% , n = 10 in A2 , 6 . 3 ± 1 . 7% , n = 10 in A3 , 0 . 7 ± 0 . 4% , n = 10 in A4 , 0 . 5 ± 0 . 5% , n = 10 in A5 , and 0 . 9 ± 0 . 6% , n = 9 in A6 ) . While high numbers of satellite boutons appeared in the trh mutant , the total numbers of boutons in individual segments were comparable between wild-type and trh1/trh2 ( Fig 7C , w1118: A2 , 176 . 6 ± 15 . 0 , n = 10; A3 , 76 . 1 ± 4 . 8 , n = 9; A4 , 85 . 8 ± 4 . 1 , n = 10; A5 , 66 . 1 ± 5 . 4 , n = 10; A6 , 40 . 0 ± 3 . 7 , n = 9; trh1/trh2: A2 , 173 . 2 ± 8 . 8 , n = 10; A3 , 106 . 6 ± 15 . 3 , n = 10; A4 , 74 . 1 ± 9 . 9 , n = 9; A5 , 60 . 7 ± 9 . 1 , n = 10; A6 , 33 . 1 ± 3 . 6 , n = 9 ) . Therefore , these analyses suggest that satellite boutons are prone to appear in more anterior than posterior segments . We further examined whether glial Sima and Wg have any role on modifying bouton morphology in the posterior A6 segment of the trh1/trh2 larvae . With the satellite boutons at a basal level in the A6 segment of trh1/trh2 ( Fig 7B ) , we tested whether overexpression of Sima or Wg could induce satellite boutons in the trh1/trh2 mutant . Overexpression of Sima by repo-GAL4 in trh1/trh2 induced some satellite boutons ( S5A and S5B Fig , 6 . 8 ± 2 . 3% , n = 10 ) , which showed no significant difference to the trh1/trh2 mutant carrying repo-GAL4 ( 2 . 8 ± 1 . 5% , n = 8 ) . Overexpression of Wg by repo-GAL4 in trh1/trh2 displayed a basal level of satellite boutons ( 2 . 6 ± 1 . 5% , n = 9 ) . Also , glial wg-RNAi knockdown suppressed satellite bouton formation in the A3 segment of trh1/trh2 ( Fig 4F ) , but had no effect on the morphological phenotype in the A6 segment ( S5C and S5D Fig ) . Thus , the analysis of these data suggests that Wg and Sima might have relatively specific roles to induce satellite bouton formation in the anterior A3 segment . Given the satellite bouton phenotype in the trh1/trh2 larvae , we assessed basal synaptic transmission properties , firstly at muscle 6 of the A3 segment . The amplitude of spontaneous release , or the miniature evoked junctional potential ( mEJP ) , was slightly but non-significantly reduced ( Fig 8A ) , from 1 . 4 ± 0 . 1 mV ( n = 10 ) in w1118 to 1 . 2 ± 0 . 1 mV ( n = 13 ) in trh1/trh2 ( Fig 8C ) . Comparable frequencies were detected between w1118 ( 1 . 9 ± 0 . 1 Hz , n = 10 ) and trh1/trh2 ( 2 . 4 ± 0 . 3 Hz , n = 13 ) ( Fig 8D ) . The amplitudes of EJP were also comparable between w1118 ( 54 . 6 ± 2 . 1 mV , n = 10 ) and trh1/trh2 ( 50 . 9 ± 4 . 2 mV , n = 13 ) ( Fig 8B and 8E ) . The quantal content , calculated by dividing the EJP amplitude with that of mEJP , was slightly but non-significantly increased , from 40 . 2 ± 3 . 3 ( n = 10 ) in w1118 to 46 . 9 ± 4 . 5 ( n = 13 ) in trh1/trh2 ( Fig 8F ) . We then evaluated the synaptic transmission properties of muscle 6 for the A6 segment . Between the w1118 control and the trh1/trh2 mutant , the mEJP amplitudes ( 1 . 2 ± 0 . 1 , n = 9 v . s . 1 . 5 ± 0 . 2 mV , n = 8 ) , the mEJP frequencies ( 2 . 0 ± 0 . 3 Hz , n = 9 , v . s . 1 . 4 ± 0 . 2 Hz , n = 8 ) , and the EJP amplitudes ( 46 . 2 ± 3 . 7 mV , n = 9 v . s . 41 . 6 ± 5 . 7 mV , n = 8 ) remained similar without significant difference ( Fig 8A–8E ) . However , the slight increase in mEJP and the slight decrease in EJP in trh1/trh2 lead to a significant reduction in the quantal content , from 38 . 9 ± 3 . 7 ( n = 9 ) in w1118 to 27 . 9 ± 1 . 5 ( n = 8 ) in trh1/trh2 ( Fig 8F ) . Thus , the impaired synaptic activity of the A6 segment in the mutant larvae may underlie the defective stride cycles of the posterior segments .
Our results suggest that Trh has a late developmental role in tracheal morphogenesis , in addition to its well-characterized role in early tracheal cell fate specification [23 , 24] . We observed defective tracheal structure in the trh mutant ( S2A Fig ) , which may result in hypoxic conditions inside the larval body . The increase in terminal branch number ( S2A and S2B Fig ) may be a response to oxygen supply deficiency [12] . Moreover , the increases in Sima protein levels and ODD-GFP reporter expression indicate reduced internal oxygen levels ( Fig 2A and 2B and S3E and S3F Fig ) . Finally , the satellite bouton phenotype in the trh mutant was recapitulated by rearing larvae under hypoxia , and it was suppressed by rearing larvae under hyperoxia . Taken together , these observations suggest that cells in the trh mutant sense low oxygen levels caused by the defective tracheal system and respond by elevating Sima protein levels . It is not clear how profound this effect is for other types of larval cells . Based on our ODD-GFP and Sima immunostaining patterns ( Fig 2A and 2B and S3E and S3F Fig ) , many types of cells are likely to be affected [40] . We suggest that glia is the major cell type mediating satellite bouton formation in the trh mutant under hypoxia . While Sima was increased ubiquitously , manipulating the levels of Sima or Ftg , the negative regulator of Sima , in glia modulates satellite bouton formation ( Fig 3 and S3A and S3B Fig ) . Elevated Sima levels induce tracheal sprouting in tracheal cells , as well as protrusions in non-tracheal cells [12] . Interestingly , we also observed protrusion of glial processes into synaptic area in the trh mutant , indicative of a glial response ( Fig 5E and 5F ) . Several types of cells in Drosophila have been shown to respond to hypoxia [32 , 41] . For instance , under hypoxia , elevated Sima levels induce the expression of Breathless ( Btl , the FGF receptor ) in tracheal cells that branch out seeking cells that express Branchless ( Bnl ) /FGF , with this latter process also being partially dependent on Sima [12 , 13] . In an alternative pathway , atypical soluble guanylyl cyclases can mediate graded and immediate hypoxia responses mainly in neurons [42 , 43] . Drosophila glia have not been reported to sense and respond to hypoxia , but mammalian astrocytes in the central nervous system have been shown to be involved in these processes . In a mouse model for middle cerebral artery occlusion , astrocyte activation was shown to play a crucial role in ischemic tolerance , which is mediated through P2X7 receptor-activated HIF-1α upregulation [18] . Under physiological hypoxia , reduced mitochondrial respiration leads to the release of intracellular calcium and exocytosis of ATP-containing vesicles , thereby signaling the brainstem to modulate animal breathing [44] . Our results reveal a role for Drosophila larval glia in sensing hypoxia via the conventional HIF-1α/Sima pathway . We also demonstrate that under hypoxia , glia modulate the formation of synaptic boutons ( Fig 3E and 3F ) . These results clearly place the glia-modulated morphology of synaptic boutons in the context of hypoxia responses . Our study further establishes that in response to hypoxia , Wg is a glial signal that modulates synaptic bouton formation . Two sources of Wg , presynaptic motor neurons and glia , are involved in synaptic growth and remodeling [34 , 36] . Our results suggest that Sima upregulates the level of Wg secreted from glia to modulate synapse formation in the trh mutant or in control larvae grown under hypoxia . In hypoxic macrophages , HIF-1α ediates the induction of Wnt11 , which is a mammalian homolog of Wg [45] . It is likely wg is a direct target of Sima in Drosophila . We found the HIF1-α binding motif ( CGTG ) at the -269 nucleotide sequence in the wg promoter . Also , direct binding of Sima to this consensus site was reported in a systematic ChIP-seq experiment [46] . We further show that the level of Wg is controlled by glial Sima in the trh mutant ( Fig 4C and 4D ) and that Wg mediates Sima-induced satellite bouton formation at trh NMJs ( Fig 3B ) . Also , glial overexpression of Sima upregulates Wg levels at NMJs ( S4E and S4F Fig ) . Wg is also expressed from presynapses [35] . Thus , neuronal wg knockdown partially suppressed satellite bouton phenotypes in the trh mutant ( Fig 5A and 5B ) , while neuronal wg overexpression in wild-type induced the phenotypes ( S4A and S4B Fig ) . These results are consistent with the idea that presynaptic Wg contributes to the overall pool of Wg at NMJs of the trh mutant . As a secreted morphogen , Wg functions in both pre-synaptic and post-synaptic sites [35–37] . At presynaptic terminals , the canonical Wg pathway induces microtubule loop formation to regulate synaptogenesis . We also detected an increase in microtubule loops in the trh mutant ( Fig 5C and 5D ) , consistent with a role for Wg signaling in modulating synaptic reorganization . Postsynaptic Wg signaling leads to subsynaptic reticulum differentiation [35] , which was not apparent in the trh mutant ( S1D Fig ) , suggesting that Wg might be a component of the complex hypoxia response that induces synaptic bouton reorganization . Brief exposure to hypoxia induces immature spines and impaired synaptic function in hippocampal neurons [17] . The morphological change to satellite boutons at the A3 segment of the trh NMJs was not accompanied by altered synaptic transmission ( Fig 8 ) , which may be compensated during long-term hypoxia . The satellite boutons , also named as bunch boutons , have been described in spastin mutants [47 , 48] . As an AAA ATPase , Spastin severs microtubules to facilitate transport to distal axon segments [49] . Accordingly , the spastin mutant also exhibits a lack of microtubules at terminal boutons [48] . In contrast , the trh mutant presented an increase of the more stabilized microtubule loops ( Fig 5C and 5D ) . Microtubule loops have been linked to synaptic bouton stabilization , and an excess of microtubule loops has been associated with increased synaptic bouton formation [50 , 51] . The altered morphology of satellite boutons may be part of the structural changes necessary to maintain normal synaptic transmission under hypoxia . The trh and spastin mutants also exhibit differences in synaptic function , with loss of spastin function slightly enhancing spontaneous synaptic transmission release but reducing evoked synaptic transmission [48] . Thus , although the morphology of synaptic boutons at trh NMJs resembles that of spastin mutants , satellite boutons at trh NMJs retain synaptic functions , unlike the impaired synaptic transmission of spastin mutant boutons . The size of NMJs in muscles 6/7 decreases from the anterior to posterior segments , which could represent a coupling with muscle growth [52 , 53] , thereby maintaining similar electrophysiological efficacy at anterior and posterior NMJs ( Fig 8 ) . Interestingly , our findings show that synaptic responses in the trh mutant differ , with satellite boutons only appearing in anterior segments ( Fig 7A and 7B ) . Furthermore , synaptic transmission at trh NMJs remained normal in the anterior A3 segment but was impaired in the posterior A6 segment ( Fig 8 ) . These observations are consistent with the idea that satellite bouton formation is a part of a homeostatic response to restore synaptic activity . Why synapses are not reorganized in the posterior segments remains elusive . We failed to detect an upregulation of Wg in the A6 segment ( S4C and S4D Fig ) , and glial Sima overexpression even in the A6 segment of the trh mutant failed to increase satellite boutons significantly ( S5A and S5B Fig ) . Thus , the upregulation of Wg by Sima may be segment-dependent , which awaits further study . Motor neurons in the ventral nerve cord project much longer axons to muscles in posterior segments compared to anterior ones . It has been shown that axonal transport to posterior segments is more vulnerable to inefficient transport conditions . For example , mutation of long-chain Acyl-CoA synthetase impairs the balance between anterograde and retrograde transport , causing distally-biased axonal aggregations and affecting the growth and functioning of synapses [54] . It is possible that glia-derived Wg signals may not be efficiently transported to posterior segments during hypoxia . This polar difference in synaptic activity and bouton morphology may contribute to the uncoordinated movements of the trh mutant larvae . Alternatively , defective locomotion in posterior segments of the trh mutant is independent of the glial modulation of bouton morphological changes . Larval forward locomotion , propelled by peristaltic contraction , is controlled by different circuits targeting anterior and posterior segments . The GABAergic SEZ-LN1 neurons specifically control posterior A6 and A7 segmental muscle contraction by inhibiting A27h premotor neurons , which promotes longitudinal muscle contraction during larval forward crawling [55] . Specific alteration of the circuit in the posterior segments may lead to the locomotion defect in the trh mutant .
All flies were reared at 25°C . w1118 was used as wild-type control and to backcross with trh1 or trh2 . The sources of fly strains are as follows: trh2 , elav-GAL4 , MHC-GAL4 , repo-GAL4 , UAS-trh , UAS-sima , UAS-sima-RNAi , Nrv2-GAL4 , and His2Av-mRFP were obtained from Bloomington Drosophila Stock Center ( BDSC ) ; trh1 , NP6293-GAL4 , and UAS-wg-RNAi from Kyoto Stock Center; and UAS-trh-RNAi , UAS-fga-RNAi , and UAS-fz2-RNAi from Vienna Drosophila Resource Center ( VDRC ) . Also used were btl-GAL4[56] , moody-GAL4 [57] , alrm-GAL4[58] , UAS-wg , and GFP-ODD [29] . The repo-cyto-GFP line was generated with the sequence for cytoplasmic GFP under the control of the 4 . 3 kb repo promoter , which recapitulates the full repo expression pattern . Larvae in a food vial were transferred at 1 day after egg laying ( AEL ) to a ProOx ( model 110 , BioSpherix , Lacona , NY ) oxygen-controlled chamber . Oxygen or nitrogen was infused into the chamber to a desired concentration ( 5% or 50% ) , which was maintained until assay . The NMJ phenotypes were analyzed as previously described [59] . For live tissue preparation to detect repo-cyto-GFP expression , non-fixed larvae were dissected and the larval fillets were incubated with anti-horseradish peroxidase ( HRP , 1:10 ) in phosphate buffered saline ( PBS ) for 10 minutes . Primary antibodies used were against Synapsin ( 3C11 , mouse , 1:100; Developmental Studies Hybridoma Bank , DSHB ) , HRP-Cy5 ( rabbit , 1:100; Jackson ImmunoResearch ) , Dlg ( mouse , 1:100 , DSHB ) , GluRIIA ( mouse , 1:100 , DSHB ) , dPAK ( rabbit , 1:1000 ) , GluRIII ( rabbit , 1:1000 ) , Brp ( nc82 , mouse , 1:100 , DHSB ) , Sima ( guinea pig , 1:1000 ) [60] , Repo ( mouse , 1:1000 , DSHB ) , Wg ( 4D4 , mouse , 1:10 , DSHB ) , and Futsch ( 22C10 , mouse , 1:100 , DSHB ) . Secondary antibodies used were anti-rabbit or -mouse 488 , Cy3 , or Cy5 ( 1:1000 , Jackson ImmunoResearch ) . For Wg immunostaining , larval preparations of different genotypes were marked and immunostaining was performed in the same test tube . NMJs in muscle 6/7 of A3 segments ( or A2-A6 in Fig 7 and A6 in S4 and S5 Figs ) of wandering third-instar larvae were analyzed . Confocal images were acquired via LSM510 confocal microscopy ( Carl Zeiss ) using 40x water , 40x water immersion ( for live tissue in Fig 5E ) , or 100x oil objectives . All presented images are projections of confocal z-stacks . Numbers of satellite boutons , total boutons , and microtubule loops were counted manually . The percentage of satellite boutons was calculated as the number of satellite boutons divided by the total boutons ( satellite + normal ones ) for each NMJ , and the average percentage is calculated from about 6–13 NMJs for each genotype . The immunofluorescence intensities of Wg and HRP were analyzed by ImageJ . HRP-positive regions were chosen to measure mean intensities of Wg and HRP . After subtracting the intensity with the background one , the ratio of Wg levels to HRP levels was presented as the normalized Wg intensity . The overlapping area of GFP and HRP projections was chosen by the “AND” operator in ImageJ , which was divided by HRP area for the percentage . Each dot in the bar graph represents the data from a single NMJ of a larva , and 6–13 NMJs from 2–5 independent experiments were pooled for quantification . Embryos were acquired by means of LSM510 confocal microscopy ( Carl Zeiss ) using a 20x objective , and were analyzed as previously described [29] . For Fig 2B , each dot in the bar graph represents data from a single embryo in which fluorescence was measured in at least 35 cells . Basal transmission properties were analyzed at NMJs of muscle 6/7 in specified segments of wandering third-instar larvae as previously described [61] , with some modifications . The larval body wall was dissected in cold calcium-free HL3 solution and recorded in HL3 solution containing 0 . 4 mM CaCl2 at room temperature . Mid third instar larvae ( feeding stage ) were placed on black agar plates ( 2% agar with black food coloring in 25 × 20 cm2 dishes ) at room temperature for filming . Video recording by a Sony Xperia Z1 camera started after 1 min habituation and lasted for 5 min , and it was analyzed using Ctrax software [62] . The ( x , y ) positions were used to calculate the crawling distance between two successive frames , and crawling speed was derived by dividing total distance travelled by time . The change in angle of larvae between two frames was divided by time to represent rotational angles . The forward crawling assay was a modification of a previous study [39] . Larvae were transferred into a tunnel ( ~1 mm width ) made in 2% black agar . Specimens were video-recorded for 3–10 minutes using a Leica S8 APO microscope . Kymographs were constructed using the MultipleKymograph plug-in for ImageJ ( NIH ) . Only forward crawling was counted , and 7–10 steps for each of ten larvae were analyzed for each genotype . Statistic data were analyzed by Mann-Whitney test or one-way ANOVA with Tukey's Multiple Comparison post-test , and shown by scatter plots with bar using GraphPad Prism .
|
Oxygen is essential for animals to maintain their life such as growth , metabolism , responsiveness , and movement . It is therefore important to understand how animal cells trigger hypoxia response and adapt to hypoxia thereafter . Both mammalian vascular and insect tracheal branches are induced to enhance the oxygen delivery . However , the study of hypoxia response in the nervous system remains limited . In this study , we assess the morphology of Drosophila neuromuscular junctions ( NMJs ) , a model system to study development and function of synapses , in two hypoxia conditions , one with raising wild-type larvae in hypoxia , and the other in the trachealess ( trh ) mutant in which the trachea is defective , causing insufficient oxygen supply . Interestingly , the glial processes , normally wrapping around the axons , invade into synaptic boutons of NMJs under hypoxia . Also , the hypoxia-induced factor Sima activates the Wg signal in glia and the secreted Wg signal reorganizes the synaptic boutons as a response to hypoxia . This synaptic bouton reorganization might maintain normal synaptic transmission and locomotion of larvae .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"oxygen",
"neuroscience",
"biological",
"locomotion",
"pulmonology",
"animals",
"animal",
"models",
"organisms",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"hypoxia",
"experimental",
"organism",
"systems",
"drosophila",
"research",
"and",
"analysis",
"methods",
"medical",
"hypoxia",
"animal",
"studies",
"animal",
"cells",
"life",
"cycles",
"chemistry",
"insects",
"arthropoda",
"cellular",
"neuroscience",
"eukaryota",
"cell",
"biology",
"neurotransmission",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"crawling",
"physical",
"sciences",
"cellular",
"types",
"larvae",
"chemical",
"elements"
] |
2019
|
Glial response to hypoxia in mutants of NPAS1/3 homolog Trachealess through Wg signaling to modulate synaptic bouton organization
|
Genomic Selection ( GS ) is a new breeding method in which genome-wide markers are used to predict the breeding value of individuals in a breeding population . GS has been shown to improve breeding efficiency in dairy cattle and several crop plant species , and here we evaluate for the first time its efficacy for breeding inbred lines of rice . We performed a genome-wide association study ( GWAS ) in conjunction with five-fold GS cross-validation on a population of 363 elite breeding lines from the International Rice Research Institute's ( IRRI ) irrigated rice breeding program and herein report the GS results . The population was genotyped with 73 , 147 markers using genotyping-by-sequencing . The training population , statistical method used to build the GS model , number of markers , and trait were varied to determine their effect on prediction accuracy . For all three traits , genomic prediction models outperformed prediction based on pedigree records alone . Prediction accuracies ranged from 0 . 31 and 0 . 34 for grain yield and plant height to 0 . 63 for flowering time . Analyses using subsets of the full marker set suggest that using one marker every 0 . 2 cM is sufficient for genomic selection in this collection of rice breeding materials . RR-BLUP was the best performing statistical method for grain yield where no large effect QTL were detected by GWAS , while for flowering time , where a single very large effect QTL was detected , the non-GS multiple linear regression method outperformed GS models . For plant height , in which four mid-sized QTL were identified by GWAS , random forest produced the most consistently accurate GS models . Our results suggest that GS , informed by GWAS interpretations of genetic architecture and population structure , could become an effective tool for increasing the efficiency of rice breeding as the costs of genotyping continue to decline .
Over the next 30 years , the production of staple cereal grains including wheat , maize , and rice must to be doubled to keep pace with global population and income growth . At the same time , agriculture , in general , is imperiled by human-induced climate change , and plant breeders and farmers together must contend with increased biotic and abiotic stresses that are the direct result of climate unpredictability . Breeding rice varieties adapted to the Asian tropics is already a challenging and resource-intesive endeavor . The number of bacterial , fungal , viral and insect pests for tropical irrigated rice outnumber those for other major cereals . For non-irrigated rice , abiotic stresses such as flooding and drought also negatively affect production [1 , 2 , 3] . Rice breeders must therefore consider a large number of simple and quantitative traits in combination when developing new lines while , at the same time , maintaining and improving quality and ensuring yield improvements over existing varieties . Using coventional breeding methods , this process is extremely time consuming—on average , it takes up to ten years for elite varieties to be developed and identified [4] . The majority of public sector rice breeding programs in Asia still use conventional breeding schemes . By far , the most common way of breeding is the pedigree method , which involves visual selection and trait screening over several successive generations [2] . With advances in rice molecular genetics and genomics , however , other potentially faster breeding methods are being developed . Marker assisted selection ( MAS ) , in which a small number of molelcular markers are used to tag genes-of-interest , has been implemented for rice improvement , but its overall impact on enhancing the efficiency of breeding has been limited [5] . MAS has been successfully used in rice to incorporate major genes and/or large-effect quantitative trait loci ( QTLs ) controlling abiotic stresses such as submergence , salinity and drought into new varieties [6] . However , most traits of interest to rice breeders are not controlled by just a few large-effect genes , but by many genes of small effect and/or by a combination of major and minor genes . MAS is far less suitable for these types of trait genetic architectures , so its utility to rice breeders is limited . Epistatic interactions and the effects of genetic background in rice furthermore make molecular breeding even more complicated . Genomic selection ( GS ) , introduced in 2001 by Meuwissen and colleagues , presents a new alternative to traditional MAS that has enormous potential to actually improve gain per selection in a breeding program per unit time , and thus breeding efficiency . In a GS breeding schema , genome-wide DNA markers are used to predict which individuals in a breeding population are most valuable as parents of the next generation of offspring [7] . These estimated values , termed the genome estimated breeding values ( GEBVs ) , are output from a model of the relationship between the genome-wide markers and phenotypes of the individuals undergoing selection . The GEBVs are then used to select the best parents for making new crosses . The GS model itself is developed from a training population that resembles the population under selection ( also referred to as the testing population ) ; it is both genotyped and phenotyped , while the testing population is genotyped only . The testing population genotypes are then entered into the model to calculate the GEBVs of all the individuals in the population , even those that have not been phenotyped . Thus , the key difference between GS and traditional MAS is that genotyping is not limited to a selected set of markers that tag putative genes , but rather breeding value is predicted based on all available marker data to avoid ascertainment bias and information loss . Including all markers in the model regardless of effect size also makes it possible for the first time to track and select for small effect genes/QTL in addition to large effect genes/QTL . Statistical shrinkage , Bayesian , and/or machine learning methods are used to fit the many thousands of effects [7 , 8] . The advantage of GS over the widely-used traditional pedigree breeding method is thus one of breeding efficiency . Gain from selection during GS is proportional to GEBV accuracy . As a result , when GEBV accuracy is high enough , GS can reduce breeding time by increasing the proportion of high-performing offspring in a breeding population , thus accelerating gain from selection [9 , 10] . GS has been most successfully implemented in dairy cattle breeding , where its efficiency is proven: the replacement of progeny testing with the genotyping of young bulls has cut generation interval time in half [11] . Genetic heterogeneity is also low for Holstein-Friesian cattle , and GxG and GxE effects are limited , which makes prediction of breeding value simpler [5] . In plant breeding , these interactions present a challenge , as does the presence of structure within and between breeding populations , but GS still holds the potential to improve breeding efficiency . In temperate crops GS can accelerate gain from selection per unit time beyond that gained by the overall population improvement described above through the use of off-season nurseries [12 , 13] , while in tropical crops like rice , GS can be used with one or more cycles of rapid generation advance [14] for a similar gain . In plants most applied GS experiments to date have been in maize and small grains , and it is quickly generating interest as a breeding tool for those crops . High GEBV accuracies for grain yield and a variety of other quantitative traits have been obtained for both maize and wheat bi-parental and double haploid populations using experimental cross-validation [15 , 16] , and GS has been demonstrated to outperform marker assisted recurrent selection ( MARS ) in at least one maize breeding program [17] . Moderate cross-validation prediction accuracies have also been obtained for yield and a variety of other traits in diverse germplasm collections and breeding populations of maize , wheat , oat , and barley [18 , 19 , 20 , 21 , 22 , 23] . Preliminary genomic selection research has also been published on several other crop plants including cassava , sugarcane , and sugar beet [24 , 25 , 26 , 27] . Several recent studies in maize , however , advise caution regarding the presence of hidden or known structure or family relatedness within a breeding population or germplasm collection when estimating GS accuracy . Windhausen et al ( 2012 ) found that within a diversity panel of 255 maize lines from eight distinct breeding populations any predictive ability in the dataset was a byproduct of the population structure , while Riedelsheimer et al ( 2013 ) found mean accuracies of 0% when trying to predict individuals in biparental families using data trained on the progeny of an unrelated cross [28 , 29] . To accurately predict the phenotypes of individuals in their biparental crosses , Riedelsheimer et al found it necessary to train the model using full sibs of the validation individuals , or half sibs representing both parents of the cross [28] . These limitations will likely also apply to rice , which is subject to deep population structure and is often bred in large , inter-related pedigree schemas . Rice is also frequently admixed , and many varieties contain introgressions from different subpopulations [30 , 31 , 32] . The work of Guo et al . , ( 2014 ) evidences this need to control for subpopulation structure when performing GS in rice . In a rice diversity panel , Guo et al . ( 2014 ) found that ∼33% of the genomic heritability was explained by subpopulation structure , and that controlling for subpopulation structure when performing cross-validation significantly decreased prediction accuracy . When prediction was performed within a specific subpopulation , however , structure was found to have little effect on prediction accuracy [33] . This is fortunate for breeding programs , which generally work within a particular subpopulation , although introgressions are frequent . Genetic architecture must also be taken into account when considering the implementation of genomic selection . GWAS results in maize have consistently found most agronomic traits to be controlled by many genes of small effect [34 , 35 , 36 , 37] . In rice , by contrast , many GWAS and QTL mapping studies have found large effect QTLs for agronomic traits , including grain yield , flowering time , plant height , aluminum tolerance , grain yield under drought stress , and submergence tolerance [38 , 39 , 40 , 41 , 42 , 43 , 44] . The difference in the genetic architecture between maize and rice , as well as the difference in the genetic architecture of different rice traits , could be expected to affect the relative efficacy of different genomic selection statistical methods . To the best of our knowledge , no research on performing GS in a rice breeding population has yet been published . Here we report the results of performing GWAS and GS cross-validation using data on a collection of 363 elite breeding lines from the International Rice Research Institute's ( IRRI's ) irrigated rice breeding program . To assess GS accuracy , we performed five-fold cross validation to predict grain yield , flowering time , and plant height in the 2012 wet and dry season in Los Baños , Philippines and compared our prediction results using GS to those using only pedigree information as well as a traditional MAS model . For the GS models , the training population composition , marker number , and the statistical method for the calculation of GEBVs were varied to determine their effect on rice GS accuracy . Finally the GWAS results published in a companion paper allowed us to analyze the effect of genetic architecture on GS prediction accuracy [45] .
384-plex Genotyping-by-sequencing ( GBS ) was used to discover and call SNPs on 369 advanced inbred breeding lines from IRRI's irrigated rice breeding program ( methods ) . SNP calling was performed using the TASSEL3 . 0 GBS pipeline with physical alignment to the MSU v6 . 0 Nipponbare rice reference genome using Bowtie2 [46 , 47] . The resulting SNP data were imputed using the TASSEL 3 . 0 FastImputationBitFixedWindow plugin [48] . After imputation , SNPs with call rates < 90% were removed along with monomorphic markers to obtain a filtered SNP dataset containing 73 , 147 SNPs . Individuals with missing data > = 60% , a total of six individuals , were dropped for a total of 363 genotyped lines ( materials and methods ) . The majority of the 363 lines were known a priori from breeding records to belong to the indica or indica-admixed subpopulation groups . In order to identify outlier individuals belonging to the japonica or japonica-admixed groups , principle components analysis ( PCA ) was performed using the 73 , 147 SNPs . Thirty one outliers were identified and excluded based on this analysis ( S1 Fig . ) . After removing these 31 outliers , the resulting PCA suggested no remaining subpopulation stratification within the dataset . Family structure , however , was presumed to still exist . As the presence of close relatives ( e . g . full sibs ) across training and testing folds in a cross-validation experiment can artificially inflate prediction accuracies , it was necessary to also control for this family structure . To do so , the remaining 332 lines were clustered using a partitioning around k-medoids algorithm ( PAMK ) based on the genotype matrix . k = 87 was found to be the most statistically favorable number of clusters in the dataset based on average silhouette width ( S2 Fig . ) . Individuals in the same cluster ( of 87 ) were then assigned to the same fold of 5 to form the five folds used for cross-validation . The most closely related individuals were thus placed within the same fold , making it impossible for them to be spread across training and testing groups [26] ( materials and methods ) . Four years of grain yield ( kg/ha ) , flowering time ( days to 50% flowering ) , and plant height ( cm ) data and related phenotypic covariates were curated from historical breeding trial records taken at a single location in Los Baños , Philippines for years 2009–2012 , two seasons per year , dry and wet , with the exception of plant height in the 2009 wet season , which was not available ( materials and methods ) . As the genotyped lines represent a subset of a working breeding program , substantial missing data are present in years 2009–2010 for all traits ( S1 Table ) . Such an unbalanced design is typical of breeding trial data and to be expected in the practical implementation of GS . Correlations among years/seasons were calculated for all three traits using the trait least squares means . For grain yield , the 2011 and 2012 data were more tightly correlated than the earlier year data . Flowering time and plant height data was well correlated for all four years and seasons ( S3 Fig . ) ( see methods ) . Narrow-sense heritabilities were calculated on a per line basis for each trait for both validation seasons—the 2012 dry season ( 2012 DS ) and the 2012 wet season ( 2012 WS ) and ranged from 0 . 31–0 . 32 for grain yield , 0 . 30–0 . 35 for plant height , and 0 . 32–0 . 44 for flowering time ( Table 1 ) ( materials and methods ) . Heritabilities for all three traits were slightly higher in the dry season than the wet season . Five-fold cross validation was performed using the full set of 73 , 147 markers to predict grain yield , flowering time , and plant height in the 2012 dry and wet seasons . The year and season data included in the training population were varied to determine which combinations of years/seasons were the most predictive of the 2012 dry and wet season ( total of twelve different combinations ) . The GEBV accuracies were calculated as the correlation of predicted GEBV and observed phenotypes in the validation population . Six statistical methods widely demonstrated to produce accurate genomics-assisted breeding models in a variety of crops were selected from the literature to test using our rice data . The selected methodologies were chosen to represent the variety of available approaches , and included one linear , parametric , and frequentist method: rrBLUP , one linear , parametric , and Bayesian method: Bayesian LASSO ( BL ) , one non-linear semi-parametric method: Reproducing Kernel Hilbert Spaces ( RKHS ) , and one non-linear machine learning method: Random Forest ( RF ) [19 , 23 , 49 , 50] . Multiple Linear Regression ( MLR ) , in which a subset of significant markers are chosen to fit a linear model , has been shown to be effective for traits with a very simple genetic architecture , and served as our non-GS method control [51] . Finally , kinship BLUP was used to predict GEBV based on the pedigree A-matrix alone ( ped ) ( methods ) [52] . We estimated accuracies using three experiment types ( CV1 , CV2 , and CV3 ) . CV1 accuracies were calculated using training populations that included data from the validation year/season , i . e , if the validation population consisted of the 2012 dry season , then data on individuals from the 2012 dry season were included in the training population , excluding data on any individuals in the validation fold . However , this is likely to upwardly bias accuracy estimates by confounding GxE and line effects [21] , so we worked to obtain an estimate of this bias by performing two other types of experiments . For CV2 accuracies we excluded the validation year/season from the training population . By removing these data from the training population , however , we introduce a different confounding factor to our accuracy estimate—a smaller training population size . We therefore performed cross validation experiment 3 ( CV3 ) in which the data from the validation year/season were retained in the training population , but the equivalent data from the respective 2011 season were not included in the training population . The overall estimate of bias for a given permutation was subsequently estimated as accuracy of CV3—accuracy of CV2 [26] ( materials and methods ) . The bias estimates were found to be very small and consistent for all tested traits and permutations ( Table 2 , S2–S4 Tables ) . It can thus be concluded that for the population and statistical methods tested here bias as a result of including data from the validation year/season in the training population is not a significant concern . Grain yield . The highest prediction accuracies for grain yield in both the 2012 dry and wet seasons were 0 . 31 , when the training populations consisted of data from all four years ( 2009–2012 ) , both seasons per year . The peak dry season accuracy was obtained when rrBLUP was used to build the model , and the peak wet season accuracy was obtained when RF was used ( Table 2 , S2 Table ) . In general , however , prediction accuracies did not significantly vary depending on the combination of years or seasons in the training population ( α = 0 . 05 ) . These results indicate that the most recent and complete years ( 2011 , 2012 ) are also the most predictive , but that adding data from earlier years to the training population and utilizing both seasons of data ( as opposed to using only the dry season to predict the dry season or only the wet seasons to predict the wet season ) can marginally increase accuracy ( Table 2 , S2 Table ) . These results make sense given the strong correlations between the wet and dry seasons within the same year and the weak correlations between the earlier and later years for grain yield ( S3 Fig . ) . The lower relative importance of the earlier year data could also be due to the large proportion of missing data in the earlier years . The statistical method used to build the prediction model had a significant effect on accuracy . RR-BLUP , Random Forest , and RKHS all performed significantly better than pedigree alone . RR-BLUP and RF , specifically , outperformed pedigree prediction by an average of ∼8% . Similar results have been documented in CIMMYT wheat populations where genetic markers have been found to add 7 . 7%-35 . 7% to the accuracy of grain yield predictions over a pedigree-only model depending on the population and environment [52] . The modest gains in accuracy of using markers to predict breeding value in our rice population suggest that larger training populations may be necessary to better model the effects of Mendelian segregation on yield , in addition to effects due to family relationships [53] . Some of the marker models performed worse than pedigree prediction . Bayesian LASSO performed significantly worse than prediction based on pedigree alone , while MLR performed worst of all . It is worth noting that the GWAS for grain yield in this population ( unlike the GWAS for flowering time or plant height ) did not identify any large effect QTL [45] , which could explain why choosing a subset of markers to predict GEBV performed so poorly relative to the genomic selection methods ( Table 2 , S2 Table ) . Flowering time . The prediction accuracies for flowering time were higher than those for grain yield at 0 . 63 and 0 . 54 for the best performing experiments in the 2012 dry and wet seasons , respectively . For the dry season , the most predictive training population was composed of the 2009–2011 data , dry seasons only , while for the wet season , the best training population included all seasons from 2010–2011 . The prediction accuracies for flowering time in the 2012 dry season were significantly higher than those for the 2012 wet season across statistical methods and experiments ( p < 0 . 0001 ) , but the differences in the performance of different training populations were not significant within a given validation population ( Table 2 , S3 Table ) . Unlike for grain yield , the best accuracies for predicting flowering time for both seasons were obtained using MLR . In fact , MLR significantly outperformed all other statistical methods and was more accurate than pedigree alone by 22% and 33% for the dry and wet seasons , respectively ( Table 2 , S3 Table ) . The higher accuracies for prediction of flowering time relative to predictions for yield , and also of the dry season predictions over the wet season predictions , can be explained by the higher trait heritabilities for flowering time of the 2012 dry season relative to the 2012 wet season ( Table 1 ) , and by the strong correlation in the phenotype data for all years and seasons ( S3 Fig . ) . The outstanding performance of MLR , on the other hand , is best explained by the genetic architecture of flowering time . Multiple large effect QTL have been cloned for flowering time [43 , 44] , and the GWAS performed on this population identified a single very large effect QTL on chromosome 3 that explained more than 40% of the variation in flowering time [45] . These results are also consistent with results for prediction of heading date using MLR versus GS in wheat [51] . Of the genomic selection methods tested ( MLR is a non-GS method ) , random forest performed the best by a significant margin , and was the next best method of predicting flowering time after MLR . This is worth noting as the random forest algorithm is also effective at capturing large-effect QTL [54] . Overall , these results suggest that the presence of large effect QTL for specific traits in rice could improve the prediction accuracy of those traits , although it remains to be seen whether genomic selection models will be the most practical means of obtaining those predictions . One promising avenue of research would be to model the large effect QTL as fixed effects using a genomic selection method such as rrBLUP [55] . Plant height . The accuracies for plant height were similar to those for grain yield , 0 . 34 for the dry season when the 2009–2011 dry seasons served as the training population , and 0 . 32 for the wet season when all seasons and years served as the training population ( Table 2 , S4 Table ) . These results further suggest that heritability has an important effect on accuracy . Both grain yield and plant height had similar heritabilities , and similar prediction accuracies ( Tables 1 , 2 , S4 Table . ) For predicting plant height , however , MLR was sometimes the best-performing statistical method , as was the case for the most accurate wet season experiment , described above , but for other experiments , MLR was the worst-performing method , as for the best performing dry season experiment , described above . Due to the inconsistent performance of MLR , the prediction method with the best performance over all experiments was random forest ( Table 2 , S4 Table ) . Across all experiments , random forest outperformed pedigree prediction by an average of 13 . 3% , an improvement in the performance of marker based prediction relative to pedigree prediction that is squarely in between the improvements seen for grain yield and plant height ( Table 2 , S4 Table ) . These results suggest that large marker effects help to make up the genetic architecture for plant height , but that plant height genetic architecture is more complicated than the genetic architecture of flowering time . This inference is borne out by the GWAS results for plant height—four large effect QTL were identified , explaining ∼74% of the total variation [45] . While these effects are large , they are not as dramatic as the one super-QTL found for flowering time on chromosome three , which may explain the difference in the performance of MLR for the two traits . As for flowering time , future research in predicting plant height could explore fitting these QTL in linear models as fixed effects . In order to determine the necessary number of markers for performing GS in a rice population of this type , we selected differently sized SNP subsets from the 73 , 147 SNP set . The subsets were selected in two ways: 1 . to be evenly distributed across the genome ( see materials and methods for details ) or 2 . at random . Ten selections were made for each subset size and type ( i . e . random versus distributed ) , and five-fold CV was performed using each selection in combination with all five marker based models ( materials and methods ) . For each trait , cross validation was run for both validation populations , with years 2009–2011 , both seasons per year , serving as the training population . ( Fig . 1 , S4 Fig . , S5 Table , S6 Table ) . For all three traits and both validation seasons , it is clear from the marker subset results that 73 , 147 markers is more than is necessary to capture the QTL segregating in this population . For almost all traits , there was no significant difference in the best-performing GS method for a given trait or validation season when 7 , 142 SNPs ( approximately 1 SNP for every 0 . 2 cM ) were used versus when 13 , 101 SNPs ( 1 SNP for every 0 . 1 cM ) or the full 73 , 147 SNPs were used . This was true for the randomly chosen SNPs as well as for the evenly distributed SNPs , however the accuracy variances were higher for the randomly chosen SNPs , so it is our recommendation that SNPs be evenly distributed across the genome when possible ( Fig . 1 , S4 Fig . , S5 Table , S6 Table ) . Although it is possible that the variation in the call rates and minor allele frequencies of the randomly selected SNPs also contributed to the larger variations in accuracy in the random SNP subsets , it is still thought that the position of the SNPs was the most important contributor to prediction accuracy . For all three traits and both validation seasons , prediction accuracies dropped significantly faster with decreasing numbers of markers when the markers were chosen at random versus when they were evenly distributed throughout the genome . The drop-off in prediction accuracy when random selections of SNPs were used was particularly acute for flowering time and plant height and is attributable to the presence of large-effect QTL for these traits; As the number of randomly chosen SNPs decreases , the odds of capturing the effect of any one QTL also decreases . The prediction results for grain yield , by contrast , did not differ as dramatically between the randomly and evenly distributed subsets as did those for flowering time and plant height . These results suggest that the genetic architecture for grain yield is more in line with an infinitesimal model , i . e . , that there are many small effect QTL throughout the genome , and are in agreement with the grain yield GWAS results [45] . It thus follows that the effect of choosing SNPs at random would not be as detrimental for grain yield as it is for flowering time or plant height when accuracy crucially depends on capturing specific regions that explain a high proportion of the phenotypic variance . At fewer than 7 , 142 SNPs , accuracies began to decrease for most traits and statistical methods , although the extent to which accuracies decayed depended on the prediction method used , the trait , and the validation season . For grain yield in the 2012 dry season , for example , there was no significant difference in the performance of rrBLUP at any marker set > = 3076 markers . For random forest , however , there was no significant difference in prediction accuracy all the way down to sets of markers > = 316 ( random or distributed ) . While it would not be advisable to use such a small number of markers , as the smaller the number of markers , the larger the variation in prediction accuracy , these results do suggest that for grain yield , at least , random forest works better with smaller numbers of markers than does rrBLUP . The results for plant height were very similar to those for grain yield . For flowering time , when SNPs were evenly distributed , variances in accuracy were very small , again , most likely as a result of the super-QTL on chromosome three . These very small variances meant that for both MLR and random forest , accuracies were significantly lower for fewer than 7142 SNPs ( distributed ) or 1553 SNPs ( random ) . Taken collectively , these results suggest that using ∼1 SNP every 0 . 2 cM ( ∼6–7K SNPs ) , could be ideal for performing genomic selection in inbred rice breeding populations like the one at IRRI . Opportunely , two Infinium 6K SNP fixed arrays have recently been developed for use within specific rice breeding/research programs [56] . Fixed arrays have established advantages in rice , including robust allele calling , cost-effectiveness per data point , and speed of genotyping turn-around [56] . 6–12K fixed arrays could thus prove to be the most affordable and efficient means of genotyping for GS in rice , especially for smaller breeding programs with less genotyping informatics expertise . The best strategy , however , will likely be to have multiple genotyping platforms available and the flexibility to switch between them as needed . Genotyping turn-around time is ultimately key for GS because genotypes must be available in time for selections and the next generation of crossing . It should be noted that depending on the platform , genotyping individuals with more markers than is necessary could be detrimental to breeding progress if it overloads the bioinformatics and computational capacities of a breeding program . The matrix of genotypes and phenotypes on a breeding population provides the opportunity to perform GWAS in addition to testing any GS models that are available . This paper describes the GS-side of a joint GS-GWAS project on a single rice breeding population , and is the first study to suggest that GWAS on a set of breeding lines might provide information about both the genetic architecture of the traits-of-interest and the population structure of the breeding materials . Specifically , our results on performing GS for grain yield , plant height , and flowering time demonstrate that performing GWAS using the inputs to GS can reveal the presence of large-effect QTL segregating in a breeding population , which can then be modeled accurately using GS . Our results are promising for the implementation of GS in rice improvement . For all traits tested , GS outperformed prediction based on pedigree alone with the use of a reasonable number of markers ( ∼7000 ) suggesting that genomic selection is accessible for moderately-resourced public programs with minimal bioinformatics capacities . For yield , which appears to be controlled by many genes of small effect [45] , RR-BLUP was the most computationally efficient of the best performing statistical methods . For plant height and flowering time , however , the highest accuracies were obtained using random forest and/or MLR , which suggests the presence of both large and small effect QTL for these traits , a hypothesis that is also supported by the GWAS results [45] . Currently , the most commonly used methods of rice improvement are pedigree breeding and traditional marker assisted selection , which mainly track large effect QTL . Our results suggest that genomic selection will make it possible for the first time to track , accumulate , and select for small effect QTL using genetic markers in addition to large effect QTL . One promising strategy is to build GS models in which large effect QTL are fit as fixed effects to capture the variance of large-effect QTL along with small effect QTLs located throughout the genome [55] . Future experiments in rice genomic selection should focus on building these models . While genomic selection has yet to be integrated into applied breeding programs in rice as it has in maize and wheat , it would be feasible to undertake small pilot programs within specific rice breeding programs , especially for irrigated rice where growing environments are generally more uniform . Such pilot programs are needed , in particular , to determine when and how to incorporate genomic selection into existing breeding programs . An example of an irrigated rice breeding pipeline that incorporates genomic selection is presented in Fig . 2 . Parents are selected and crossed and the resulting F1 progeny fixed over seven generations with selection of families for heritable traits . Traditionally , selection during pedigree line fixation would be based only on phenotype . Here , we propose incorporating selection based on GEBV at least once during fixation , as resources allow . Early generation GEBV-based selection would help to avoid eliminating families that carry beneficial rare or recessive alleles and would increase the proportion of top performers that are advanced to the observational yield trials ( OYT ) . Late-generation selection based on GEBVs could be used to select fixed lines to advance to the OYT . The top lines advanced to the OYT based on GEBV could be used simultaneously as parents of the next generation of breeding ( Fig . 2 ) . From the OYT , the best performing lines could be identified and advanced to the replicated yield trials ( RYT ) by a combination of phenotypic and genomic selection . Phenotypic selection by the breeder has the potential to compensate for the fact that the GS model is always a generation or more behind the current breeding population . This means that any favorable new GxG interactions will not be captured by the model and cannot be selected by GEBV alone . In species where the majority of the genetic variation under selection is controlled by many additive , small effect loci , this should not be a problem . However , in rice and other inbreeding crops , the genetic architecture of many important agronomic traits contains important non-additive features and transgressive variation is common [41 , 43 , 57 , 58 , 59 , 60] . The selected lines from the RYT are subsequently advanced to the multi-environment trials ( MET ) where the GEBVs can be used to select parents for the next generation of hybridization . In order to build or update the genomic selection model at any stage of selection , a training set consisting of a fraction of the breeding population ( ∼300 individuals ) representing different families would need to be both phenotyped and genotyped . The rest of the lines would be genotyped only to calculate the GEBVs ( Fig . 2 ) . The above genomic selection models would ideally account for multiple environments and GxE interactions , however current programs such as the one at IRRI and many other national breeding institutes do not make use of multi-environment data until very late stages of the breeding process , after the population has already been reduced to a manageable number of lines . Thus , even GS models that do not account for multiple environments , like those presented here , are of use to plant breeders and have the potential to improve breeding outcomes . The data from the Multi-environment trials on the IRRI breeding lines used in this experiment is currently being accumulated and vetted and will be a subject for future GS research . In order to fully exploit the benefits of GS , however , new rice breeding schemes will need to be implemented to further reduce the breeding cycle and increase genetic gain . Heffner et al . ( 2010 ) proposed a GS scheme for winter wheat using rapid generation advance ( RGA ) to generate F5 lines , and multi-location field trials to test F5-derived material , which was further used to train the GS model [13] . A similar scheme should also be effective for rice and a modified scheme is currently being implemented at IRRI within the irrigated breeding program . By using the genotype and phenotype inputs from pilot programs for both GS and GWAS , the accuracy of GS models could be improved while , at the same time , helping to answer basic biological questions about the genes underlying agronomic traits of interest . Ultimately , in order for genomic selection to be of practical use , it must be possible to select lines with combinations of phenotypes that are routinely measured by breeders , such as disease and insect resistance and grain quality . GEBVs could be used to select for traits that are either difficult or expensive to phenotype or are late in development ( e . g . panicle or post-harvest traits ) , while phenotypic selection , such as in the OYT and RYT in Fig . 2 , could be used for other important variety parameters . The use of multi-variate GS models or selection indices as GS phenotypes are other potential solutions to this problem , but both require additional research and computational/statistical inputs to implement . In practice , determination of whether GS can cost-effectively increase genetic gains relative to utilizing pedigree data alone or simply phenotyping more lines requires a careful consideration of the relative cost of phenotyping compared to genotyping plus line development [61] . For GS to provide increased genetic gain in a pedigree breeding program , the prediction approach must either increase accuracy relative to phenotyping or permit a substantial increase in selection intensity . It is possible to increase selection intensity through the use of rapid generation advance , as mentioned above , but the selection intensity increase must be very large because the response of genetic gains to increasing selection is logarithmic rather than linear . In this study , GEBV accuracy for yield averaged about 0 . 3 for the most effective prediction methods ( Table 2 ) . The corresponding accuracy for phenotypic selection is the square root of heritability , or about 0 . 55 for evaluation in a single three-replicate trial ( Table 1 ) . An accuracy of 0 . 3 corresponds to a heritability for yield evaluation of 0 . 1 , which is roughly the accuracy achievable by screening for yield in a single unreplicated irrigated rice trial at IRRI ( e . g . Bernier et al . , 2007 ) . Currently , the cost of phenotyping a single rice plot for yield and genotyping via GBS is roughly equivalent ( $20-$30 ) , so there is no clear advantage for GS over simply phenotyping more materials in unreplicated trials . However , genotyping costs are likely to continue to drop , whereas phenotyping costs are generally steady or rising . Furthermore , continued refinement of GS models by incorporating fixed effects and accumulation of high quality data over years and environments is expected to increase GEBV accuracy . As a result , we predict that in the near future , GS will become a cost-effective means of performing line selection in rice .
369 elite breeding lines were selected for genotyping from the International Rice Research Institute ( IRRI ) irrigated rice breeding program based on the planned inclusion of the lines in the 2011 Multi-Environment Testing Program and presence in the 2011 and 2012 Replicated Yield Trials ( RYT ) at IRRI ( Los Baños ) . Approximately half of the lines were also included in the 2009–2010 RYTs at IRRI ( S1 Table ) . The other lines were promoted from the observational yield trial ( OYT ) to the RYT in 2011 . Phenotypes for the replicated yield trials ( RYT ) were used for all the experiments and curated from the IRRI database for years 2009–2012 , including wet and dry seasons each year . All of the RYT breeding lines , of which our selected 369 lines are a subset , were grown in a randomized complete block design with three replicates in the same field location at IRRI every season and year . The following data were curated for each year , with the exception that plant height data was not available for the 2009 wet season: plant height: the actual measurement in cm from soil surface to tip of tallest panicle ( awns excluded ) flowering time: days to when 50% of flowers were visible in whole plot maturity date: days to when 85% of grains on panicle were mature number of effective tiller or panicle per plant: count of the number of panicles on each plant lodging score: percent of plants that lodged grain yield ( kg/ha ) : grain yield from a representative plot was harvested and weighed , from this sample the grain yield per hectare was calculated from an inner harvested area of the plot excluding border rows rep: replication number of observation The plant height , flowering time , and grain yield phenotypes were selected for prediction using the genomic selection models . DNA extraction . Young leaf tissue was collected from each of the 369 breeding lines from plants grown in Gutterman Greenhouse in Ithaca , NY . DNA was extracted using the Qiagen 96-plex DNeasy kit as per the Qiagen fresh leaf tissue 96-plex protocol ( www . qiagen . com/HB/DNeasy96Plant ) . Library preparation . 384-plex genotyping-by-sequencing ( GBS ) libraries were prepared using the protocol by Elshire et al . 2011 [62] , as described previously in Spindel and Wright et al 2013 [63] . GBS data analysis . SNPs were discovered and called from the raw 384-plex GBS data using the TASSEL3 . 0 GBS pipeline with physical alignment to the MSU version 6 . 0 Nipponbare rice reference genome using Bowtie2 , as described in Spindel and Wright et al 2013 [47 , 63 , 64] ( S5 Fig . ) . The IRRI breeding materials genotyped here are a collection of multi-parent related and unrelated inbred lines , so the GBS-PLAID algorithm for imputation , which was developed specifically for imputation of biparental rice mapping populations , was not useful [63] . Imputation of missing data was instead performed using the TASSEL3 . 0 FastImputationBitFixedWindow plugin with default settings [48] . The algorithm works by dividing the entire SNP dataset into small SNP windows , then identifying the most similar inbred line within each window to fill the missing data . The algorithm takes advantage of small IBD regions shared between pairs of inbred lines in the collection; if the window from the closest neighbor has more than 5% difference from the line being imputed , the data point is left as missing [48] . The imputation error rate using this algorithm was estimated for each chromosome in our dataset by masking a fraction of the un-imputed allele calls and comparing the imputed and actual calls . The average imputation error rate across the twelve rice chromosomes was estimated in this way to be less than 1% . SNPs that still had 10% or more data missing after imputation ( or call rates of < 90% ) were removed from the dataset along with all monomorphic SNPs , for a total SNP set of 73 , 147 SNPs . After the SNP filtering described above , individuals with more than 60% missing data were dropped from the dataset , which resulted in the removal of six individuals that failed sequencing for the total of 363 genotyped lines used throughout the study ( S5 Fig . ) . The final dataset was then transformed from nucleotide genotype coding ( i . e . , 'A' , 'C' , 'T' , 'G' ) to numeric coding ( 1 , 0 , -1 for class I homozygotes , heterozygotes , and class II homozygotes , respectively ) to facilitate statistical analysis . The minimal remaining missing data were filled using the numeric genotype means of each line in order to perform PCA and genomic selection modeling ( S5 Fig . ) . The majority of the 363 lines were characterized a priori from pedigree records to belong to the indica or indica-admixed subpopulation groups . In order to identify outlier individuals belonging to the japonica or japonica-admixed groups , principle components analysis ( PCA ) was performed in R ( version 3 . 0 . 1 ) using the imputed 73 , 147 SNPs , with remaining missing data filled using the line means . The first principal component of high density SNP data in rice can separate the indica and japonica subgroups [30] , so by plotting the first four principal components using JMP Pro 10 , 13 japonica outliers were identified as a tight cluster that was pulled apart from the rest of the 350 lines ( S1A Fig . ) . These 13 lines were removed from the dataset , and a second PCA was performed using the same methodology as the first to identify any admixed outliers , i . e , outlier lines containing greater percentages of japonica derived SNPs . By plotting the first four principal components of the second PCA , another 18 lines were judged on a visual basis to be outliers and removed from the dataset , leaving a total of 332 lines to be used for the cross-validation experiments ( S1B Fig . ) . A third PCA was performed using the remaining 332 to confirm that there were no additional subpopulation outliers . It was also known from studying the breeding program pedigrees that differing degrees of family relatedness existed within the remaining 332 lines , including half sibs , full sibs , parents and offspring , and unrelated lines . The presence of highly related individuals in the dataset could have the effect of artificially inflating prediction accuracy if the most closely related individuals are randomly assigned to different folds , and one of those folds is then used as training , while the other is used as testing . Or , in other words , the training fold could end up as unusually predictive of the testing fold if , for example , a pair of full sibs is split across training and testing folds . To control for this possibility when designing our folds , we performed a partitioning around k-medoids analysis ( pamk ) using the R fpc package ( function pamk ) with the 73 , 147 SNPs . k values from 2 to 332 were tested to determine the most statistically probable k-value by average silhouette width ( S2 Fig . ) . The largest average silhouette width was found to occur at k = 87 ( S2A Fig . ) . Individuals found within same cluster of 87 were then assigned to the same fold , making it impossible for the most closely related individuals to be split across training and testing folds . Full clusters were assigned to one of five folds randomly , controlling only for cluster size in order to produce three folds of 66 individuals and two folds of 67 individuals . A similar procedure was used by Ly et al . , 2013 [26] . For each cross validation experiment , one of the five folds served as the validation fold , and the other four folds served as the training folds . The process was repeated five times so that each fold served once as the validation fold , resulting in predicted GEBV values for all individuals . Accuracy was assessed as the mean Pearson Correlation of the predicted GEBV and observed phenotype in the validation population . The cross validation experiments shown in Table 3 were performed in order to test all logical combinations of years and seasons in the training and validation populations . Note that a year's wet season was never used to predict the same year's dry season because in Southeast Asia , the dry season arrives first chronologically . We did , however , predict the 2012 wet season both with and without the 2012 dry season present in the training population . We tested scenarios in which both seasons per year were included in the training population as well as scenarios where only the data from the seasons matching the validation population were included in the training data ( e . g . , using only the wet season data to predict the wet season ) . We also sought to test scenarios using only more recent year data in the training population ( e . g . only 2011 , or 2010–2011 ) and scenarios using more historical year data in the training population ( e . g . 2009–2011 ) ( Table 3 ) . Cross validation experiment 1 ( CV1 ) accuracies were calculated for all experiments with the validation year/season included in the training population , excluding individuals in the validation fold . Including the validation year/season in the training population can bias accuracies upwards by confounding GxE and line effects , however , so in order to obtain an estimate of this bias , we also performed cross validation experiments 2 and 3 ( CV2 , CV3 ) for CV permutations 1–5 , see above table . For CV2 , we excluded the validation year/season from the training population . These results are not directly comparable to those in which the training population contained the validation year/season ( CV1 ) , however , because the training population for CV2 is smaller than was used for CV1 and training population size can have an important effect on prediction accuracy . For this reason , we performed CV3 , in which we included the validation year/season in the training population , but removed the equivalent seasons from 2011 , e . g . , for the first cross-validation permutation in the above table , CV2 would not include the 2012 dry season in the training population , and CV3 would include the 2012 dry season but would not include the 2011 dry season . Thus , the estimate of bias can be calculated for a given CV permutation experiment as CV3 accuracy minus the CV2 accuracy [26] . The bias was only estimated for the first five CV permutations because the bias estimates turned out to be small and similar to each other for all five CV permutations . For all three traits , multiple years , seasons , and replicate yield entries existed along with the previously described covariates for all 332 individuals . In order to build genomic selection models , it was necessary to convert these raw yields into a single , adjusted yield for each individual . Adjusted yields , plant heights , or days to flowering were calculated for each year/season combination by fitting an initial linear model of the observations y , by line ID ( GHID ) x1 , and phenotype covariates described above ( e . g . lodging ) x2…n for the given Year x Season in JMP . Non-significant covariates as determined by an F-test ( α > = 0 . 05 ) or covariates that resulted in singularities were removed , and the model re-fit . When all covariates included in the model were statistically significant , the least squares mean yield for each line ID was exported as the adjusted yield . Missing phenotype data were coded as null data for the above analysis , or , in other words , no imputation or numeric filling of phenotypic values was performed . The least square means for each year and season were also used to calculate a correlation matrix for each trait ( S3 Fig . ) . For each experiment , adjusted yields were calculated for each of the five training folds separately by fitting a linear model for each training fold as described above with the difference that data from all years and seasons for a particular CV experiment was including in the x matrices for all lines not in the validation fold . Year , season , and a year x season interaction were also included as covariates in the model , and subject to the same significance requirements as the other model covariates . Six statistical methods were used for each experiment , including four genomic selection methods: RR-BLUP , Bayesian LASSO ( BL ) , Reproducing Kernel Hilbert Spaces ( RKHS ) , and Random Forest ( RF ) , and two non-genomic selection methods: Multiple Linear Regression ( MLR ) and Pedigree-BLUP ( PED ) . The four genomic selection methods were chosen based on their demonstrated success in accurately predicting GEBV in variety of crops and because they represent the different types of statistical methodologies used to build GS models , i . e . , Linear parametric methods ( RR-BLUP , BL ) , non-linear semi-parametric methods ( RKHS ) , non-linear , non-parametric methods ( RF ) , as well as Frequentist methods ( RR-BLUP , RKHS ) , Bayesian methods ( BL ) , and machine learning methods ( RF ) [19 , 23 , 49 , 50 , 65 , 66 , 67] . For an overview of the methods , see Lorenz et al . , 2011[8] . Multiple linear regression using a subset of markers derived from single marker regressions ( MLR ) , another linear , parametric statistical method was the fifth statistical method tested to predict breeding value , and served as our used as a non-GS marker-based prediction control . For each fold , single marker regression was run for all markers and p-values determined for each marker by f-test . Note that this is the statistical equivalent of a crude GWAS . Linear models were then tested using 1 through the first 100 most significant markers , and the model with the best fit was returned . The returned model was then used to calculate the accuracy for the given fold . For the marker subset experiments where the number of markers ( p ) was less than 100 , models were tested using 1 through p markers . MLR has been shown to be effective for agronomic traits with very simple genetic architectures , but is otherwise not expected to perform well [51] . Prediction based on pedigree alone was the sixth statistical method and was performed in order to determine if a . ) the fold design method properly controlled for family structure within the dataset , and b . ) if GS could outperform prediction based on pedigree alone [52] . All statistical modeling was done in R . For the pedigree models an A-matrix was calculated using a three-generation pedigree file for all individuals in the training and validation populations using a custom R function . The models themselves were calculated using package rrBLUP ( function kin . BLUP ) . RR-BLUP models were also calculated using package rrBLUP ( function kinship . BLUP ) . RKHS models were calculated using kinship . BLUP , K . method = "GAUSS" , modified so that parameter theta was always equal to 2 . 5 , as per guidelines in the BGLR package documentation [68] . Random Forest was performed using package randomForest ( function randomForest ) . Bayesian LASSO was performed using package BLR ( function BLR ) . Narrow sense heritabilities were calculated for each trait on a per line basis using the rrBLUP package , function mixed . solve , with the least square means for the complete validation populations used as input . The narrow sense heritabilities were calculated as the additive genetic variance divided by the total phenotypic variance . The set of 73 , 147 SNPs was used for all experiments with the exception of the marker subset experiments described below . The cross-validation results were analyzed using ANOVA and pairwise student's t to determine: a significant difference in the accuracy of prediction of the two validation populations across statistical methods , i . e . , where yi ( accuracy ) = μ + xijβj + εij , and i is one RYT experiment and stat method for validation population j ( e . g . xi = CV experiment 1 for method RR-BLUP and j = validation population 2012 DS ) . b significant difference in the performance of the six statistical methods across the different experiments , i . e . , where yi ( accuracy ) = μ + xijβj + εij , and i is one RYT experiment for stat method j ( e . g . xi = CV experiment 1 and j = RR-BLUP ) . c significant difference in the performance of each experiment across statistical methods , after excluding the three worst-performing statistical methods ( Bayesian LASSO , MLR , and pedigree only ) , i . e . , where yi ( accuracy ) = μ + xijβj + εij , and i is one statistical method for RYT experiment j ( e . g . xi = RR-BLUP and j = CV experiment 1 ) ( Table 1 , S2–S4 Tables ) . Distributed . To select subsets of SNPs that were evenly distributed across the genome , 11 bin parameters were selected: 25Kb ( 0 . 1 cM ) , 50 Kb ( 0 . 2 cM ) , 120 Kb ( . 5 cM ) , 240Kb ( 1 cM ) , 480 Kb ( 2 cM ) , 840 Kb ( 3 . 5 cM ) , 1200 Kb ( 5 cM ) , 1800 Kb ( 7 . 5 cM ) , 2400 Kb ( 10 cM ) , 3600 Kb ( 15 cM ) , 4800 Kb ( 20 cM ) . For each bin parameter , all SNPs in the 73 , 147 SNP set were placed into bins according to the bin parameter . To select subsets of SNPs for a given bin size , the SNPs in each bin were sorted first by minor allele frequency , largest to smallest , and then by call rate , largest to smallest . Ten selections of SNPs were made for each bin size—the first subset consisted of the top ranked SNP in each bin , i . e . , the SNP with the highest MAF and call rate , the second subset consisted of the second ranked SNP in each bin , and so on for the top ten SNPs in each bin . If a bin had fewer than ten SNPs , then the top SNP in each bin was chosen for all ten selections . Each subset was then used as the genotype matrix to perform five-fold cross-validation using the same folds as for the original RYT cross validation experiments . The RYT 2012 wet season and the RYT 2012 dry season served as the validation populations and RYT years 2009–2011 , all seasons , served as the training population . The five marker-dependent statistical methods tested previously were used once more: RR-BLUP , RKHS , Random Forest , Bayesian LASSO , and MLR . Accuracy was calculated for each of the ten selections ( for each bin parameter ) as previously . A mean accuracy , standard deviation , and standard error for each bin parameter were also calculated by averaging the cross-validation results of the 10 selections for each bin parameter ( S5 Table ) . The average accuracies with standard error as the error bars were plotted versus the number of SNPs in each subset ( as determined by the bin size parameter ) using JMP ( Figs . 1 , S4 ) . The results for full 73 , 147 SNP set were included on these plots as a reference , although these accuracies are not averages . ANOVA and pairwise students were used to test for significant difference in the performance of the five statistical methods across the different bin parameter sizes , and for significant differences in the performance of the various bin parameter sizes ( and thus total SNP number ) across the five statistical methods ( S5–S6 Tables ) . Random . Ten random selections of SNPs were chosen from the 73 , 147 SNP set for 15 subset sizes: 24 , 48 , 65 , 83 , 96 , 109 , 161 , 212 , 316 , 448 , 781 , 1553 , 3076 , 7142 , 13101 using a pseudo-random numbers generator . Subset sizes 83 , 109 , 161 , 212 , 316 , 448 , 781 , 1553 , 3076 , 7142 , and 13101 were chosen to match the number of SNPs in the distributed SNP subsets described above . The additional SNP subset sizes were included to improve resolution . Cross validation experiments and analysis were performed for the random subsets as described above for the distributed subsets ( Fig . 1 , S4 Fig . , S6 Table ) .
|
Genomic selection is a promising breeding technique that aims to improve the efficiency and speed of the breeding process . While it has been shown to be effective in crops such as wheat and corn , it has not yet been applied to rice breeding . Genome-wide association studies ( GWAS ) , by contrast , are used to identify genes or QTLs that underlie traits of importance to breeding such as yield , flowering time , or plant height , and has been performed successfully in rice . Here , we experiment with applying genomic selection in conjunction with GWAS to a rice breeding program at the International Rice Research Institute in the Philippines and show that genomic selection can result in more accurate predictions of breeding line performance than pedigree data alone and that GWAS results can inform the results of GS . Our results suggest that GS could be an effective tool for increasing the efficiency of rice breeding .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Genomic Selection and Association Mapping in Rice (Oryza sativa): Effect of Trait Genetic Architecture, Training Population Composition, Marker Number and Statistical Model on Accuracy of Rice Genomic Selection in Elite, Tropical Rice Breeding Lines
|
Calmodulin ( CaM ) is a calcium sensing protein that regulates the function of a large number of proteins , thus playing a crucial part in many cell signaling pathways . CaM has the ability to bind more than 300 different target peptides in a Ca2+-dependent manner , mainly through the exposure of hydrophobic residues . How CaM can bind a large number of targets while retaining some selectivity is a fascinating open question . Here , we explore the mechanism of CaM selective promiscuity for selected target proteins . Analyzing enhanced sampling molecular dynamics simulations of Ca2+-bound and Ca2+-free CaM via spectral clustering has allowed us to identify distinct conformational states , characterized by interhelical angles , secondary structure determinants and the solvent exposure of specific residues . We searched for indicators of conformational selection by mapping solvent exposure of residues in these conformational states to contacts in structures of CaM/target peptide complexes . We thereby identified CaM states involved in various binding classes arranged along a depth binding gradient . Binding Ca2+ modifies the accessible hydrophobic surface of the two lobes and allows for deeper binding . Apo CaM indeed shows shallow binding involving predominantly polar and charged residues . Furthermore , binding to the C-terminal lobe of CaM appears selective and involves specific conformational states that can facilitate deep binding to target proteins , while binding to the N-terminal lobe appears to happen through a more flexible mechanism . Thus the long-ranged electrostatic interactions of the charged residues of the N-terminal lobe of CaM may initiate binding , while the short-ranged interactions of hydrophobic residues in the C-terminal lobe of CaM may account for selectivity . This work furthers our understanding of the mechanism of CaM binding and selectivity to different target proteins and paves the way towards a comprehensive model of CaM selectivity .
Calmodulin ( CaM ) , Fig 1a and 1b , is a promiscuous Ca2+-sensing protein that plays part in many physiologically important cellular processes [1] . Its two lobes , connected by a flexible linker , have one beta sheet and two EF-hand motifs each , Fig 1a–1c . The EF-hand binds a Ca2+ ion which induces tertiary structure rearrangements of lobe helices , exposing hydrophobic residues [2] . This allows CaM to bind and regulate a myriad of target proteins such as ion channels , kinases and G-protein coupled receptors . The Ca2+-signaling and olfactory transduction pathways ( Fig 1d and 1e ) are two examples of cell-signaling pathways where CaM is involved . In the Ca2+-signaling pathway , CaM activates and regulates the myosin light chain kinase IV ( MLCK ) and calcineurin ( CaN ) , among others [3] . Through this , CaM plays part in regulating a variety of biological processes including metabolism , proliferation and learning . In olfactory transduction , CaM’s role is instead to inhibit the CNG channel and activate the Ca2+/CaM-dependent protein kinase ( CAMK ) which inhibits the adenylate cyclase 3 ( ADCY3 ) . CaM’s promiscuity can be mainly ascribed to its ubiquity and interactions to specific target proteins depend partly on local cell environments such as availability of target proteins . Moreover , high affinity binding to target proteins is partly due to the flexible linker that allows wrapping around target proteins but also stems from the structural properties of the two lobes , thus linked to the pockets formed by their hydrophobic interfaces [4] . Here , we study the interactions between CaM and various target proteins , including proteins that are involved in the Ca2+-signaling and olfactory transduction pathways . Ca2+-independent CaM regulation , with Ca2+-free CaM ( apo CaM ) regulating the target protein may occur , although not as frequently as Ca2+-dependent CaM regulation . The IQ binding motif , named after the first two conserved residues of the motif , is often associated to apo-CaM binding while 1-5-10 or 1-5-8-14 motifs , named after the position of conserved hydrophobic residues , are associated with Ca2+-bound CaM ( holo CaM ) binding . [5–7] These motifs are , however , lacking the complexity to completely distinguish holo-CaM and apo-CaM binding , as well as different types of binding . An example is seen in Fig 2 , where apo and holo CaM C-terminal domain ( C-CaM ) expose different hydrophobic interfaces , but both bind the voltage-gated sodium channel NaV1 . 2 at the same IQ binding motif . Holo CaM often interacts with a higher affinity , but in some cases , such as the IQ motif of Nav1 . 2 , apo-CaM binding is more favorable than holo-CaM binding [8 , 9] . Both CaM dynamics and motif-dependent target-protein binding have been extensively studied via both experiments [2 , 8 , 10–25] and simulations . Indeed , all-atom molecular dynamics ( MD ) simulations allow for a detailed description with molecular insights . Early MD simulations of the holo CaM conformational ensemble have shown the holo-CaM linker to be flexible [26–28] and indicated structural destabilization when removing Ca2+ from one site in the N-term domain [29] . Moreover , CaM target-protein binding was approached by studying the dynamics as well as the thermodynamics of CaM or CaM complexes [4 , 26 , 30] , with insights into binding of specific target proteins using both regular MD and metadynamics [30] . However , little is known about the conformations of the binding interfaces and different binding modes of the two lobes , as well as their relation to protein-unbound CaM conformational ensemble . Part of this was addressed through the analysis of multiple aggregated MD simulations of C-CaM using Markov state modeling ( MSM ) , in which target protein binding was proposed to occur by C-CaM adopting the bound conformation before binding to the protein [31] . However , the analysis did not cover the differences of CaM N-terminal domain ( N-CaM ) and C-CaM dynamics , mechanisms and binding modes , nor did the study explore the possibility of different binding mechanisms linked to different binding modes . Binding a ligand can be conceptualized using two frameworks: conformational selection and induced fit . In pure conformational selection , the apo protein adopts a holo-like state before binding [32] . In pure induced fit , the ligand binds in a typical apo state that is not ideal , which causes subsequent rearrangements before reaching a high-affinity holo state [33] . In reality , binding likely involves a combination of the two mechanisms . However , spectroscopy experiments , as well as extensive MD simulations , may shed light on which mechanism is dominating . If the apo protein samples the holo-like state , conformational selection is typically assumed to be dominating , otherwise induced fit is assumed [34] . Here , we analyze thermally enhanced MD simulations of calmodulin with different Ca2+-occupancy and use spectral clustering to elucidate calmodulin selective promiscuity . We search for indicators of conformational selection by mapping solvent exposure of residues from sampled states ( clusters ) to contacts of already existing structures of CaM/peptide complexes . Moreover , we gain knowledge about the characteristics behind different binding modes of the two lobes , as well as the difference between holo and apo binding modes .
For this project , we considered four different binding states: holo and apo calmodulin , as well as Ca2+ bound only in the N-term ( N-holo ) and Ca2+ bound only in the C-term ( C-holo ) , Table 1 . The simulations of N-holo , C-holo and holo CaM used structure 3CLN [35] , while the apo simulations used structure 1LKJ [36] . N-holo was generated by removing Ca2+ from C-CaM and C-holo by removing Ca2+ from N-CaM from the holo structure . The systems were built using Charmm-gui [37 , 38] , where the protein was solvated in a box of about 21000 water molecules . The systems were then ionized with 0 . 15 M NaCl . Charmm36 was chosen as force field [39] , and TIP3P [40] as water model . The modified parameters of Charmm27 force field from Marinelli and Faraldo-Gomez [41] were used to parameterize Ca2+ ions . 5000 steps of minimization were carried out , followed by a 50 ps NVT ensemble equilibration with strong harmonic restraints on the protein atoms . The box was then scaled , relaxing pressure with Berendsen barostat [42] , while gradually releasing the position restraints for 350 ps . The MD parameters used in these simulations are extensively described elsewhere [43] . Calmodulin was simulated in an NPT ensemble with a 1 atm pressure and 2 fs time step . The short-ranged electrostatic interactions were modeled with a 1 . 2 nm cutoff where the switching function started at 1 . 0 nm , and the long-ranged ones with PME [44] . We used a Nose-Hoover thermostat [45] , an isotropic Parinello-Rahman barostat [46] , and constrained hydrogen bonds with LINCS [47] . The MD simulations were run at a constant temperature of 303 . 5 K . In addition to regular MD simulations , temperature enhanced simulations were performed; temperature replica exchange [48] ( T-REMD ) and replica exchange solute tempering [49–51] ( REST ) . In T-REMD , parallel simulations of independent replicas are run at different temperatures . A random walk in temperature space is achieved by employing the Metropolis criterion periodically to accept coordinate exchanges between neighboring replicas . Conformations obtained at higher temperatures are propagated down to the lower temperatures through exchanges . Because barriers are more easily passed at higher temperatures , the efficiency is increased when the free energy landscape is rugged . However , the number of replicas needed to span a certain temperature interval scales with system size [52] . For this reason , T-REMD may not always be efficient . To alleviate this , REST only modifies the hamiltonian of the system for the solute ( protein ) , and not the solvent [51] . However , the relative efficiencies of regular MD , T-REMD and REST depend on the ruggedness of the free energy landscape [43] . We used 25 replicas in both T-REMD and REST . The replicas of T-REMD spanned a temperature range of 299 . 13-326 . 09 K , while the REST replicas were simulated at temperatures between 300 . 0-545 . 0 K . The temperature ranges for REST and T-REMD were chosen using the “Temperature generator for REMD simulations” [53] , considering only the protein for REST . Exchanges between neighboring replicas were attempted every 2 ps , where half of the replicas were involved in each attempt . The REST simulations were performed with the Plumed 2 . 3b [54] plug-in patched with Gromacs version 5 . 1 . 2 [55] , where the charge of the atoms in the hot region were scaled , as well as the interactions between the two regions and the proper dihedral angles [56] . Analysis was performed using the replica at the lowest temperature . and was carried out on the protein heavy atoms . The first four residues in the apo structure were removed , because those are missing in the 3CLN structure . The apo CaM ensemble is more diffusive and generally well sampled by regular MD , while holo CaM free energy landscape is more hierarchical , and thus sampled more efficiently by temperature enhanced methods [43] . For this reason , more MD data is used in the apo analysis while more REST data is used for holo , Table 1 . The three methods may sample different regions with varying efficiency , which is why data from all three simulations is used . Fig 3 illustrates the steps carried out to analyze the MD simulation trajectories . In a first step , the protein was divided into quasi-rigid domains using SPECTRUS [57] . This procedure exploits fluctuations between residues to determine which parts of the protein move together . The clustering and post-processing described hereafter were done on each quasi-rigid domain . To further reduce dimensionality and complexity of the dataset while preserving the most important features , we used spectral clustering [58] . The advantage of spectral clustering compared to regular clustering techniques like k-means is three-fold . First , it is able to accurately assign points to non-convex clusters . Second , non-linear dimensionality reduction is intrinsic to the algorithm . For high-dimensional data , such as MD trajectories , non-linear dimensionality reduction improves clustering by circumventing the curse of dimensionality where the sparsity of the data increases with increased dimensionality [59] . Third , the number of clusters is the same as the number of dimensions onto which the points are projected . This feature becomes advantageous as the number of free parameters is reduced . In spectral clustering , the data manifold is first approximated by a graph with adjacency , or similarity , matrix A . It contains the local relationships between points and is constructed given the matrix of distances d . The distances dij are passed through a radial basis function , or Gaussian kernel , with scaling parameter σ , yielding the graph edge-weights A i j = e - d i j 2 2 σ 2 , ( 1 ) where dij is the dissimilarity between conformation xi and xj . The geodesic distance between two points is the distance between these points along the manifold , the shortest distance on the graph . The size of the scaling parameter , σ , influences the accuracy of geodesic distances , and should not be too large nor too small . A too large σ would yield short-cuts , thus causing non-convex clusters to be poorly defined . A too small σ , on the other hand , would result in a disconnected graph . Here , σ is the average nearest neighbor distance between configurations . The random walk matrix , related to the Laplacian [58] , is then constructed L = D - 1 / 2 A D - 1 / 2 . ( 2 ) The degree matrix , D , is diagonal with Dii being the degree of node i . The first k eigenvectors are computed and normalized per row to obtain points projected onto the k-sphere . These points are clustered into k clusters using k-means with centers projected onto the sphere . The choice of number of clusters is guided by the maximum eigengap , the difference between two consecutive eigenvalues . The representative structure of a cluster is chosen as the structure with smallest RMSD with respect to the other structures in the cluster . A cluster with all structures represents the conformational heterogeneity of one state . Here , in practice , the dissimilarity between conformations , dij , is measured as the distance between contact maps of inverse inter-atomic distances ( iiad-cmap ) . This general metric is effective for all proteins and , unlike root-mean-square-deviation ( RMSD ) , does not rely on structural alignment . The inverse distances make larger distances small so that far-away motions are cancelled out , without requiring a cutoff . The frame-to-frame distance matrix is compiled by computing the distances between each frame iiad-cmap . Each state , or cluster of frames , was analyzed to provide statistical information about its specific characteristics and molecular features . The features included interhelical angles , secondary structure and importance of states for target protein binding .
We sampled conformations of apo and holo CaM using MD , T-REMD and REST , Table 1 . The analysis began by identifying two quasi-rigid domains , namely N-CaM and C-CaM , with SPECTRUS [57] . The identified domains are shown as red and blue CaM ribbons in Fig 3 . The clustering and post-processing were therefore carried out on these two domains separately . S1–S4 Figs show the representative structures of holo and apo CaM obtained after clustering on C-CaM and N-CaM . Together with these states , experimental structures of similar conformations are shown for comparison . The molecular features were extracted from the conformational states by computing the interhelical angles and secondary structures . Fig 4 shows the interhelical angles of each state ( cluster ) from the holo simulations , while Fig 5 shows the interhelical angles for each state of the apo simulations . Each dot is a projection of a conformation in interhelical angle space , colored according to the state ( cluster ) it belongs to . On top of this , experimentally obtained structures are plotted as squares . The white squares denote structures where CaM is not bound to target peptides , Table 3 , while pink squares correspond to structures with CaM bound to a target peptide , Tables 4 and 5 . The black circles are mean values of CaM interhelical angles inferred from NMR data [2 , 24] , used for comparison . Furthermore , to allow a comparison between interhelical angles in the holo and apo ensemble , the sampled apo conformations are shown as a gray shadow in Fig 4 ( holo ) and the sampled holo conformations are shown as a gray shadow in Fig 5 ( apo ) . A simple sanity check , S5–S10 Figs , showed that most states were sampled by all three simulation methods , with a few exceptions . Examples of such are states that are separated by a high free energy barrier and therefore not sampled by the regular MD simulations within these timescales , or states that could not be sampled by the replica exchange simulations because of the relatively short time scales simulated . Similar conclusions can be drawn from the kinetic analysis of MD trajectories ( S11 Fig ) . A change in interhelical angles due to Ca2+-binding can be seen in Fig 4 ( holo ) and Fig 5 ( apo ) , in agreement with previous work [2 , 24 , 26] ( see S1 Text for more extensive description and details ) . We note also that the conformational ensemble generated by MD simulations is broader than the ones observed experimentally for both protein-bound and unbound CaM . In agreement with NMR studies of holo CaM [60] , the protein-unbound CaM adopts the conformations of CaM bound to specific target proteins , which is seen as an overlap between the pink squares and the MD data ( colored dots ) in Fig 4 , thus hinting at a possible conformational selection mechanism of target protein binding . Fig 6 displays the difference in secondary structure frequency for each residue compared to holo . Red dots denote the helical , black the strand and blue the coil content . Positive values indicate a gain in secondary structure element compared to holo , and negative ones a loss . The helical break around residue 74-82 [26–28] is seen to different extents in all Ca2+-binding states of CaM , Fig 6 . This break occurs in the linker and yields a compact state with the two lobes in contact . This compact state resembles the binding mode where CaM is wrapped around the target protein and was suggested to be an intermediate conformation during target protein- and Ca2+-binding [69 , 75] . Furthermore , binding Ca2+ rearranges the helices in the lobes and pins the beta sheets between residues 27/63 and 100/136 instead of 28/62 and 99/137 , Fig 6 . Apo CaM is more flexible , allowing the beta sheets to shift between 26-28/62-64 and 99-101/135-137 , Fig 6 . This highlights the rigidity of holo compared to apo CaM . Due to the flexibility of apo , a C-CaM state ( Fig 5b , state 2 ) is found where the end of the fourth binding loop is involved in a beta sheet , deforming the binding loop , in agreement with [31] . We hypothesize that this state may inhibit Ca2+-binding . In this state , residues 129-131 are more prone to form a beta sheet with 99-101 and 135-137 , S12 Fig . This extra sheet does not form in apo N-CaM because the coiled loop between helices C and D is slightly shorter than the one in apo C-CaM between helices G and H . The flexibility obtained by removing Ca2+ from C-CaM allows these beta sheet formations to occur . To investigate the possibility for conformational selection during the Ca2+-binding process , we searched for potential overlaps in interhelical angle space between apo and holo . We observe holo-like states in apo N-CaM , Fig 5a . ( states 3 , 5 and 8 , plotted with colors lime , cyan and cerise ) , as well as C-CaM holo-like states that have been reported in previous work [31 , 76] , Fig 5b ( states 1 and 5 , depicted as red and blue ) . These states could potentially aid Ca2+-binding though conformational selection . The holo N-CaM states 4 and 6 ( shown by cyan and magenta dots ) overlap with the apo ensemble , Fig 4a . Their secondary structure frequencies are plotted in S13 Fig , which show similar secondary structure frequencies as apo ( Fig 6b ) , with a shift of beta sheet to residues 28/62 . The C-CaM interhelical angles of holo and apo overlap more than the N-CaM interhelical angles do . The overlap in interhelical angles and shift of beta sheet in N-CaM , make these states likely intermediate states in Ca2+-binding , involved in conformational selective binding of Ca2+ . To assess whether Ca2+-binding allosterically modulates the conformational ensemble of the other lobe , we used the MD , T-REMD and REST datasets of C-holo ( Ca2+ in C-CaM ) and N-holo ( Ca2+ in N-CaM ) , Table 1 . Solvent exposure for each residue in N-holo and C-holo was compared to the apo ( Fig 7a and 7c ) , and holo ( Fig 7b and 7d ) ensembles . This shows that the Ca2+-free lobe tends to transition to an apo-like conformational ensemble , while the Ca2+-bound lobe stays in the holo-like ensemble . Although there is a small allosteric influence between the lobes , as seen in Fig 6c and 6d , where N-CaM mostly lacks beta sheet in C-holo , the solvent exposure analysis indicates that the overall conformation of one lobe is mostly independent of the conformation of the other . These results indicate that binding of Ca2+ to one of the lobes does not yield a significant population shift in the other lobe and thus that binding is likely not conformationally cooperative between the two lobes . Cooperative binding between sites and lobes has been investigated but the results obtained in different studies have been largely inconsistent , thus making it difficult to validate our results . This is because many different parameter sets in the models used to fit the experimental data perform equally well [77] . To investigate conformational selection aspects of calmodulin binding to target proteins , the relative solvent exposure of the different CaM residues in different states was calculated and mapped to the contacts in a set of CaM complexes , Tables 4 and 5 and Fig 3 . The underlying idea is that in a conformational selection mechanism , water molecules around an exposed residue will be replaced by the target peptide . Therefore , the residues that are in contact with the target peptides should be found exposed to solvent in a state involved in a binding mechanism dominated by conformational selection . To then characterize different modes of binding , the distributions of total solvent exposure per state were clustered and divided into classes , as described in the Materials and Methods section . We used molecular dynamics and temperature enhanced MD with an agnostic spectral clustering scheme to search for conformational selection of CaM Ca2+ and target protein binding . We found that the Ca2+-state of one lobe does not significantly influence the conformation of the other lobe , but binding Ca2+ may occur through conformational selection facilitated by a transition state that is visited by both apo and holo CaM . It is observed here by overlapping conformations in interhelical angle space . In N-CaM , the transition between apo and holo likely involves a beta sheet residue shift . The target protein binding modes of apo CaM are more shallow than those of holo CaM , and the charged interactions dominate due to the induced burial of the hydrophobic binding interface from Ca2+-depletion . The residues involved in binding are equally exposed in apo CaM states and thus all observed states are equally likely to initiate binding to the target protein . The notion that not only the linker , but also the binding interface of C-CaM shows configurational flexibility has previously been proposed both through extensive MD simulations [31] , but also in structural studies [2 , 24 , 25] . The conformation of C-CaM interface is suggested to vary more than the N-CaM interface [25] . Here , holo C-CaM shows distinct states exposing hydrophobic residues that are otherwise screened from water . Such states are absent in N-CaM , which also exhibits less binding heterogeneity than C-CaM in the CaM-complex structures . This is seen as the shallow class of C-CaM show more shallow binding than the shallow class of N-CaM , while the deep binding class of C-CaM shows deeper binding than the deep binding class of N-CaM . The distinct states of holo C-CaM with a clear mapping to bound structures indicate a tendency for C-CaM binding to be selective and dominated by conformational selection , while N-CaM , which lacks this clear mapping , likely binds through a more flexible mechanism involving intermediate states . For general protein-ligand binding , strong and long-ranged interactions have been thought to favor binding dominated by induced fit , while short-ranged interactions would favor conformational selection . [84] Our results support and extend this idea to CaM by showing that weak hydrophobic interactions dominate deep binding modes which indeed tend to be dominated by conformational selection in C-CaM . In the flexible binding of N-CaM , on the other hand , hydrophobic interactions occur less frequently . We hypothesize that the long-ranged electrostatic interactions of the N-CaM charged residues may initiate fast binding while the hydrophobic pocket in C-CaM may account for selectivity . Previously published studies using NMR [23] proposed that C-CaM is selective while N-CaM binds afterwards through induced fit mechanism , or a coupled conformational selection mechanism initiated by C-CaM [60] . This is in line with the argument given here on C-CaM selectivity and N-CaM flexibility . This study opens the way towards understanding the process and mechanisms behind calmodulin Ca2+-sensing . In a next step , the second aspect of binding , induced fit , may be studied . The full mechanism of specific target peptide binding could also be simulated starting from the conformational selection state of CaM identified here .
|
Calmodulin is a protein involved in the regulation of a variety of cell signaling pathways . It acts by making usually calcium-insensitive proteins sensitive to changes in the calcium concentration inside the cell . Its two lobes bind calcium and allow the energetically unfavorable exposure of hydrophobic residues to the aqueous environment which can then bind target proteins . The mechanisms behind the simultaneous specificity and variation of target protein binding is yet unknown but will aid understanding of the calcium-signaling and regulation that occur in many of our cellular processes . Here , we used molecular dynamics simulations and data analysis techniques to investigate what effect calcium has on the binding modes of calmodulin . The simulations and analyses allow us to observe and differentiate specific states . One domain of calmodulin is shown to be selective with binding involving short-distance interactions between hydrophobic residues , while the other binds target proteins through a more flexible mechanism involving long-distance electrostatic interactions .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] |
[
"chemical",
"characterization",
"medicine",
"and",
"health",
"sciences",
"molecular",
"dynamics",
"protein",
"interactions",
"electrophysiology",
"neuroscience",
"ion",
"channels",
"molecular",
"motors",
"actin",
"motors",
"protein",
"structure",
"thermodynamics",
"motor",
"proteins",
"research",
"and",
"analysis",
"methods",
"contractile",
"proteins",
"proteins",
"chemistry",
"binding",
"analysis",
"biophysics",
"molecular",
"biology",
"free",
"energy",
"physics",
"biochemistry",
"biochemical",
"simulations",
"cytoskeletal",
"proteins",
"cell",
"biology",
"neurophysiology",
"physiology",
"myosins",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"chemistry",
"computational",
"biology",
"macromolecular",
"structure",
"analysis"
] |
2018
|
Effect of Ca2+ on the promiscuous target-protein binding of calmodulin
|
In 2005 , there were outbreaks of febrile polyarthritis due to Chikungunya virus ( CHIKV ) in the Comoros Islands . CHIKV then spread to other islands in the Indian Ocean: La Réunion , Mauritius , Seychelles and Madagascar . These outbreaks revealed the lack of surveillance and preparedness of Madagascar and other countries . Thus , it was decided in 2007 to establish a syndrome-based surveillance network to monitor dengue-like illness . This study aims to evaluate the use of capillary blood samples blotted on filter papers for molecular diagnosis of CHIKV infection . Venous blood samples can be difficult to obtain and the shipment of serum in appropriate temperature conditions is too costly for most developing countries . Venous blood and dried-blood blotted on filter paper ( DBFP ) were collected during the last CHIKV outbreak in Madagascar ( 2010 ) and as part of our routine surveillance of dengue-like illness . All samples were tested by real-time RT-PCR and results with serum and DBFP samples were compared for each patient . The sensitivity and specificity of tests performed with DBFP , relative to those with venous samples ( defined as 100% ) were 93 . 1% ( 95% CI:[84 . 7–97 . 7] ) and 94 . 4% ( 95% CI:[88 . 3–97 . 7] ) , respectively . The Kappa coefficient 0 . 87 ( 95% CI:[0 . 80–0 . 94] ) was excellent . This study shows that DBFP specimens can be used as a cost-effective alternative sampling method for the surveillance and monitoring of CHIKV circulation and emergence in developing countries , and probably also for other arboviruses . The loss of sensitivity is insignificant and involved a very small number of patients , all with low viral loads . Whether viruses can be isolated from dried blood spots remains to be determined .
Chikungunya virus ( CHIKV ) is an arthropod-borne alphavirus transmitted by mosquitoes of the Aedes genus . Chikungunya has spread widely through Africa [1] , [2] , [3] , South-East Asia , the Indian subcontinent [4] , [5] , [6] , and most recently Southern Europe [7] , [8] . There was a large outbreak in La Réunion in 2005–2006 with an estimated 266 , 000 individuals infected [9] . Populations on other islands in the Indian Ocean ( Mauritius , Madagascar , Mayotte , the Seychelles and islands of the Union of the Comoros ) were also affected by this virus [3] , [10] , [11] . The Northern and Eastern parts of Madagascar experienced their first documented Chikungunya outbreak in 2006 , concomitantly with the circulation of dengue virus serotype 1 [11] . Since then , CHIKV has become endemic and now circulates continually in the eastern part of the island . In February 2010 , the primary health care centre of Mananjary ( southeast of Madagascar ) reported an outbreak of dengue-like illness ( fever , arthralgia , myalgia , rash , and headache ) . Laboratory tests confirmed CHIKV infection in more than 90% of the patients sampled . Due to logistic constraints and lack of resources , it was decided to monitor the outbreak using an alternative support for blood specimens . Arboviruses are heat-labile RNA viruses , and diagnosis of these virus traditionally requires a blood sample stored at 4°C for only a short time , and quickly transported to the laboratory at a temperature not exceeding 4°C [12] . Dried-blood blotted on filter papers ( DBFP ) are a possible alternative , cost-effective and technically appropriate for low-incomes countries . Several studies have demonstrated that DBFP are suitable for serological and molecular diagnosis of bacterial , viral , and parasitic diseases [13] , [14] , [15] , [16] , . The diagnosis of dengue infection using DBFP has already been described [21] , [22] , this support has not been evaluated for the diagnosis of Chikungunya , other than during a serological study [23] . During a CHIKV outbreak on the East coast of Madagascar , we evaluate the value of DBFP for the molecular diagnosis of CHIKV infection and its usefulness for monitoring a CHIKV outbreak in a low-resources country .
Specimens and data were collected within the activities of the national public health sentinel surveillance systems and therefore this was considered to be non-research activity . Fever surveillance , including arbovirus protocols , was approved by the respective ministries of health and the National Ethics Committee of Madagascar ( FWA00016900 ) . Before taking each specimen , physicians explained the purpose of the surveillance system . Patients were then free to refuse to participate . Oral consent was documented in the patient form . This research did not cause any additional trauma and all injuries suffered by individuals were associated with routine care . The specimens used in this study were collected as part of the routine care , and this study was designed retrospectively . As specimens were anonymous , written consent could not be obtained . The National Ethics Committee approved the use of oral consent . In 1996 , in accordance with World Health Organisation resolution AFR/RC43/R7 , the Integrated Diseases Surveillance and Response system was implemented by the ‘Direction des Urgences et de la Lutte contre les Maladies Négligées’ ( DULMN ) of the Malagasy Ministry of Public Health . Chikungunya fevers are a notifiable disease . To detect such event , a sentinel surveillance network for fever syndrome was established in 2007 and now includes 34 sentinel sites in 32 health districts of Madagascar [24] . All patients presenting at one clinic with fever were tested for malaria using a rapid diagnostic test ( RDT ) ; the general practitioner filled case report forms for all febrile patients . In some clinics involved in virological surveillance , patients that fulfilled the case definition for dengue-like illness with onset of fever 5 or fewer days earlier were sampled . Specimens were stored in a liquid nitrogen tank and shipped weekly to the National Reference Laboratory for Arboviruses ( NRLA ) at the Institut Pasteur from Madagascar . Specimens used in this study were collected during two different periods . The first period ( from February to March 2010 ) was during an outbreak of Chikungunya in two districts on the Southeast coast of Madagascar ( Mananjary and Farafangana ) . The second period was the post-outbreak period ( from July to September 2010 ) . All patients , visiting sentinel centres , that fulfilled the case definition for dengue-like illness with onset within the previous 5 days were included in the study and both venous blood and capillary blood spotted onto a clean Whatman 3MM filter paper ( Sigma-Aldrich , St . Louis , MO , USA ) were collected . Dengue-like illness was defined as a presence of fever ≥38°C and two or more of the following: retro-orbital or ocular pain , headache , rash , myalgia , arthralgia , leukopenia , and haemorrhagic manifestations [25] . Cases and controls were defined retrospectively as follows: sera from patients included in our study and who fulfilled the dengue-like syndrome case definition were tested for Chikungunya . Cases were then defined as patients with Chikungunya infection confirmed by real-time RT-PCR or Indirect Immunofluorescence assay after viral isolation on Vero E6 or AP61 cells . Controls were defined as patients included in the study but tested negative for Chikungunya . All sera were kept in a liquid nitrogen tank before shipping to the NLRA then at −80°C from arrival until analysis . DBFP were kept at room temperature ( 25°C ) until analysis as previously described [26] . CHIKV RNA was obtained after propagation of CHIKV in Vero E6 cells for 5 days . Cell supernatants , enriched with CHIKV ( CHIKV-SP ) , were then collected and a volume of 100 µL was tested to confirm the presence of CHIKV ( see below ) . RNA was extracted from 100 µL aliquots of serum or CHIKV-SP , using Trizol LS ( Invitrogen Life Technologies , Paisley , Refrewshire , United Kingdom ) , according to the manufacturer's recommendations . The RNA was precipitated with 400 µL of isopropanol ( SIGMA ) , air-dried and suspended in 50 µL of RNase-free water . For DBFP , a 6 mm diameter disk containing the dried blood spot was cut using a paper puncher and placed in a 1 . 8 mL tube as previously described [27] . To avoid contamination and false positive results , puncher was soaked in NaOH ( O . 1N ) and rinsed with RNase free water between each DBFP . Viral RNA was then extracted using Trizol reagent ( Invitrogen Life Technologies , Paisley , Refrewshire , United Kingdom ) , according to the manufacturer's recommendations . RNA was precipitated with 500 µL of isopropanol . To validate each series of extraction , 15 µL of CHIKV-SP was spotted onto Whatman 3 MM filter Papers and dried . Six mm diameter disks were cut and put in a sealed plastic bag to prevent moistening and stored at +4°C until use as controls [26] . DBFP from healthy individuals previously tested negative for CHIKV were used as negative controls . CHIKV RNA was detected by one-step real-time RT-PCR assays in a Rotorgene 6000 apparatus ( Corbett life science ) . Oligonucleotide primers were used with dual-labelled hydrolysis ( Taqman ) probes adapted from Laurent P et al , that target the E1 region ( GenBank AF369024 ) ( Table 1 ) [28] . The Ag-Path One Step RT-PCR kit ( P/N: 4387391 , Ambion , Foster City , CA , USA ) was used for amplification . The reaction mix , in a final volume of 25 µl , consisted of 2 . 5 µL of RNA extract , primers at a final concentration of 0 . 5 µM , and the CHIK probe at a final concentration of 0 . 3 µM . The RT-PCR conditions were as follows: a 40 min reverse-transcription step at 50°C followed by denaturation for 10 min at 95°C and 45 cycles of denaturation at 95°C for 10 s and annealing/extension at 56°C for 60 s . Positive controls , negative controls and no template controls ( NTC ) were included in each series . Runs were validated only if the NTC and the negative control did not exhibit fluorescence curves that crossed the threshold line , and the positive control gave a fluorescence curve that crossed the threshold line within 39 cycles ( Ct≤39 ) . A specimen was considered positive for CHIKV if it gave a positive reaction with a Ct≤39 . Dilutions ( 10−3 to 10−7 ) of CHIKV-SP were mixed with blood from a healthy donor . Fifteen µL of this mix was blotted on Whatman 3 MM filter Paper and a 6 mm diameter disk containing the dried spot was cut . Viral RNA was extracted and detected in blood samples mixed with different dilutions of CHIKV-SP before ( 15 µL ) and after being blotted onto filter papers . The sensitivity and specificity of the assays were evaluated using two-by-two tables . The sensitivity and specificity of the DBFP specimens were determined by comparison with the results obtained with venous blood samples ( sera ) by the routine diagnostic test ( real-time RT-PCR ) . Data were recorded and analysed statistically by R Analysis with R version 2 . 7 . 0 software [29] . Results with a two-sided p value≤0 . 05 were scored as being significant . A non-parametric test , the Wilcoxon test , was also carried out . Receiver Operating Characteristic ( ROC ) plot analysis was performed to determine the best threshold value for the CHIKV RNA load by real-time RT-PCR obtained with DBFP compared with the values obtained from sera . As the mean value for the negative reference sample is expected to be smaller than the mean value for the positive reference sample , inverse transformation of the test data was used to prepare data for analysis as previously described [30] . AUC ( Area under Curve ) values were used to assess the discrimination of the test compared with the reference .
During the first week of February 2010 , the health authorities of the health district of Mananjary ( Southeast coast of Madagascar ) reported an increased incidence of febrile syndrome with arthralgia . Sera from 11 suspected cases were shipped to NRLA at the Institut Pasteur from Madagascar . All specimens tested positive for CHIKV by real-time RT-PCR . From February to October 2010 , 3 , 177 suspected cases were recorded by DULMN and 191 cases were confirmed by laboratory tests . Overall , 181 samples from patients presenting dengue-like illness were included in our study: 73 ( 40 . 3% ) were CHIKV confirmed cases and 108 ( 59 . 7% ) were negative controls; the median age were 18 years and 32 years and sex ratio ( M/F ) were 1 . 2 and 0 . 5 , respectively . No dengue virus infection was detected . Among the 181 patients tested , DBFP for 74 ( 40 . 9% ) and sera for 73 ( 40 . 3% ) scored positive for CHIKV ( Table 2 ) . Results for DBFP and sera were concordant for 170 ( 93 . 9% ) patients and discordant for 11 ( 6 . 1% ) . The Kappa coefficient was 0 . 87 ( p<0 . 001; 95% CI:[0 . 80–0 . 95] ) . The sensitivity and the specificity of the test performed with DBFP were 93 . 1% ( 68/73; 95% CI:[84 . 7–97 . 7] ) and 94 . 4% ( 102/108; 95% CI:[88 . 3–97 . 9] ) , respectively ( Table 2 ) . The quantities of RNA obtained from filter paper and serum are shown in Figure 1 . The difference of the Cycle threshold ( Ct ) between viral amplification from 15 µl of whole blood and DBFP containing the same viral dilution varied from 2 . 5 to 3 . 48 . As 3 . 32 Ct is equivalent to 1 Log of RNA quantity , the loss of viral RNA associated with drying samples on filter paper can be estimated to be between 0 . 5 Log to 1 . 05 Log ( mean 0 . 825 Log ) . The value of the Area Under Curve ( AUC ) was 0 . 96 ( 95% CI:[0 . 93–0 . 99] . The optimal Ct cut-off for real-time RT-PCR with dried blood spot samples was 40 with an AUC = 0 . 96; this value correctly classified 95 . 3% of the data ( sensitivity = 93 . 4% and specificity = 96 . 2% ) ( Figure 2 ) . All 11 discordant cases had Ct values between 37 and 39 cycles . A second real-time RT-PCR was performed in triplicate for 8 of these 11 discordant paired samples and both concordant and discordant results were obtained with the triplicate repeats ( data not shown ) .
The aim of this study was to validate an alternative method for sampling and the diagnosis of CHIKV infection suitable for surveillance in low income countries . Serological methods ( IgM detection ) have been used to confirm CHIKV infection or circulation [23] , but these methods have various limitations: for example , it is generally necessary to obtain paired sera , one during the acute phase and one during the convalescent phase such that seroconversion or an increase in IgM titres can be detected . In most of poor settings countries like Madagascar or in remote regions , it can be difficult to obtain serum during the convalescent phase . The use of molecular techniques has already been described for the diagnosis of infection caused by several viruses , including dengue [21] , [22] , measles [31] , and Rift Valley fever virus [32] . Genotyping of measles virus has also been reported [33] . Our study showed that for diagnosis , outbreak monitoring and virological surveillance of CHIKV infection and circulation , capillary blood samples taken from the finger and spotted onto filter paper is a cost-effective alternative with a good sensitivity and specificity ( 93 . 1% and 94 . 4% , respectively ) . Together with the detection of Dengue virus from DBFP , that has been found to perform well with a sensitivity and a specificity of 90 . 7% and 82 . 9% respectively when compared to the detection from sera [21] , [22] . We implement our sentinel surveillance for the detection of both CHIKV and Dengue Virus using DBFP as specimen collection system . Despite its good sensitivity , we have observed some discordant results between DBFP and sera . One explanation could be an intrinsic low viral load in the samples and the limit of detection of the real-time RT-PCR . A similar observation has been reported in other studies [22] . Viral RNA in DBFP may decay , if stored for long periods , and this is a possible limitation , as the virus may become undetectable . However , some studies have shown that viral RNA in filter paper was stable for several weeks at room temperature ( 25°C ) [26] . In our laboratory , we were able to detect CHIKV in DBFP after 6 months of storage at 25°C ( data not shown ) . Similarly , Gauffin F . et al . found no significant decay of RNA in dried blood spots stored for up to 20 years [34] . Another possible limitation of our study is that the viral load may differ between capillary and venous blood . However , this does not appear to be a problem because we restricted blood collection to the first 5 days after the onset of fever . Nevertheless , both the decay of CHIKV RNA upon storage and viral loads in capillary and venous blood should be further studied . The use of DBFP instead of serum eliminates the need for a cold chain or for a nitrogen tank for transportation . For our current arbovirus surveillance system , the overall cost for sampling and shipping of specimens using a 10 L nitrogen tank by road once a week from/to one sentinel site is around 300 US$ per week whereas using DBFP this cost has decreased to less than 10 US$ per week . For centres only accessible by air , costs of shipment using nitrogen tank or isothermal boxes is not sustainable for a country like Madagascar . This method has other advantages . In particular , capillary blood collection is easier in young children . Self-collection is also possible during outbreaks when health workers may be overloaded with work . Self-collection of DBFP samples has already been used successfully for serosurveys during the 2006 Chikungunya outbreak in La Réunion [23] . Currently , this method of collection is used by the Ministry of Health from the Union of the Comoros to collect and ship specimen to our NRLA for dengue-like syndromes investigations . Despite these various advantages of the use of DBFP , it is important to note that this method is probably not suitable for subsequent analyses involving growing the virus . Nevertheless , it has been shown that recovery of some flaviviruses ( Dengue , West Nile and Yellow fever ) and an alphavirus ( Venezuelan equine encephalitis ) was possible from DBFP stored for up to 90 days [12] , [26] . More work is needed to evaluate the viability of CHIKV in dried blood . In conclusion , we demonstrate that DBFP is a cost-effective method for surveillance and for the monitoring of viral outbreaks in low income countries , and especially in large countries where the access to laboratory facilities is limited . This method can facilitate the extension of surveillance system networks , and may be useful to public health authorities for rapid identification of Chikungunya outbreaks and , by extension , those of other arboviruses ( e . g . Dengue fever , Rift Valley Fever , West Nile ) .
|
Chikungunya is a mosquito-transmitted viral disease . No treatment is currently available . The only way to prevent infection is to avoid mosquito bites . Surveillance of circulation by early diagnosis is useful to prevent or limit outbreak . CHIKV , like all RNA viruses , is heat-labile . Consequently , confirmatory diagnosis classically requires blood samples that are transported in appropriate conditions ( i . e . at 4°C within 48 hours , in liquid nitrogen , or frozen at −80°C and transported on dry ice ) to prevent false negative results . This is not always possible in field conditions in low income countries . Dried blood spots are already used to diagnose parasitic , bacterial and viral infection . We compared venous sample to dried blood sample to make diagnosis of Chikungunya infection . We demonstrate the usefulness of this sampling method for the molecular diagnosis of Chikungunya infection . In particular , dried blood spots were very nearly as suitable as frozen serum specimens for the diagnosis of recent infection by CHIKV .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"test",
"evaluation",
"survey",
"methods",
"diagnostic",
"medicine",
"emerging",
"infectious",
"diseases",
"infectious",
"disease",
"epidemiology",
"virology",
"epidemiology",
"emerging",
"viral",
"diseases",
"neglected",
"tropical",
"diseases",
"arboviral",
"infections",
"biology",
"microbiology",
"viral",
"disease",
"diagnosis"
] |
2013
|
Dried-Blood Spots: A Cost-Effective Field Method for the Detection of Chikungunya Virus Circulation in Remote Areas
|
The bank vole is a rodent susceptible to different prion strains from humans and various animal species . We analyzed the transmission features of different prions in a panel of seven rodent species which showed various degrees of phylogenetic affinity and specific prion protein ( PrP ) sequence divergences in order to investigate the basis of vole susceptibility in comparison to other rodent models . At first , we found a differential susceptibility of bank and field voles compared to C57Bl/6 and wood mice . Voles showed high susceptibility to sheep scrapie but were resistant to bovine spongiform encephalopathy , whereas C57Bl/6 and wood mice displayed opposite features . Infection with mouse-adapted scrapie 139A was faster in voles than in C57Bl/6 and wood mice . Moreover , a glycoprofile change was observed in voles , which was reverted upon back passage to mice . All strains replicated much faster in voles than in mice after adapting to the new species . PrP sequence comparison indicated a correlation between the transmission patterns and amino acids at positions 154 and 169 ( Y and S in mice , N and N in voles ) . This correlation was confirmed when inoculating three additional rodent species: gerbils , spiny mice and oldfield mice with sheep scrapie and 139A . These rodents were chosen because oldfield mice do have the 154N and 169N substitutions , whereas gerbil and spiny mice do not have them . Our results suggest that PrP residues 154 and 169 drive the susceptibility , molecular phenotype and replication rate of prion strains in rodents . This might have implications for the assessment of host range and molecular traceability of prion strains , as well as for the development of improved animal models for prion diseases .
The conversion of the cellular prion protein ( PrPC ) into an abnormally-folded isoform ( PrPSc ) that accumulates in the brain of affected individuals represents the key feature of transmissible spongiform encephalopathies ( TSEs ) , or prion diseases [1] . They include bovine spongiform encephalopathy ( BSE ) in cattle , scrapie in sheep , Creutzfeldt-Jakob disease ( CJD ) and variant CJD ( vCJD ) in humans . According to the “prion theory” , PrPSc is the major or the sole component of the TSE agents , named prions . These unusual agents are believed to self-propagate by catalyzing the conversion of PrPC into PrPSc which acts as a template [2] . Experimental animals are of paramount importance for the study of TSEs . However , very long incubation periods or even unsuccessful transmissions are observed when a given model is challenged with prions from a different species . Prion transmission to a new species is in fact limited by a phenomenon known as “species barrier” [3] . Early studies argued that the main factor influencing interspecies transmission resides in the homology degree of the amino acid sequence of PrP between the donor and recipient species [4] . Differences in the PrP sequence can result in a non-effective interaction between PrPSc and PrPC and in an inefficient propagation of PrPSc that produce prolonged incubation periods . The sequence of the donor PrPSc and host PrPC is identical on second passage in the same species and this can explain the adaptation of the agent to the new host , which results in shorter incubation periods . Transgenic mice carrying the PrP gene of the donor species have been generated with the aim of removing the transmission barrier . These models provided evidence that not only PrP homology but also the prion strain played a prominent role in the transmission barrier . Although highly susceptible to sporadic CJD , transgenic mice over-expressing human PrP showed lower susceptibility to vCJD than wild-type mice [5] . Individual strains represent different PrPSc conformations within the framework of the “prion theory” [6] . Recently , we reported that the interspecies transmission of prions from humans to bank voles ( Myodes glareolus ) can occur without an apparent species barrier despite a low degree of PrP sequence homology between voles and humans [7] . Studies of transmission barrier are important for elucidating the basis of prions replication and acquiring knowledge to decipher the risk of interspecies transmission . The availability of animal models susceptible to different prion strains is of crucial relevance for such kind of studies . We recently showed that the bank vole is very susceptible to TSEs [7] , [8] , [9] . Here , we studied the transmission features of different TSEs in a panel of seven rodent species showing various degrees of phylogenetic affinity and specific PrP sequence divergences in order to investigate the molecular basis of the high susceptibility of voles in inter- and intra-specific transmissions .
Transmission studies were first set up in bank and field voles in comparison to C57Bl/6 and wood mice . Concerning natural scrapie ( SS3 ) , the results of primary transmission to bank voles and C57Bl/6 mice were previously reported [9] . One hundred per cent of bank and field voles developed obvious clinical signs and were sacrificed after short survival times following inoculation of natural scrapie ( Table 1 ) . The first signs of disease in both vole species were hyperactivity/reactivity followed by the progressive disappearance of the typical behaviour of hiding under the cage's sawdust . Overt neurological signs appeared later and consisted of incessant walking along the cage and characteristic upward movements of the head ( head bobbing ) , accompanied by severe and progressive ataxia . Hunched posture , apathy and pronounced hypo-activity/reactivity preceded sacrifice or death which occurred 10–20 days after the onset of neurological signs . In contrast , the inoculation of natural scrapie in C57Bl/6 and wood mice produced very long survival times without overtly suggestive signs of prion disease . C57Bl/6 mice rarely showed subtle and equivocal signs such as nervousness and hyper- reactivity , followed by apathy . PrPSc brain accumulation was detected by Western blot in 100% of bank and field voles whereas it was not detected in wood mice and in only three out of nine C57Bl/6 mice ( Table 1 ) . The reverse situation was apparent following the inoculation of BSE . Neither of the two vole species showed clinical signs or PrPSc accumulation during their lifetime . This is in contrast to C57Bl/6 and wood mice which showed overt neurological signs characterized by hyper-activity/reactivity and followed by hind limb incoordination , hunched posture and apathy . Survival times were rather long in both mice species , but the transmission rate was high ( Table 1 ) . Western blot showed that the apparent molecular weight ( MW ) of proteinase K-treated PrPSc from BSE and natural scrapie was maintained after transmission to rodents . Indeed , the unglycosylated isoform of PrPSc from BSE was ∼1 kDa lower than scrapie ( Fig . 1 ) . Glycoform analysis showed the typical three bands whose rank order of density was diglycosylated>monoglycosylated>unglycosylated in all species and with both TSE sources ( Fig . 2 ) . The BSE glycoprofile in C57Bl/6 and wood mice was characterized by a very high intensity of the diglycosylated isoform . Second passages were performed on C57Bl/6 mice and bank voles , which were considered representative of the two rodent groups . They were successful in all inoculated animals ( Table 1 ) . Survival times shortened compared to primary transmissions , which demonstrated the existence of obvious transmission barriers . Despite the lack of any evidence of BSE transmission to voles upon primary passage , a second “blind” passage of BSE was carried out by using the brain of the individual showing the longest survival time upon primary transmission ( 1 , 044 days post-infection , d . p . i . ) for the preparation of the inoculum . This led to the appearance of overt clinical signs with severe excitability and ataxia . Voles showed 483±85 d . p . i . survival time and 100% transmission rate based on both spongiform change and PrPSc accumulation; this latter retained the typical signature of BSE ( Fig . 3 , lane 1 ) . The third passages of SS3 and BSE were carried out in order to investigate their adaptation to bank voles and C57Bl/6 mice ( Fig . 4 ) . The survival time was unchanged after the second passage of SS3 in both species . On the contrary , it further shortened in bank voles that were challenged with BSE but not in C57Bl/6 mice . This is likely the consequence of a low level of replication upon primary transmission . This resulted in subclinical infection and possible low infectious titre of the inoculum used for the second passage . Both SS3 and BSE adapted to voles as very rapid strains , while longer survival times were observed after their adaptation to C57Bl/6 mice ( Fig . 4 ) . The primary transmission results of natural scrapie and BSE suggested a differential susceptibility of the two vole species on one side , and of C57Bl/6 and wood mice on the other . Bank voles , field voles and wood mice were inoculated with the 139A strain and the transmission characteristics compared to those observed in C57Bl/6 mice in order to investigate if such a pattern of susceptibility was maintained even after the inoculation of a well characterized mouse-adapted scrapie strain . 139A was transmitted very efficiently ( 100% transmission rate ) to all species . All inoculated animals showed clinical signs and revealed spongiform degeneration and PrPSc accumulation in their brain ( Table 2 ) . Strikingly , both vole species showed shorter survival times than C57Bl/6 mice which is the species to which that strain is adapted . Wood mice showed the longest survival times among the four species ( Table 2 ) . Clinical signs of disease were similar in C57Bl/6 and wood mice and characterized by progressive weight loss , dorsal kyphosis , incoordination of hind limbs and plastic tail . The clinical picture in voles was clearly different from that observed after inoculation of natural scrapie . It included hyperactivity/excitability , followed by 10–15 days of reduced activity and behavioural depression . Motor dysfunctions were much less evident compared to what was observed after inoculation of natural scrapie . Spongiosis was widespread in the brain of all species with the exception of the cerebellar cortex . Both granular and molecular layers of cerebellar cortex were targeted by moderate/high vacuolar degeneration in C57Bl/6 and wood mice , while spongiosis was only occasional and confined to the granular layer in field and bank voles ( Fig . 5 ) . Interestingly , the molecular analysis of PrPSc provided further evidence of the differences in the transmission features of prions between voles and mice . As a matter of fact , the typical 139A glycoprofile in mice , monoglycosylated>diglycosylated>unglycosylated , was faithfully maintained in wood mice , while it clearly changed to a diglycosylated>monoglycosylated>unglycosylated pattern in voles ( Fig . 1 and 2 ) . Second passage of 139A was carried out in the three rodent species under investigation . Survival times were very short in both vole species , while in wood mice they were rather long and similar to those observed in C57Bl/6 mice ( Table 2 ) . Molecular analysis showed that the PrPSc glycoprofiles seen in primary transmissions were maintained upon second passages ( data not shown ) . The adaptation of 139A confirmed the very short survival times of vole-adapted strains , which were already observed with SS3 and BSE . The hypothesis of a high expression level of PrPC which accounted for these findings , was ruled out by Western-blot and Histo-blot analyses , because they did not show any significant differences either in the distribution or in the level of PrPC expression in the brain of bank voles , field voles , wood mice and C57Bl/6 mice ( data not shown ) . 139A was fully adapted and stabilized in bank voles with the third passage and subsequently inoculated back into C56Bl/6 mice in order to investigate if the novel PrPSc glycoprofile observed in voles inoculated with 139A could have been considered as the emergence of a different strain with a new stable molecular signature . The third passage of 139A to bank voles produced the same survival time ( 76±8 d . p . i . ) and PrPSc characteristics ( data not shown ) as the second passage . This suggested that the strain had already been adapted to the new host at the second passage . After inoculation , all C57Bl/6 mice ( n = 20 ) developed the disease showing spongiform change and PrPSc accumulation in their brain . Survival times were long ( 463±62 d . p . i . ) , suggestive of the existence of a transmission barrier also during the transmission from voles to mice , the species to which 139A was originally adapted . Worth mentioning is the fact that the molecular characteristics of PrPSc reverted to that of the original mouse inoculum ( Fig . 3 , lane 5 ) . The comparison of PrP sequences of the bank vole , field vole , wood mouse and laboratory mouse displayed a high homology degree , although a number of substitutions were found in the N-terminal cleaved signal peptide and in the C-terminal signal sequence that is also cleaved when the GPI-anchor is added . Sequence comparison showed relevant amino acid substitutions at only five positions ( Fig . 6 ) . For the sake of clarity , the numbering system used throughout the text for amino acid residues refers to the mouse PrP sequence . The first substitution , G89S , was at the N-terminus non-structured tail of PrP and was observed only in the field vole . The second replacement , L108M , was in the N-terminal disordered tail and is known to influence the susceptibility of both voles [8] and mice [10] . Two substitutions were found in the structured C-terminal domain . A replacement Y154N was found in the loop region between the first α-helix and the second β-strand , while a substitution S to S169N was in the loop between the second β-strand and the second α–elix ( Fig . 7 ) . They both distinguished the sequences of laboratory and wood mice from those of voles . Finally , the substitution D226E was in the C-terminal region , and also differentiated the PrP sequences of laboratory and wood mice from those of the two vole species . We especially focused our attention on the two variations observed in the structured C-terminal domain of PrP , which are located into regions that supposedly contribute to the species barrier because they apparently function as selective protein-protein interaction sites or are involved in the specificity of intermolecular interactions [11] . In order to test the hypothesis of the role of Y154N and S169N substitutions in influencing the transmission and phenotype characteristics of prions to rodents , we analyzed the PrP sequence of other rodents frequently bred under laboratory conditions and hence selected for transmission studies three additional species: the oldfield mouse , the Mongolian gerbil and the spiny mouse . They were chosen because oldfield mice showed Y154N and S169N substitutions , whereas gerbil and spiny mice did not show them ( Fig . 6 ) . Furthermore they have different levels of phylogenetic relationship with the previously inoculated rodent species ( Fig . 8 ) [12] . Groups of oldfield mice , spiny mice and gerbils were challenged with the same inocula of 139A and natural scrapie used in previous transmissions . Following inoculation of both scrapie sources , oldfield mice developed the disease with short survival times , comparable to those of voles , while gerbils showed a very inefficient transmission of natural scrapie and long survival times after inoculation of 139A ( Table 3 ) . Overall results confirmed that Y154N and S169N were the only variations that correlated with the different transmission patterns observed . Besides the vole species , residue 108M also occurred in gerbils and spiny mice , while D226E was found in the two voles , but not in oldfield mice which showed a similar susceptibility to voles . N99G was exclusive to spiny mice and might explain the apparent resistance of this species to both 139A and natural scrapie . In fact , this amino acid substitution has been reported to have an inhibitory effect on PrPSc formation in rabbits , a species thought to be resistant to TSEs [13] . In accordance with the overall data , phenotypic analysis of this second set of transmissions by brain histopathology and molecular analysis of PrPSc revealed characteristics which paralleled those observed in voles , wood mice and laboratory mice . A severe vacuolar degeneration of molecular and granular layers of the cerebellum was evident in gerbils , while the molecular layer was completely spared in oldfield mice ( Fig . 5 ) . Furthermore , the glycoprofile of 139A changed in oldfield mice similarly to that previously observed in voles , with the di-glycosylated band appearing the most prominent while it retained the mice-like pattern in gerbils ( Fig . 3 ) . Finally , since we observed that scrapie and BSE adapted to bank voles as much faster strains than in mice , we checked if this also applied to oldfield mice by setting up the second passage of SS3 in that species . The survival time of SS3 in oldfield mice was indeed short ( 103±11 d . p . i . ) and comparable to that observed in bank voles .
We showed that the rodents under investigation can be subdivided into three groups . The first included voles and the oldfield mouse and was characterized by: i ) high susceptibility to scrapie , ii ) low susceptibility to BSE , iii ) extremely short incubation times with adapted strains and iv ) change in the 139A glycoprofile . The second group comprised C57Bl/6 mice , wood mice and gerbils and displayed: i ) low susceptibility to scrapie , ii ) relatively high susceptibility to BSE , iii ) longer incubation times with adapted strains and iv ) no change in the 139A glycoprofile . The third group included only spiny mice which showed a distinctive resistance to prions . These findings were consistent with the inefficient transmission of natural scrapie to wild type mice reported by several authors [14] , [15] , but they also confirmed old observations by Chandler and Turfrey [16] , who reported that 50% of field voles inoculated with a rat- or a mouse-passaged scrapie isolate died after 2 . 5 months , well before the other rodent species which were also challenged . PrP sequence comparison indicated that Y154N and S169N correlated with the different transmission patterns observed . Overall , species with Y154–S169 were resistant to scrapie , permissive to BSE and reproduced a mouse-like phenotype when infected with 139A , while species with 154N–169N displayed rather opposite features . The inverted susceptibility of rodents to scrapie and BSE underlined the role of strains in the transmission barrier: amino acid exchanges could either enhance or reduce the efficiency of transmission , depending on the prion strain . In particular , we showed that Y154N–S169N exchanges , which appeared to confer in vole-related species a high susceptibility to scrapie , had the opposite effect with BSE . This is concordant with in vitro studies showing that the alteration of the conversion efficiency induced by Y154N–S169N mutations in the vole PrP is strain-dependent , leading to differential effects with vole-adapted BSE and scrapie [9] . The change in the 139A glycoprofile further corroborated the distinction between vole- and mouse-related species . It is known that the PrPSc glycoform pattern is not necessarily preserved upon interspecies transmission [7] , [17] . This may suggest that the glycoprofile is a phenotypic characteristic which is not intrinsic to strains , but it might also reveal a more general phenomenon of strain components selection during interspecies transmission [18] . The recovery of the original 139A glycotype after back passage from voles to mice demonstrated that the change did not imply a permanent mutation , thus suggesting a possible direct effect of Y154N and S169N variations in the PrP sequence of the recipient species on this strain-related characteristic . This confirms previous observations that the glycosilation pattern of PrPSc can be also influenced by the host [7] , [17] . Piening et al [9] analyzed the role of PrP sequence by an in vitro conversion assay in the aim to investigate the basis of the higher susceptibility of bank voles to natural scrapie in comparison to mice . In agreement with our in vivo results , in vitro studies identified the Y154N and S169N substitutions as being responsible for the different conversion efficiency obtained with mouse and vole PrPC . Notably , ovine and murine PrP have the same amino acids at positions 154 and 169 , while bovine PrP differs only at codon 154 , having H instead of Y . In agreement with such differences , in vitro assays showed that the vole PrPC was less efficiently converted than that of mouse by both scrapie and BSE [9] . However , by introducing the murine double mutation 154N–169N into the bank vole sequence , the conversion efficiency was enhanced up to a level comparable to the efficiency achieved with mouse PrPC , irrespective of remaining mismatches at residues 108 and 226 . These findings suggested that the similarity at positions 154 and 169 represented a major determinant of species barrier between the above species . Nevertheless , the different conversion efficiency of mouse and vole PrPC by sheep scrapie did not correlate with the in vivo susceptibility of the two species . Assuming that the conversion of PrPC is caused by a direct interaction with PrPSc [19] , in vitro studies implied that the recognition and conversion of mouse PrPC by sheep PrPSc were more efficient than those of vole PrPC . However , our in vivo results suggested that other factors subsequent to such interaction might have influenced the pathogenesis , leading voles to develop the disease more easily than mice . On this basis , it is tempting to speculate that voles allow a particularly efficient adaptation and/or rapid replication of prions , as suggested also by the unusually short incubation times of adapted strains . This latter was a striking feature of bank vole , given that almost all vole-adapted prions showed survival times ranging from ∼35 to 130 d . p . i . , irrespective of whether they derived from humans , cattle , sheep , deer , mice , or hamsters ( [7] , [9]; Agrimi , unpublished observations ) . Herein , we showed that this feature also applies to field voles and oldfield mice . In these rodents the second passage of SS3 produced survival times comparable to bank voles . Furthermore 139A induced disease with survival times shorter than in the donor species , C57Bl/6 mice , even upon primary transmission . The Syrian hamster model has represented a major advance in prion research owing to the extremely short incubation period of the hamster-adapted strain 263K [20] . Interestingly , both hamsters and voles are 154N–169N . However , the comparison of their susceptibility leads to contrasting observations . It is known that hamsters resist BSE challenging [21] , similarly to voles . Furthermore we found that the glycoprofile of the 139H hamster strain , which derived from mouse 139A [22] , is shifted toward diglycosylated PrP , similarly to voles ( data not shown ) . On the other hand , our attempt to transmit SS3 to hamsters was unsuccessful [9] . The presence of amino acids that are unique to hamster species ( V111M , I138M , V202I , M204I , V214T ) offers a potential explanation for these discrepancies . Indeed , at least in the case of 138M , Priola and Chesebro [23] demonstrated in a cell system that this single hamster-specific residue could influence the transmission barrier between mouse and hamster by blocking the conversion of PrP . The molecular basis of interspecies transmission and adaptation of prions are unknown . Nevertheless , evidence suggests that the PrP sequence of the recipient species acts by dictating the range of possible PrP conformations and hence conditioning the susceptibility to different prion strains [3] . According to this model , the vole sequence would be particularly prone to adopting a wide range of conformations . This would explain the high susceptibility of voles to a variety of TSEs upon primary transmission , although with important exceptions such as BSE . In agreement with the low efficiency of transmission of BSE , also vCJD , which derives in humans from infection by the BSE agent , showed in bank voles very low transmission rate and extremely long survival time ( Agrimi , unpublished observations ) . This supports the idea that the BSE agent transmits poorly to species carrying the Y154N–S169N substitutions , irrespective of the PrP sequence of the donor species . Positions 154 and 169 are quite variable among mammalian PrPs . Human and bovine sequences are 154H–169S , sheep and goat 154Y–169S , elk and deer 154Y–169N . Considering the strain-related effect of variations at these positions , it could be speculated that such differences may account for the apparent limitation of prion interspecies transmission observed among humans , cervids and small ruminants . Actually , the only TSE proven to have crossed a species barrier naturally is BSE , which transmitted from cattle to humans; two species that share the same amino acids at positions 154 and 169 . In the three-dimensional representation of mouse PrPC , residues 154Y and 169S , corresponding to 154N and 169N of the vole prion protein , are exposed on the protein's surface ( Fig . 7 ) and are therefore accessible for potential interactions with PrPSc . Interestingly , it has been shown that also in a model of PrPSc based on electron micrographs of two-dimensional crystals [24] , 154Y–169S residues are located on accessible surfaces of the β-helical core structure potentially important for PrPSc-fibril formation [9] . Interestingly , position 169 lies in the loop connecting the second β sheet and the second α helix ( β2-α2 ) ( Fig . 7 ) , a region which is critical in conditioning the PrPC three-dimensional structure [25] , the formation of fibrils [26] , the susceptibility of sheep to scrapie [27]–[29] , the replication of prions [30] and the transmission barrier ( [11] , [19] , present paper ) . Furthermore , 169N has recently been identified as controlling the conformational plasticity of the β2-α2 loop [31] . As a whole , these data highlight the relevance of these positions when modelling the interspecies barrier . For instance , this could be significant when estimating the risk of prions for humans in primate models , which show a high variability at position 154 and in the β2-α2 loop , including position 169 . The distinction between vole- and mouse-related species inferred by transmission studies is paralleled by the taxonomy , which classifies voles and oldfield mice in the family of Cricetidae , while the remaining species in that of Muridae [12] . This suggests the need to consider the possible existence of host factors in addition to PrP which differently modulate the transmission barrier in the Cricetidae and Muridae families . Finally , our study showed that the range of rodent models with improved susceptibility to TSEs is wider than it has appeared in studies up to date . Moreover , the high susceptibility of voles and oldfield mice to TSEs gave rise to questions about the possible role of wild rodents in the natural spread of animal TSEs suggesting an intriguing field for epidemiological investigations .
Bank voles ( Myodes glareolus , formely Clethrionomys glareolus ) , field voles ( Microtus agrestis ) , wood mice ( Apodemus sylvaticus ) , oldfield mice ( Peromyscus polionotus ) and spiny mice ( Acomys cahirinus ) were obtained from breeding colonies at the Istituto Superiore di Sanità , Rome , Italy . Mongolian gerbils ( Meriones unguiculatus ) and house mice ( C57Bl/6 ) ( Mus musculus ) were purchased from Charles River ( Como , Italy ) . The research protocol was approved by the Service for Biotechnology and Animal Welfare of the Istituto Superiore di Sanità and authorized by the Italian Ministry of Health , according to Legislative Decree 116/92 , which implemented the European Directive 86/609/EEC on laboratory animal protection in Italy . Animal welfare was routinely checked by veterinarians from the Service for Biotechnology and Animal Welfare . Subjects were individually identified by passive integrated transponders , inoculated when weanlings ( 40–60 days ) and kept in groups of two-four individuals per cage . Scrapie-infected brain tissue ( SS3 ) was obtained from the thalamus of a naturally-affected sheep of Sarda breed from Tuscany , which carried the AA136RR154QQ171 PrP genotype . The mouse-adapted scrapie strain 139A was kindly provided by Prof . M . Pocchiari ( Istituto Superiore di Sanità ) . The BSE inoculum was prepared from the medulla oblongata of clinically-affected cattle diagnosed in Italy in 1994 . All inocula consisted of 10% ( w/v ) brain homogenate in sterile saline . Animals were anaesthetized with ketamine and inoculated intracerebrally ( i . c . ) into the left hemisphere with 20 µl brain homogenate . Beginning one month after inoculation , animals were examined twice per week until the appearance of clinical symptoms , and then examined daily . We measured the survival time instead of the incubation time because of the differences among species in the clinical phenotype of the disease . Diseased animals were sacrificed with carbon dioxide at the terminal stage of disease but before neurological impairment was such as to compromise welfare and , especially , adequate drinking and feeding . Survival time was calculated as the interval between inoculation and sacrifice or death . After collection at sacrifice , each brain was cut parasagitally into two parts . The smaller one was stored at −80°C for biochemical studies . The other part was fixed in formalin for histology and immunohistochemistry analysis as described previously [7] . Total PrP as well as PK-resistant PrP were examined by Western blotting in SDS-PAGE gels , as previously described [7] . Genomic DNA was extracted from frozen brain samples using standard procedures . The coding region of the PrP gene from each species was amplified from 100 ng of genomic DNA using the polymerase chain reaction ( PCR ) . PCR reactions were performed with either MoPrP5 ( TGGGCACTGATACCTTGTTCCTC ) and MoPrP3 ( CCCAGCCTAGACCACGAGAATG ) primers ( wood mouse ) or PrP5uni ( TYAGYCATCATGGCRAACCTTRGC ) and PrP3uni ( TCATCCCACBATCAGGAAGATGAG ) ( bank and field voles ) . The latter primers were moderately degenerated on the basis of known rodent PrP sequences and located within the coding region of the PrP gene . The purified PCR products were re-amplified in a ‘nested’ PCR to attach sequences corresponding to standard sequencing primers . The re-amplified products were cycle sequenced using Thermo sequenase ( Amersham Pharmacia , Freiburg , Germany ) and 5′-IRD-800 labelled primers according to the manufacturer's recommendations . Sequences were determined with the help of an automated system ( Model 4000L , LI-COR , Lincoln , NB ) . The spiny mouse , the oldfield mouse and the bank vole PrP coding regions were successfully amplified with primer C1-for ( TGTAAAACGGCCAGTCCTCATTTTGCAGATCAG ) and C1-rev ( CAGGAAACAGCTATGACCGGTCCTCCCAGTCATTGCC ) or with C2-for ( TGTAAAACGACGGCCAGTGGCACCCACAATCAGTGG ) C2-rev ( CAGGAAACAGCTATGACCCACGATCAGGAAGATGAG ) . Details on the primers and PCR conditions are available from the authors upon request . PCR products were purified and sequenced with the Big Dye primer cycle sequencing kit ( Applied Biosystems , CA , USA ) . Sequences were determined with an ABI Prism 310 apparatus ( Applied Biosystems ) . The PrP sequence of gerbils was obtained from GenBank ( AF117314 ) . The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession numbers for the prion proteins discussed in this paper are: C57Bl/6 mouse ( M18070 ) , wood mouse ( AF367623 ) , spiny mouse ( EF467171 ) , bank vole ( AF367624 ) , field vole ( AF367625 ) , oldfield mouse ( EF467170 ) , Mongolian gerbil ( AF117314 ) , Syrian hamster ( M14054 ) . The Protein Database ( PDB ) accession number of mouse PrPC in Figure 7 is 1AG2 .
|
Prions are unconventional infectious agents that cause fatal neurodegenerative diseases in animals and humans . A pathological form of the cellular prion protein ( PrPC ) , named PrPSc , appears to be the major or the sole component of prions . These agents are transmitted by inducing the conversion of host PrPC into PrPSc that accumulates in the brain of affected individuals . Different factors are believed to modulate such events , which explains the variable transmission efficiency observed under inter-species experimental inoculation . These factors are still fairly unknown , although evidence exists that some kind of structural compatibility between PrPSc of the infectious inoculum and PrPC of the host has a role in making transmission more or less efficient . We investigated the transmission of prions to different rodents and showed that specific amino acid substitutions ( Y154N and S169N ) in the prion protein are major determinants of susceptibility to prions . In particular , we showed that these specific variations i ) direct the transmission rate of prions between different species in a way that is dependent on the prion strain , ii ) affect the molecular characteristics of prions , and iii ) influence their replication efficiency .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"infectious",
"diseases/prion",
"diseases",
"pathology/histopathology",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration"
] |
2008
|
Prion Protein Amino Acid Determinants of Differential Susceptibility and Molecular Feature of Prion Strains in Mice and Voles
|
The adaptation of Chikungunya virus ( CHIKV ) to a new vector , the Aedes albopictus mosquito , is a major factor contributing to its ongoing re-emergence in a series of large-scale epidemics of arthritic disease in many parts of the world since 2004 . Although the initial step of CHIKV adaptation to A . albopictus was determined to involve an A226V amino acid substitution in the E1 envelope glycoprotein that first arose in 2005 , little attention has been paid to subsequent CHIKV evolution after this adaptive mutation was convergently selected in several geographic locations . To determine whether selection of second-step adaptive mutations in CHIKV or other arthropod-borne viruses occurs in nature , we tested the effect of an additional envelope glycoprotein amino acid change identified in Kerala , India in 2009 . This substitution , E2-L210Q , caused a significant increase in the ability of CHIKV to develop a disseminated infection in A . albopictus , but had no effect on CHIKV fitness in the alternative mosquito vector , A . aegypti , or in vertebrate cell lines . Using infectious viruses or virus-like replicon particles expressing the E2-210Q and E2-210L residues , we determined that E2-L210Q acts primarily at the level of infection of A . albopictus midgut epithelial cells . In addition , we observed that the initial adaptive substitution , E1-A226V , had a significantly stronger effect on CHIKV fitness in A . albopictus than E2-L210Q , thus explaining the observed time differences required for selective sweeps of these mutations in nature . These results indicate that the continuous CHIKV circulation in an A . albopictus-human cycle since 2005 has resulted in the selection of an additional , second-step mutation that may facilitate even more efficient virus circulation and persistence in endemic areas , further increasing the risk of more severe and expanded CHIK epidemics .
The potential of RNA viruses to emerge into new environments often depends on their ability to efficiently adapt to new hosts . These adaptations sometimes comprise a stepwise process that includes 1 ) initial viral introduction/establishment in the recipient species , followed by 2 ) finite adjustment/optimization of the virus replication and transmission strategies for specific environments associated with a new host [1] , [2] . This process has been well documented for several single-host viruses such as pandemic influenza A virus , the SARS coronavirus and canine parvovirus ( reviewed in [3] , [4] ) that do not rely on alternating infection of disparate hosts for their maintenance in nature . However , much less is known about the adaptive processes that mediate cross-species jumps for double-host viruses such as arthropod-borne viruses ( arboviruses ) . Several recent studies documented that the acquisition of a single mutation in an arbovirus genome can mediate their cross-species transfer [step ( 1 ) ] [5]–[8] . However , in none of these cases have subsequent , additional adaptive mutations been detected , posing the question of whether selection of second-step adaptive mutations is possible or necessary for these viruses to persist in nature . This information is critical for understanding and predicting the long-term consequences of pathogen emergence and maintenance in affected areas , which in turn could influence the development and success of targeted intervention strategies for managing outbreaks . A new lineage of Chikungunya virus ( CHIKV ) [arbovirus in family Alphavirus , genus Togaviridae] emerged in 2004 in Kenya and subsequently spread into many countries in the Indian Ocean basin [hence the name: Indian Ocean lineage ( IOL ) ] , causing devastating outbreaks of arthritic disease [9] . In India , IOL strains were first detected in December 2005 followed by extensive geographic expansion during 2006–2011 into 19 Indian states with a total number of human cases estimated in 2007 at between 1 . 4 and 6 . 5 million [10] , [11] . During 2006 , the states most affected by CHIKV were Karnataka and Maharashtra , with a subsequent shift to Kerala , Coastal Karnataka and West Bengal [12] , [13] . Several hypothetical factors may have contributed to the CHIKV emergence/spread on the Indian subcontinent [14] , including: 1 ) the use of immunologically naïve human populations for maintenance , amplification and virus dispersal among localities , 2 ) reliance on peridomestic and anthropophilic mosquitoes as vectors , and 3 ) the IOL-specific genetic predisposition for rapid adaptation to Aedes ( A . ) albopictus , which was previously considered only a secondary CHIKV vector [9] . The mode of CHIKV maintenance in nature is complex and appears to be region-specific . In Africa , CHIKV is maintained in enzootic cycles involving transmission between non-human primates and canopy-dwelling , primatophilic Aedes mosquitoes , primarily A . furcifer , A . taylori , A . africanus , A . luteocephalus and A . neoafricanus [15]–[19] . In contrast , CHIKV transmission in Asia is believed to rely on humans alone as reservoir/amplification hosts , with the domestic A . aegypti and to lesser extent the peridomestic A . albopictus serving as primary urban mosquito vectors [19] , [20] . Recent evidence , however , suggests the possibility of additional sylvatic , zoonotic transmission cycles [21] , [22] . In India , both urban CHIKV vectors are present , although their distributions differ , and their epidemiologic significance for CHIKV transmission probably varies locally . A . aegypti was considered to be the most important during the early phase of the CHIK epidemic in 2006 [23] . However , in subsequent years ( 2007–2009 ) , the involvement of A . albopictus as the principal vector was documented at least in the states of Kerala and Coastal Karnataka [24]–[27] . Interestingly , CHIKV transmission by A . albopictus was shown to be associated with the acquisition of the A226V amino acid substitution in the E1 envelope glycoprotein [24] , [28]–[32] ( Figure S1 ) that is responsible for alphavirus virion assembly and virus fusion in endosomes of target cells [33]–[35] . The role of the E1-A226V substitution on CHIKV adaptation to A . albopictus was directly demonstrated in laboratory studies , including those using reverse genetics , showing that this mutation is directly responsible for increased CHIKV infection , dissemination and transmission by this vector species [6] , [36] . In India , evidence that CHIKV was undergoing genetic adaptation to A . albopictus via the E1-A226V substitution first came from Kerala State . During 2006 , only the E1-226A variant was recovered there; however , during subsequent years ( 2007–2008 ) , all isolates sequenced possessed the E1-226V residue [24] ( Figure S1 ) . In 2008 the E1-A226V substitution was also found among the majority of CHIKV isolates from Coastal Karnataka , adjacent to Kerala [37] , suggesting introduction from the latter state . In a follow-up study conducted in the state of Kerala , a novel substitution in the E2 envelope glycoprotein , L210Q , was discovered in all human and mosquito CHIKV isolates collected during 2009 [27] ( Figure S1 ) . The E2 protein is located on the tips of alphavirus spikes and interacts with host cell receptors as well as with neutralizing antibodies [38] , [39] . The L210Q substitution has not been reported in any other CHIKV strains , including those isolated in Kerala State during 2006–2008 . This suggests that E2-L210Q substitution was selected as a result of CHIKV adaptation to specific ecological conditions present in Kerala State . Position E2-210 is located in the domain B of the E2 glycoprotein [39] , and several earlier studies demonstrated that mutations in this domain mediate host specificity of several alphaviruses [5] , [7] , [40]–[42] as well as the selection of escape mutants by neutralizing antibodies [43]–[45] . Moreover , we recently demonstrated that epistatic interactions between mutations at positions E1-226 and E2-211 of CHIKV influence the penetrance of the E1-226V substitution for fitness in A . albopictus [46] . The E2-I211T substitution was probably acquired by IOL CHIKV strains around 2004–2005 [47] , and provides a suitable background to allow CHIKV adaptation to A . albopictus via the subsequent E1-A226V substitution . Considering that A . albopictus was a principal CHIKV vector in the state of Kerala in 2009 , it was hypothesized that the novel substitution E2-L210Q provided an additional selective advantage for CHIKV transmission by this mosquito [27] . To test this hypothesis we undertook a comprehensive reverse genetic analysis of the effects of E2-L210Q in various CHIKV hosts . Our observations demonstrate that the E2-L210Q substitution mediates a significant increase in CHIKV dissemination in A . albopictus by increasing initial infectivity for midgut epithelial cells . In addition , we show that the E1-A226V substitution has a significantly stronger effect on CHIKV fitness in A . albopictus than E2-L210Q , probably explaining the observed time differences required for selective sweeps of these mutations in nature .
To investigate the effect of the E2-L210Q substitution on CHIKV fitness in A . albopictus mosquitoes , we employed a reverse genetics approach based on the SL-CK1 strain of CHIKV ( hereafter abbreviated SL07 ) , isolated in 2007 in Sri Lanka [9] . Previous phylogenetic studies indicated that SL07 evolved from the Indian subgroup of IOL and represents one of the most closely related isolates to strains responsible for CHIKV outbreaks in India ( including the Kerala state ) [9] , [48] . The SL07 isolate was passed only twice on Vero cells since its isolation from a febrile patient , thus limiting the potential for cell culture-adaptive mutations that can artificially influence alphavirus fitness in vertebrate and/or mosquito hosts . The SL-07 strain has an alanine residue at E1 position 226 and a leucine residue at E2-210 , corresponding to prototype IOL strain introduced into India in late 2005 . Since the E2-L210Q substitution was only detected in CHIKV strains form Kerala that had previously acquired the A . albopictus-adaptive E1-A226V substitution [24] , single E1-A226V and double ( E1-A226V and E2-L210Q ) substitutions were introduced into an infectious clone ( i . c . ) , generated from the SL07 strain using site-directed mutagenesis . In addition , a clone with the single E1-A226V substitution ( SL07-226V ) was genetically marked by introducing a synonymous mutation 6454A→C that creates an ApaI restrictase site ( SL07-226V-Apa ) . Previously we demonstrated that the 6454A→C substitution does not influence CHIKV fitness in vitro or in vivo [6] . The infectious viruses SL07-226V-Apa and SL07-226V-210Q were rescued by electroporation of in vitro-transcribed RNA into Vero cells . The specific infectivity and viral titers in cell culture supernatants were almost identical for all constructs ( Table S1 ) , indicating that the introduced mutations did not attenuate CHIKV in Vero cells . Although a variety of mechanisms may be involved , adaptation of arboviruses for enhanced transmission by mosquitoes is typically expected to result in an increased ability to develop a disseminated infection leading to salivary gland infection . To investigate the effect of the E2-L210Q substitution on CHIKV fitness in A . albopictus mosquitoes , direct competition experiments were performed using SL07-226V-Apa and SL07-226V-210Q viruses ( Figure 1 ) . For these experiments , A . albopictus ( Thailand strain ) was presented with blood meals containing a mix of 5x105 plaque forming units ( pfu ) /mL of SL07-226V-Apa and 5x105 pfu/mL of SL07-226V-210Q viruses ( combined titer 106 pfu/mL ) and 10 days post-infection ( dpi ) , the presence of disseminated viral infection as sampled from individual mosquito legs and heads was analyzed as described in the Materials and Methods . The dissemination of the SL07-226V-210Q in the Thailand colony of A . albopictus was 4 . 3 times more efficient compared to SL07-226V-Apa ( Figure 1A ) ( p = 0 . 021 ) , supporting the hypothesis that glutamine at E2-210 was selected in CHIKV population in Kerala State due to its positive effect on CHIKV transmission . To corroborate these findings , the ApaI site was introduced into the backbone of SL07-226V-210Q , and the resultant virus ( SL07-226V-210Q-Apa ) was tested in direct competition in A . albopictus ( Thailand colony ) against SL07-226V that was produced by reverting the ApaI site in SL07-226V-Apa to the wild-type ( w . t . ) nucleotide sequence ( Figure 1B ) . The dissemination of SL07-226V-210Q-Apa was 3 . 4 times higher than that of SL07-226V ( p = 0 . 017 ) , indicating that the genetic marker was not responsible for the competition outcome , and supporting the role of the E2-L210Q substitution in increased CHIKV dissemination in A . albopictus . To demonstrate that the outcome of competition experiments was not affected by CHIKV propagation in Vero cells ( which were used to identify mosquitoes with disseminated infection , prior to CHIKV genotype analysis ) , these cells were infected at a multiplicity of infection of 0 . 1 pfu/cell in triplicate with 1∶1 mixtures of viruses that were used in mosquito competition experiments . At 2 dpi , cell culture supernatants were collected for viral RNA extraction and processed as described in the Materials and Methods . No detectable differences in viral fitness ( changes in the ratios of the 2 viruses ) were observed after Vero cell passage , indicating that E2-L210Q substitution does not affect CHIKV fitness in these cells ( Figure S2 ) . Early studies of CHIKV competence to infect A . albopictus demonstrated significant variation in susceptibility among different geographic strains of this mosquito [49] . To demonstrate that the effect of the E2-L210Q substitution on CHIKV dissemination in A . albopictus was not limited to a particular geographic strain , we also compared dissemination efficiency of the SL07-226V-Apa versus SL07-226V-210Q viruses in mosquitoes derived from Galveston , USA . Similar to the results with Thailand mosquitoes , the E2-L210Q substitution provided a mean 4 . 5-fold increase in dissemination efficiency of CHIKV ( p = 0 . 003 ) ( Figure 1A ) . These data suggest that the E2-L210Q substitution would likely have a similar effect on CHIKV fitness in A . albopictus from Kerala State , India , and from other parts of the world . To investigate if the fitness advantage associated with the E2-L210Q substitution is sufficient for selection of mutant viruses in a w . t . CHIKV population , the SL07-226V-210Q was serially passaged in the presence of a 100-fold excess of SL07-226V-Apa ( surrogate “wild-type” virus ) in an alternating cycle between A . albopictus mosquitoes and Vero cells . To initiate the cycle , A . albopictus ( Galveston colony ) were presented with blood meals containing 5x105 pfu/mL SL07-226V-Apa and 5×103 pfu/mL SL07-226V-210Q viruses ( 100-fold excess of SL07-226V-Apa ) . After three consecutive passages , heads and legs of individual mosquitoes were processed as described above to determine if selection of virus with E2-L210Q substitution had occurred ( Figure 2A ) . Despite being present in 100-fold lower quantity in the initial virus population , the SL07-226V-210Q virus alone was detected in 31 . 6% of mosquitoes after 3 alternating mosquito-Vero passages , whereas SL07-226V-Apa ( w . t . ) alone was found in 52 . 6% of mosquitoes , while 15 . 8% of mosquitoes had both competitors in their heads and legs ( Figure 2B ) ( p = 0 . 227 one-tailed McNemar test ) . These data indicate that the E2-L210Q substitution has the potential to be selected in CHIKV populations in locations where A . albopictus serves as the primary vector . The 31-fold increase over 3 artificial transmission cycles in the relative frequency of SL07-226V-210Q over its initial ratio in blood meals corresponds to a ca . 3-fold increase per cycle , which is in agreement with fitness advantage of the E2-L210Q substitution observed earlier in direct competition experiments ( Figure 1 ) . Historically , A . aegypti mosquitoes were the primary vector of CHIKV in Asia [19] , [20] , and this species still plays a significant role in CHIKV transmission in India [23] , [50]–[52] . To investigate if the E2-L210Q substitution also affects CHIKV fitness in A . aegypti , we analyzed the effect of E2-L210Q on CHIKV dissemination in this vector using competition experiments as described above . Since CHIKV transmission by A . aegypti has never been associated with the E1-A226V substitution , we first used the w . t . genetic background of the SL07 strain ( E1-226A and E2-210L ) to introduce the E2-L210Q substitution . Also , because A . aegypti is less susceptible to CHIKV than A . albopictus for strains with E1-226V , we used higher total oral doses up to 2 . 4×107 pfu/mL ( Figure 3A , 3B ) . The dissemination efficiency of SL07 and SL07-210Q-Apa viruses in A . aegypti ( Thailand strain ) were almost identical ( p = 0 . 395 ) ( Figure 3A ) . Similarly , no difference in the dissemination efficiency between SL07 and SL07-210Q-Apa viruses was detected in Galveston A . aegypti mosquitoes ( Figure S3 ) . Additionally , competition between SL07-226V-210Q-Apa and SL07-226V viruses , which express E2-210Q and E2-210L residues in the background of E1-226V , respectively , was also analyzed in Thailand A . aegypti , again revealing no statistically significant differences in dissemination efficiency [p = 0 . 402] ( Figure 3B ) . These data indicate that it is unlikely that the polymorphism at E2-210 affects CHIKV transmission by A . aegypti . Alternatively , E2-L210Q could have been selected as a result of CHIKV adaptation to a vertebrate host in India , probably humans . Although we did not observe any fitness change associated with this mutation during propagation of CHIKV in Vero cells ( derived from African green monkey kidneys ) , to extend our analysis we repeated competition experiments using the human-derived cell line 293 ( embryonic human kidney ) because earlier studies showed that CHIKV can infect and replicate in various primary human cell lines including epithelial , endothelial , fibroblast , muscle satellite and macrophages ( reviewed in [11] ) . No detectable difference in fitness resulting from the E2-L210Q substitution was observed in this cell culture model ( Figure 3B , 3C ) . Although cell lines are not ideal surrogates for in vivo infections , our data further support the conclusion that the E2-L210Q substitution was most likely selected only by A . albopictus . Previous studies determined that the A . albopictus-adaptive E1-A226V substitution acts primarily at the level of midgut infectivity . It was suggested that efficient CHIKV infection of and replication in midgut cells promotes more active CHIKV dissemination and transmission by this vector [6] , [36] , [46] , thus allowing the selection of A . albopictus-adapted CHIKV strains in nature . To explore which step during CHIKV infection of A . albopictus mosquitoes is affected by the E2-L210Q mutation , we first compared the relative ratios of SL07-226V-Apa and SL07-226V-210Q RNAs in whole mosquitoes , mosquito midguts and mosquito carcasses ( bodies without midguts ) after oral infection ( Figure 4A ) . We observed a marked increase in the relative amount of E2-210Q RNA in all samples analyzed , including midguts at 7 dpi . Furthermore , similar increases in the relative amount of E2-210Q RNA in mosquito midguts were observed as early as day 1 post infection , regardless of which of the two competitors was marked by the ApaI site ( Figure 4B , 4C ) . In contrast , no difference in the relative amounts of E2-210Q versus E2-210L RNAs were observed 2 days after intrathoracic infection of A . albopictus , when CHIKV titers peak ( Figure 4D ) . When injected intrathoracically , CHIKV does not require infection of and replication within mosquito midguts to disseminate to other organs and tissues , suggesting that the initial infection/replication of the midgut epithelium is a major site of selection of the E2-L210Q substitution in A . albopictus . To further test the hypothesis that the E2-L210Q substitution affects CHIKV fitness only during initial infection of the A . albopictus midgut , we first compared infection rates of mosquitoes presented orally with serial dilutions of the viruses expressing either E2-210L or E2-210Q residues in the backbone of the SL07 strain that has the E1-A226V substitution . The E2-210Q residue was associated with significantly higher infectivity ( p = 0 . 006 and p = 0 . 034 , Fishes exact test ) for A . albopictus ( Thailand ) at the blood meal titers of 3 . 5 and 2 . 5 Log10 ( pfu ) /mL , and the oral infectious dose 50% ( OID50 ) value calculated for SL07-226V-210Q was 8 . 9 times lower ( higher infectivity ) than that for SL07-226V ( Figure 5A , 5B ) . The lack of a significant difference in infection rates after the highest dose ( 4 . 54 Log10 ( pfu ) /mL ) probably reflected the oral dose nearing saturation . By way of comparison , earlier studies , including those using the SL07 strain , determined that the well-established A . albopictus-adaptive substitution E1-A226V mediates a much greater , ∼100-fold decrease in OID50 values [6] , [9] . To directly study the effect of E2-L210Q substitution on initial CHIKV infection of A . albopictus midgut cells , we developed a replicon/helper system for the SL07 strain . Sub-genomic replicons of alphaviruses can be packaged into virus-like particles ( VLPs ) by co-transfection of replicon and helper RNAs into susceptible cells [53] . The helper RNA provides the structural genes that package replicon RNA into VLPs , but the helper RNA itself is not packaged into the VLPs . Therefore , the VLPs are capable of primary infection and replicon RNA replication within cells , but cannot spread to neighboring cells due to the lack of the structural genes in the replicon . Thus , replicon VLPs allowed us to investigate the effect of mutations of interest on initial infection of midgut cells . Since transfection efficiency of viral RNA is critical in determining the efficiency of VLP production , we switched to BHK-21 cells that have superior RNA susceptibility compared to Vero cells . Earlier , we observed that CHIKV isolates that have not been passaged in rodent-derived cells lines ( including SL07 ) are impaired in their ability to replicate in BHK-21 cells ( KT , SCW , unpublished ) . Therefore , to ensure efficient recovery of CHIKV VLPs from BHK-21 cells , double BHK-adaptive substitutions ( nsP1-L407P and nsP3-T348A ) were introduced into the SL07 i . c . ( see Materials and Methods for details ) . Although these substitutions increase replication capacity , rather than electroporation efficiency , of CHIKV in BHK-21 cells , they have no effect on mosquito infection ( data not shown ) . The modified SL07 i . c ( contains nsP1-L407P and nsP3-T348A substitutions ) was subsequently used to generate all CHIKV replicons used in the mosquito infectivity study . The SL07 replicon expressing green fluorescent protein ( GFP ) was packaged into VLPs using w . t . SL07 helper ( with E2-210L and E1-226A residues ) or using a modified helper encoding E2-L210Q and E1-A226V substitutions . The SL07 replicon expressing cherry fluorescent protein ( CFP ) was packaged into VLPs using a helper encoding E2-210L and E1-226V residues ( Figure 6A ) . In addition , the ApaI marker was introduced into the GFP-expressing replicon . The infectious titers of all recovered VLPs , as determined by titration on Vero cells , were identical ( Figure 6A ) . Infection of Vero and C6/36 cells with 1∶1 mixtures of GFP and CFP expressing VLPs [based on infectious unit ( i . u . ) titers] yielded equal number of cells expressing these fluorescent proteins ( data not shown ) . For mosquito experiments , GFP- and CFP-expressing VLPs were mixed 1∶1 ( based on i . u . titers ) and presented in blood meals to A . albopictus as shown in ( Figure 6A ) . At 1 and 2 dpi , midguts of individual mosquitoes were dissected and analyzed by fluorescent microscopy to determine a number of cells expressing GFP and CFP in the same fields of vision ( Figure 6B ) . We found that , on average , midgut cells were 4–5 times more likely to become infected with VLPs expressing the E2-210Q residue as compared with VLPs expressing E2-210L ( Figure 6D , 6E ) . Similarly , 4–5 fold increases in relative amounts of E2-210Q RNA were observed after an ApaI digestion of RT-PCR amplicons derived from VLP-infected midguts ( Figure S4 ) . Infectious viruses were not recovered after infecting Vero cells with homogenates of 30 mosquitoes infected with VLPs mixes ( see Materials and Methods for details ) , indicating that the hypothetical formation of full-length viruses via recombination between helper and replicon RNAs , which could confound the interpretation of this experiment , did not occur . Altogether , these data demonstrate that the E2-L210Q substitution acts specifically by increasing initial CHIKV infectivity for midgut cells of A . albopictus . In a parallel experiment using VLPs , we also compared the effect of the previously characterized E1-A226V substitution on CHIKV infectivity for midguts of A . albopictus ( Figure 6A ) . The CFP-expressing replicon packaged using a helper encoding E2-210L and E1-226V residues ( CFP/E2-210L/E1-226V ) was competed against GFP-expressing replicon packaged using w . t . SL07 helper encoding E2-210L and E1-226A residues ( GFP/E2-210L/E1-226A ) . In contrast to the polymorphism at E2-210 , the valine residue at position E1-226 provided a far greater ( 41-43 fold ) increase in a midgut cell infection compared to the alanine residue at the same position ( Figure 6C , 6D , 6E ) , which agrees with previous results using infectious viruses [6] , [36] . These data also indicate that the results of experiments using VLPs with 2 different fluorescent reporter proteins ( GFP and CFP ) cloned into replicons RNAs are not influenced by those reporter proteins themselves . The significant difference of ∼10-fold between the effects of the polymorphisms at positions E1-226 versus E2-210 on CHIKV infectivity ( p = 0 . 026 and p = 0 . 005 for 1 and 2 dpi respectively ) ( Figure 6D , 6C ) indicates that the E1-A226V substitution exerts significantly stronger selection compared to E2-L210Q , and thus would be expected to be selected faster during CHIKV transmission by A . albopictus . To corroborate these findings we also analyzed effect of the E2-L210Q substitution on CHIKV infectivity for midgut cells of A . albopictus when this substitution is expressed in the background of w . t . CHIKV with the E1-226A residue . For this experiment , a GFP-expressing replicon was packaged using a w . t . SL07 helper encoding E2-210L and E1-226A residues ( GFP/E2-210L/E1-226A ) , and was competed against a CFP-expressing replicon packaged using a helper encoding E2-210Q and E1-226A residues ( GFP/E2-210Q/E1-226A ) . The E2-L210Q substitution caused a 2 . 3–2 . 4-fold increase in CHIKV infectivity for A . albopictus midgut cells ( Figure 7 ) , which was about 17 . 5 times weaker than the effect of the E1-A226V substitution in the same genetic background . Similarly , using direct competition experiments between infectious viruses SL07 and SL07-210Q-Apa ( both have the E1-226A residue ) we observed that the E2-L210Q substitution provided a mean 2 . 0-fold increase in dissemination efficiency of CHIKV ( p = 0 . 022 ) ( Figure S5 ) in the Thailand strain of A . albopictus . These data indicate that the E2-L210Q substitution would be selected more efficiently in CHIKV strains that previously acquired the E1-226V mutation .
In this study we showed that an E2-L210Q substitution recently identified in CHIKV populations of Kerala State , India , when expressed in the background of the initial adaptive E1-226V substitution , confers a selective advantage by increasing initial infection of A . albopictus midgut epithelial cells . Efficient infection of midguts promotes subsequent CHIKV dissemination into the hemocoel and transmission by this vector . However , the E2-L210Q substitution has no apparent effect on CHIKV fitness in the other primary mosquito vector , A . aegypti , or on fitness in cell culture models for primate infection ( Vero and 293 cells ) . These results as well as surveillance data indicating that CHIKV was transmitted primarily by A . albopictus in Kerala state of India when the E2-L210Q substitution was first detected [24]–[27] , provide a comprehensive evolutionary explanation for its appearance in 2009 . These results also indicate that adaptation of CHIKV to A . albopictus mosquitoes mediated by the previously characterized E1-A226V substitution was probably just a beginning of multi-step adaptive process that included the selection of a second ( E2-L210Q ) and possibly additional , future mutational steps by IOL strains now circulating in urban areas . These mutations , which have no deleterious effect on transmission by A . aegypti , will enable CHIKV to even more efficiently exploit urban transmission in environments populated by A . albopictus , but also to maintain the ability to utilize A . aegypti , which tends to occur in major urban centers [54] . Thus , our findings regarding the continued adaptation of CHIKV to A . albopictus raise serious public health concerns that even more efficient transmission may exacerbate the already devastating CHIK epidemics in India and Southeast Asia . Furthermore , the introduction of the E2-L210Q strain into new areas like Italy and France , where autochthonous cases have already occurred [55]–[57] , could spread epidemics into temperate climates where A . albopictus thrives . Considering the broad global distribution of A . albopictus , including nearly throughout the Americas , the E2-L210Q substitution may significantly increase the risk of CHIKV becoming endemic in additional locations . Interestingly , Niyas et al . ( 2010 ) demonstrated that CHIKV strains with the E2-L210Q substitution can be isolated from adult A . albopictus mosquitoes that were reared from wild-caught larvae collected in Kerala State , suggesting that transovarial transmission ( TOT ) may also play a role in CHIKV maintenance , especially during dry seasons [27] , [58] . Also , evidence suggests that TOT occurred in a small percentage of wild mosquitoes during recent CHIK outbreaks on Reunion Island , Madagascar , and in Thailand [59]–[61] . Although we did not attempt to study the effect of the E2-L210Q substitution on TOT , and at least one laboratory study failed to demonstrate TOT in A . albopictus of CHIKV strains with the E1-A226V substitution [58] , so the possibility that CHIKV mutations could influence rates of TOT warrants a thorough investigation . The molecular mechanism explaining the effects of the E2-L210Q substitution on CHIKV infectivity for A . albopictus midgut cells remains unknown . Earlier , we hypothesized that the E2 region around position 211 could be directly involved in interactions with a specific cell surface receptor [46] . We showed that the E2-211T residue mediates a significant increase in infectivity for A . albopictus in concert with the E1-A226V substitution , and that residue E2-211I , which is common among CHIKV strains , blocks this effect . Moreover , using virus overlay protein binding assays ( VOPBA ) to study CHIKV binding to the proteins associated with the brush border membrane fraction of A . albopictus midguts , we demonstrated that the E2-T211I substitution dramatically alters CHIKV interactions with as yet unidentified proteins ( KT unpublished ) . The recently determined crystal structure of the CHIKV E2 glycoprotein [39] provides additional insights into the possible involvement of residues E2-211 and E2-210 in interactions with a putative mosquito receptor ( Figure 8 ) . Both positions are located at the C'B sheet of the E2 protein , which is exposed on the virion surface on the lateral side of domain B , suggesting that these positions could be involved in interactions with cellular proteins . Substitutions of the aliphatic moieties with polar residues in this region may therefore be directly responsible for changing CHIKV affinities to as yet unidentified receptor ( s ) . Interestingly , positions E2-207 , E2-213 and E2-218 , which have been shown to be involved in VEEV adaptation to equine and mosquito hosts [5] , [7] , [42] , are also located in the same lateral surface of domain B , further supporting the hypothesis that E2-L210Q enables CHIKV to interact with a particular protein expressed on the surface of midgut cells . The studies to identify these protein ( s ) are underway . In the study by Niyas et al ( 2010 ) that discovered the E2-L210Q substitution in CHIKV strains from Kerala , only limited portions of CHIKV genomes including the nsP2 , E2 and E1 genes were sequenced [27] . Since we did not have an access to these isolates or to complete sequence of these strains , we cannot rule out the possibility that other genome regions could be influencing CHIKV evolution in Kerala State . Epistatic mutations in different genome positions can dramatically affect CHIKV infection of A . albopictus [9] . For example , the recently determined , lineage-specific epistatic interactions between positions E1-226 and E1-98 probably limited for at least 60 years the emergence and establishment of new CHIKV strains in Asian regions inhabited by A . albopictus [9] . This suggests that Kerala strains of CHIKV might have acquired adaptive substitutions in addition to E2-L210Q that promote efficient transmission in the human-A . albopictus cycle , and indicates the need for a more detailed , continuous molecular characterization of CHIKV strains from throughout its distribution . We also investigated if residue E1-226 has an epistatic effect on amino acid E2-210 . The E2-L210Q substitution was detected only in CHIKV strains that already acquired the E1-A226V substitution . We observed that the E2-L210Q substitution mediates a 4–5-fold increase in A . albopictus midgut infectivity when expressed in the background of E1-226V , whereas the same substitution caused only a 2 . 3–2 . 4-fold increase when expressed in the background of E1-226A ( Figure 6 and 7 ) . These results indicate that selection of this mutation would have been even less efficient if it had occurred in a CHIKV strain that did not yet acquire E1-A226V change . Interestingly , our data show that , with regard to CHIKV infectivity of A . albopictus midgut cells , E2-L210Q has an approximately 17-fold ( E1-226A background ) or 10-fold ( E1-226V background ) weaker effect compared with E1-A226V ( Figure 6 and 7 ) . This could explain why E1-A226V was selected convergently by unrelated CHIKV strains on at least 4 well documented occasions , while selection of E2-L210Q has thus far been observed only once in Kerala State ( Figure S1 ) . The stronger fitness effect of E1-A226V is consistent with its historically faster selection , which resulted in a selective sweep in parts of the Indian Ocean , India and Southeast Asia , compared with E2-L210Q , which has predominance in only one location . After CHIKV introduction into a region with large A . albopictus populations , the E1-A226V substitution has consistently taken about 0 . 5-1 year to appear [24] , [28] , whereas the E2-L210Q change was observed after at least 3 years of circulation in Kerala State [27] ( Figure S1 ) . More studies are needed to determine the precise dynamics of the selective sweeps associated with both mutations . Another interesting observation is that both A . albopictus-adaptive substitutions exert their effect on CHIKV fitness primarily at the level of midgut infectivity ( Figure 4 and 6 ) . The overall increase in the number of midgut cells infected with CHIKV VLPs expressing E2-210Q correlates with the increase in dissemination efficiency observed for infectious viruses . Also , the relative increase in the amount of E2-210Q RNA in midguts infected with VLPs is almost indistinguishable from the relative increase in amount of E2-210Q RNA in midguts exposed to infectious viruses ( Figure 4 and S4 ) . Although we did not examine replication in a comprehensive set of mosquito tissues , these results suggest that , after establishing an initial infection from the midgut lumen , the subsequent spread of viruses among neighboring cells is not influenced by position E2-210 . Moreover , no differences were observed in CHIKV replication in A . albopictus bodies after intrathoracic infection , indicating that replication of CHIKV in secondary mosquito organs also is unaffected by residue E2-210 ( Figure 4 ) . Similar observations were provided earlier for position E1-226 [62] . Experimental studies of epizootic versus enzootic VEEV VLP interactions with the epidemic vector , A . taeniorhynchus , also indicated that midgut epithelia is the target organ for VEEV adaptation to this vector [63] . These findings suggest that adaptation of alphaviruses to a mosquito vector primarily occurs at the level of midgut infection . In summary , we demonstrated that adaptation of CHIKV to a new mosquito vector can be a multistep process that , since 2005 , has involved at least 2 amino acid substitutions in the envelope glycoproteins . The substitution that provides the strongest selective advantage , E1-A226V , was followed by second adaptive mutation ( E2-L210Q ) that has resulted in a strain circulating in India with the fittest phenotype detected yet for transmission by A . albopictus . We hypothesize that this sequential adaptation will facilitate even more efficient circulation and persistence of the A . albopictus-adapted strains in endemic areas and will further increase the risk of expanded and more severe CHIK epidemics in new geographic ranges . This underscores the need for continued surveillance and studies of ongoing CHIKV evolution , as well as the molecular mechanisms that govern CHIKV adaptation to new environments .
The SL07 ( SL-CK1 ) strain of CHIKV was isolated in 2007 from a human in Sri-Lanka ( GenBank Acc . No . HM045801 . 1 ) . This strain belongs to Indian subgroup of the IOL [9] and was obtained from the World Reference Center for Emerging Viruses and Arboviruses ( WRCEVA ) at the University of Texas Medical Branch , Galveston , TX after its generous submission by Aravinda de Silva of the University of North Carolina . Since its isolation the strain was passed twice on Vero cells before being used for i . c . construction . Viral RNA was extracted from lyophilized virus stock using TRIzol reagent ( Invitrogen , Carlsbad , CA ) , reverse-transcribed using Superscript III ( Invitrogen , Carlsbad , CA ) and cDNA was amplified using Pfu DNA polymerase ( Stratagene , La Jolla , CA ) and PCR . To assemble the i . c . , overlapping RT-PCR amplicons were cloned into modified pSinRep5 vector ( Invitrogen ) under the control of an SP6 promoter using a strategy described previously for strain LR2006 OPY1 [64] . Point mutations 10670C→T ( E1-A226V ) , 9170T→A ( E2-L210Q ) and 6454A→C ( synonymous , ApaI marker ) were introduced in various combinations into the i . c . of SL07 using conventional PCR-based cloning methods [65] , and the PCR-generated regions were completely sequenced . Plasmids encoding sub-genomic replicons of strain SL07 were generated from the i . c . of the BHK-21 cell-adapted version of this strain [SL07-BHK that contains 1296T→C and 5087A→G ( nsP1-L407P and nsP3-T348A ) substitutions] which was reported previously [9] . These mutations were identified by electroporation of the SL07 i . c . into BHK-21 cells , followed by sequencing of the recovered , plaque purified viruses . Replicons were generated by replacing the structural gene region of SL07-BHK with the sequence of eGFP or CFP genes utilizing standard techniques [64] , [66] . In addition , a point mutation 6454A→C ( synonymous , ApaI marker ) was introduced into the pRep-GFP construct that allows comparison of the relative RNA quantities in an experimental , mixed infection sample . The helper plasmids were generated by deleting the 373–7270 nt . cDNA fragment from i . c . of SL07 that has mutations of interest at E1-226 and E2-210 . Plasmids were propagated using the MC1061 strain of E . coli in 2xYT medium and purified by centrifugation in cesium chloride gradients . Detailed information for all plasmids is available from the authors upon request . Vero cells ( African green monkey kidney ) were propagated at 37°C , with 5% CO2 , in Minimal Essential Medium ( MEM; Invitrogen , Carlsbad , CA ) supplemented with 5% fetal bovine serum ( FBS ) . BHK-21 ( S ) [Baby Hamsters Kidney] and 293 ( Human Embryonic Kidney ) cells were maintained at 37°C with 5% CO2 in MEM-alpha ( Invitrogen ) supplemented with 10% FBS and 1x MEM vitamin solution ( Invitrogen ) . The Galveston colonies of A . albopictus and A . aegypti mosquitoes were established from the mosquitoes collected in Galveston , TX ( USA ) . Thailand colonies of A . albopictus and A . aegypti mosquitoes were established from mosquito eggs collected in Bangkok , Thailand . All manipulations and handling of mosquitoes were done as described previously [67] . Infectious viruses were generated by electroporation of the in-vitro transcribed RNA into Vero cells . RNA was transcribed from SP6 promoter of the NotI linearized i . c . DNA using the mMESSAGE mMACHINE kit ( Ambion , Austin TX ) . Ten µg of RNA were electroporated into 107 Vero cells using a BTX-Harvard Apparatus ECM 830 Square Wave Electroporator ( Harvard Apparatus , Holliston , MS ) and 2mm cuvette at the following conditions: 680V , pulse length 99 µs , 5 pulses , with an interval between the pulses of 200ms . Cells were transferred to a 75 cm2 tissue culture flasks with 14 mL of Leibovitz L-15 ( L-15 ) medium supplemented with 10% FBS and 5% tryptose phosphate broth ( Sigma-Aldrich , St . Louis , MO ) . At 3 h post electroporation the cell supernatant was replaced with 14 mL of L-15 medium and maintained at 37 °C without CO2 . Cell culture supernatants were collected at 24 and 48 h and stored at −80°C . To estimate the specific infectivity of electroporated RNAs , an aliquot containing 1x105 electroporated Vero cells was serially ten-fold diluted and cells were allowed to attach to sub-confluent monolayers ( 1x106 cells/well ) of uninfected Vero cells in six-well plates [64] . After 2 h of incubation at 37°C , cells were overlaid with 0 . 5% agarose in MEM supplemented with 3 . 3% FBS and incubated for 48 h until plaques developed . The results ( specific infectivity values ) were expressed as pfu/µg of electroporated RNA ( Table S1 ) . Titers of the viruses recovered after electroporation and all experimental samples were determined by titration on Vero cells by plaque assay as previously described [7] . To generate CHIKV VLPs expressing residues of interest in E2 and E1 glycoproteins , BHK-21 ( S ) cells were used , which have superior RNA susceptibility compared to Vero cells . To ensure efficient recovery of CHIKV VLPs from BHK-21 cells , all CHIKV replicons were designed to include BHK-adaptive mutations ( nsP1-L407P and nsP3-T348A ) identified after rescue of w . t . i . c's in BHK-21 ( S ) cells . Ten micrograms of in-vitro transcribed replicon and helper RNA were mixed and electroporated into 107 BHK-21 ( S ) cells as described above for Vero cells . Cells were maintained in L-15 medium at 37°C , followed by harvesting supernatants at 30 h post-electroporation . The titer of VLPs was determined by titration on Vero cells as described earlier [68] . Briefly , 1×106 Vero cells were seeded in six-well plates and , after a 16 h incubation at 37°C , monolayers were infected with 10-fold dilutions of the samples for 1 h at 37°C , followed by adding 2 mL of MEM . After 24 h of incubation at 37°C the numbers of GFP- or CFP-expressing cells were quantified by fluorescent microscopy and titers were expressed as infectious units ( i . u . ) /mL . The role of viral mutations at position E2-210 on CHIKV dissemination in A . albopictus and A . aegypti mosquitoes was analyzed using direct competition experiment as described earlier [6] , [9] . A pair of viruses that differed by mutations of interest in the E2 protein was mixed at a 1∶1 ratio , with one of the viruses containing the ApaI marker . Viral mixes were used to prepare infectious blood meals by dilution in an equal volume of the defibrinated sheep blood ( Colorado Serum , Denver , CO ) , then orally presented to 4–5 day old female mosquitoes at 37°C as described previously [6] , [67] . Ten days post infection , heads and legs of individual mosquitoes were triturated in 500 µL of MEM media containing 5 µg/mL of Amphotericin B ( Fungizone ) , and 100 µL of clarified supernatant were added to duplicate wells of a 96-well plate containing 5x104 Vero cells/well . At 3 dpi , supernatant from virus-induced CPE ( cytopathic effect ) -positive wells was used for RNA extraction followed by RT-PCR with 41855ns-F5 ( 5`-ATATCTAGACATGGTGGAC ) and 41855ns-R1 ( 5`-TATCAAAGGAGGCTATGTC ) primers sets using One-Step RT-PCR kit ( Qiagen , Valencia , CA ) . The PCR products were digested with ApaI restrictase ( NEB , Ipswich , MA ) and separated on 1 . 5% agarose gels followed by ethidium bromide staining . One PCR band in the digested sample corresponded to disseminated infection for one out of two viruses in the pair; two bands indicated that both viruses disseminated in the same mosquito . Differences in dissemination efficiencies were tested for significance with a one-tailed McNemar test . Viral competition experiments with serial , alternating CHIKV passaging in A . albopictus and Vero cells were performed as described above with minor modifications . For the first passage , virus SL07-226V-210Q was mixed with 100-fold excess SL07-226V-Apa to generate infectious blood meals containing 5x105 pfu/mL ( combined ) . The blood meal was used for oral infection of A . albopictus ( Galveston colony ) followed by virus extraction from combined head and leg homogenates derived from 50 individual mosquitoes in 1 . 5 mL of MEM medium at 10 dpi . Homogenates were filtered and used to infect 75 cm2 flasks of Vero cells . At 2 dpi , cell culture supernatants were diluted 1∶10 in L-15 medium and mixed with equal volumes of defibrinated sheep blood to prepare a blood meal for the second passage . The cycle was repeated a total of 3 times . At 10 dpi of third mosquito passage , heads and legs of individual mosquitoes were processed as described above . For CHIKV competition experiments in specific body parts of A . albopictus , the mosquitoes were exposed to blood meals containing 1∶1 mixes of [SL07-226V-Apa and SL07-226V-210Q] and [SL07-226V and SL07-226V-210Q-Apa] . Depending on the experiment , at 1 , 2 , 3 and 7 dpi whole mosquito bodies , mosquito carcasses , or mosquito midguts were collected in pools of ten , and were used for RNA extraction using TRIzol ( Invitrogen , Carlsbad , CA ) . RNA was RT-PCR amplified , followed by ApaI restrictase digestion of amplicons as described above . Gel images were analyzed using TolaLab ( version 2 . 01 ) and relative fitness for a given virus during competition was determined as the ratio between E2-210L and E2-210Q bands in the sample , divided by the starting ratio of E2-210L and E2-210Q in the blood meal . The results were expressed as an average value of 2 pools of 10 mosquitoes midguts per pool . For CHIKV competition experiments in intrathoracically infected mosquitoes , 5 pfu of SL07-226V-Apa and 5 pfu of SL07-226V-210Q in 0 . 5 µL of L-15 media were directly injected into thoraxes of cold-anesthetized A . albopictus ( Galveston colony ) using capillary needles as described previously [69] . RNA from 2 pools , 5 mosquitoes/pool , was extracted at 1 and 2 dpi and processed as described above . To investigate the relationship between the blood meal titers and infection rates in A . albopictus , the SL07-226V and SL07-226V-210Q viruses individually were serially 10-fold diluted , mixed with defibrinated sheep blood and presented orally to A . albopictus ( Thailand ) . At 10 dpi individual mosquitoes were triturated in one mL of MEM and used to infect 5x104 Vero cells in duplicate in 96 well plates . CHIKV was detected by observing virus-induced CPE . The difference in the infection rates between SL07-226V and SL07-226V-210Q was tested for significance with a two-tailed Fishes exact test . The oral infectious dose 50% ( OID50 ) values were calculated using the PriProbit program ( version 1 . 63 ) . For VLP experiments , A . albopictus ( Thailand ) were infected with 1∶1 mixes ( based on i . u . titers ) of GFP- or CFP-expressing subgenomic replicons packaged into VLPs using CHIKV helpers that differed by substitutions at positions E1-226 and E2-210 ( Figure 6A and 7A ) . At 1 and 2 dpi , 5–10 mosquito midguts were dissected in PBS , and cut longitudinally to generate monolayers of epithelial cells . These sheets were rinsed in PBS to remove residual blood and gently spread out on a glass slide . A cover slip was applied and the midgut sheets were immediately analyzed by fluorescent microscopy to determine the numbers of cells expressing GFP and CFP in the same field of vision . One or two fields of vision were analyzed for each midgut sheet . In parallel experiment , midguts infected with VLPs packaged using helpers that differ by substitutions at position E2-210 were dissected at 1 , 2 and 3 dpi , collected in pools of ten , which were used for RNA extraction using TRIzol ( Invitrogen ) . The RNA was processed as described above . To demonstrate that replicon and helper RNAs did not recombine to generate infectious virus capable of autonomous replication , 30 mosquitoes were infected with VLPs mixes and at 7 dpi were triturated in 1 mL of MEM , filter sterilized and 300 µL of homogenate was used to infect each of 3 wells of confluent Vero cells in six-well plates . After 1 h of infection at 37°C , 2 mL of MEM was added to each well , followed by incubation at 37°C with 5% CO2 . Cells were observed daily for signs of CPE for 5 days . To investigate the effect of substitutions at E2-210 on CHIKV fitness in Vero and 293 cells , these cells were infected at a multiplicity 0 . 1 pfu/cell in triplicate with 1∶1 mixtures of [SL07-226V-Apa and SL07-226V-210Q] and [SL07-226V and SL07-226V-210Q-Apa] viruses . Cells were maintained at 37 °C with 5% CO2 in MEM and at 2 dpi , supernatants were collected for RNA extraction and processed as described above .
|
Since 2004 , chikungunya virus ( CHIKV ) has caused a series of devastating outbreaks in Asia , Africa and Europe that resulted in up to 6 . 5 million cases of arthritic disease and have been associated with several thousand human deaths . Although the initial step of CHIKV adaptation to A . albopictus mosquitoes , which promoted re-emergence of the virus , was determined to involve an E1-A226V amino acid substitution , little attention has been paid to subsequent CHIKV evolution after this adaptive mutation was convergently selected in several geographic locations . Here we showed that novel substitution , E2-L210Q identified in Kerala , India in 2009 , caused a significant increase in the ability of CHIKV to infect and develop a disseminated infection in A . albopictus . This may facilitate even more efficient virus circulation and persistence in endemic areas , further increasing the risk of more severe and expanded CHIK epidemics . Our findings represent some of the first evidence supporting the hypothesis that adaptation of CHIKV ( and possible other arboviruses ) to new niches is a sequential multistep process that involves selections of at least two adaptive mutations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"organismal",
"evolution",
"infectious",
"diseases",
"forms",
"of",
"evolution",
"microevolution",
"microbial",
"evolution",
"virology",
"biology",
"evolutionary",
"biology",
"microbiology",
"evolutionary",
"genetics",
"viral",
"evolution"
] |
2011
|
Sequential Adaptive Mutations Enhance Efficient Vector Switching by Chikungunya Virus and Its Epidemic Emergence
|
During lytic replication of Kaposi’s sarcoma-associated herpesvirus ( KSHV ) , a nuclear viral long noncoding RNA known as PAN RNA becomes the most abundant polyadenylated transcript in the cell . Knockout or knockdown of KSHV PAN RNA results in loss of late lytic viral gene expression and , consequently , reduction of progeny virion release from the cell . Here , we demonstrate that knockdown of PAN RNA from the related Rhesus macaque rhadinovirus ( RRV ) phenocopies that of KSHV PAN RNA . These two PAN RNA homologs , although lacking significant nucleotide sequence conservation , can functionally substitute for each other to rescue phenotypes associated with the absence of PAN RNA expression . Because PAN RNA is exclusively nuclear , previous studies suggested that it directly interacts with host and viral chromatin to modulate gene expression . We studied KSHV and RRV PAN RNA homologs using capture hybridization analysis of RNA targets ( CHART ) and observed their association with host chromatin , but the loci differ between PAN RNA homologs . Accordingly , we find that KSHV PAN RNA is undetectable in chromatin following cell fractionation . Thus , modulation of gene expression at specific chromatin loci appears not to be the primary , nor the pertinent function of this viral long noncoding RNA . PAN RNA represents a cautionary tale for the investigation of RNA association with chromatin whereby cross-linking of DNA spatially adjacent to an abundant nuclear RNA gives the appearance of specific interactions . Similarly , PAN RNA expression does not affect viral transcription factor complex expression or activity , which is required for generation of the late lytic viral mRNAs . Rather , we provide evidence for an alternative model of PAN RNA function whereby knockdown of KSHV or RRV PAN RNA results in compromised nuclear mRNA export thereby reducing the cytoplasmic levels of viral mRNAs available for production of late lytic viral proteins .
Kaposi’s sarcoma-associated herpesvirus ( KSHV ) is an opportunistic pathogen of human immunodeficiency virus ( HIV ) patients and the etiological agent of several human cancers , including Kaposi sarcoma and primary effusion lymphoma [1] . The KSHV life cycle includes a latent phase , when viral gene expression is largely absent and no progeny virions are produced , and a lytic phase , characterized by robust viral gene expression and virus replication . The most abundant lytic phase viral RNA is a 1 kb , noncoding , polyadenylated nuclear RNA called PAN [2–4]; PAN RNA accounts for up to 80% of the polyadenylated RNA present in a lytically infected cell . A 3′-end element called the element for nuclear expression ( ENE ) is required to maintain PAN RNA at these elevated levels . Crystallographic analysis of the PAN ENE complexed with an A9 oligonucleotide revealed that the U-rich internal loop of the ENE forms a triple-stranded interaction with the poly ( A ) tail of its own transcript [5] . This triple-helical RNA structure that shields the 3′ end [6–8] robustly inhibits nuclear RNA decay of PAN RNA . Although the ENE has been well studied , we still know little about the function of PAN RNA . Previous work has demonstrated that loss of PAN RNA , either through genetic knockout from the viral genome or antisense depletion of the transcript , results in misregulation of late lytic viral genes and host immune response genes [2 , 9] . The accompanying reduction in virion release upon KSHV PAN RNA knockdown highlights its essential role during the lytic phase , but the mechanism underlying the phenotypes associated with loss of PAN RNA is unknown . PAN RNA is exclusively nuclear , which prompted efforts to demonstrate that KSHV PAN RNA associates directly with the human and viral genomes [10 , 11] . A potential mechanism for PAN RNA function emerged from two studies: interaction of KSHV PAN RNA with viral latency-associated nuclear antigen ( LANA ) [12] and chromatin isolation by RNA purification ( ChIRP [13] ) studies of KSHV-infected human B-cells [10] . In vivo data suggested that PAN RNA may regulate chromatin states by competitively preventing LANA from associating with histone H3 , an interaction required for regulating and maintaining latency [12] . On the promoter of the KSHV master lytic activator , RTA/ORF50 , PAN RNA was shown to interact with demethylases JMJD3 and UTX [10] . KSHV PAN RNA ChIRP studies extended these conclusions to suggest ubiquitous PAN RNA binding to both the host and KSHV genomes [11] . Thus , the literature posits that PAN RNA regulates late gene expression by a mechanism dependent on chromatin association . Genes encoding PAN RNA homologs map to syntenic regions within gammaherpesvirus genomes . As is common for many long noncoding RNAs ( lncRNAs ) [14 , 15] , PAN RNAs are poorly conserved at the sequence level , but most contain ENEs homologous to that originally identified at the 3′ end of KSHV PAN RNA [16] . Bioinformatic studies revealed ENEs , thus indicating the presence of PAN RNAs , in four other γ-herpesviruses: retroperitoneal fibromatosis-associated herpesvirus Macaca nemestrina ( RFHVMn ) , Rhesus macaque rhadinovirus ( RRV ) , and equine herpesvirus 2 and 5 ( EHV-2 and EHV-5 ) [8] . A fifth PAN RNA homolog lacking an ENE was identified as a highly abundant noncoding RNA expressed from a syntenic genomic locus of bovine herpesvirus 4 ( BHV-4 ) [8 , 17] . RRV PAN RNA is a ~1 . 3-kb transcript found in the nuclei of lytic RRV-infected cells . As is the case for KSHV PAN RNA , the master herpesviral lytic activator ORF50 binds the promoter and activates expression of RRV PAN RNA [8] . Although RRV PAN RNA appears not to contain the small 5′-hairpin motif that binds viral ORF57 in KSHV PAN RNA , the RRV PAN RNA homolog is highly abundant in RRV-infected cells [8] . Moreover , similar to KSHV , RRV PAN RNA binds nuclear relocalized cytoplasmic polyA binding protein ( PABPC ) and is upregulated by the viral SOX protein in lytically infected cells [2 , 8] . KSHV and RRV have well-established cell culture models . The BCBL-1 cell line is a clinical isolate from a body cavity-based lymphoma and is a naturally KSHV-infected human B-lymphocyte cell line . A comparable RRV-infected rhesus B-cell line does not exist; all known rhesus B-lymphocyte lines were immortalized using another human herpesvirus , Epstein-Barr virus ( EBV ) . EBV and KSHV co-infections are known to collude and produce results different from those with KSHV alone [18] and similar interactions would be expected to occur between EBV and RRV . Therefore , RRV is studied instead in the human EBV-negative B-cell line , BJAB , which was de novo infected with RRV [19] . In this study , we demonstrate that knockdown of PAN RNA from the related herpesvirus RRV phenocopies that of KSHV PAN RNA knockdown . Furthermore , despite lacking nucleotide sequence conservation , expression of either KSHV or RRV PAN RNA can rescue production of progeny virions by either KSHV or RRV bacmids lacking the PAN RNA locus . Using CHART ( Capture Hybridization Analysis of RNA Targets ) [20] , we revisited the hypothesis that PAN RNA associates with specific chromatin loci to modulate gene expression during the KSHV lytic phase . Our analysis of PAN RNA from two related viruses , KSHV and RRV , suggests that the primary function of herpesviral PAN RNA is not regulation of gene expression by interacting with specific sites on chromatin . Instead , we find that PAN RNA expression is required for efficient mRNA export from the nucleus . These data emphasize the need for unbiased approaches for ascribing functions to ncRNAs .
To determine whether KSHV and RRV PAN RNA homologs perform the same function during the viral life cycle , we characterized the phenotype associated with knocking down RRV PAN RNA expression in lytic BJAB RRV cells . Anti-RRV antibodies useful for investigating the effect of PAN RNA expression on late lytic gene expression are not available at this time; however , quantitative PCR ( qPCR ) can assess viral DNA replication and virion production to confirm the downstream phenotypic changes associated with KSHV PAN RNA loss [2] . BJAB RRV cells were transfected with 2′-O-methylated and phosphorothioate-substituted antisense oligonucleotides ( ASOs ) that target RRV PAN RNA for RNaseH cleavage . 40 h after lytic induction with trichostatin-A ( TSA ) , total RNA was harvested from a subset of the cells and analyzed by Northern blot to confirm knockdown of RRV PAN RNA expression ( Fig 1A ) . Seven days later , the extracellular encapsulated and intracellular DNA were harvested from the remaining cells . Knockdown of RRV PAN RNA modestly reduced ( by 40% ) the yield of DNase-resistant , encapsulated virus released into the media , as assayed by qPCR of viral DNA ( Fig 1B ) . This phenotype mimics that of a KSHV PAN RNA knockdown [2] , but is less severe ( only 2-fold ) , perhaps attributable to less efficient knockdown of RRV compared to KSHV PAN RNA . In contrast , qPCR analysis of intracellular DNA confirms that knockdown of RRV PAN RNA does not affect accumulation of intracellular viral DNA during the lytic phase ( Fig 1C ) . Thus in B-cells , loss of RRV PAN RNA , like that of KSHV , decreases the production of encapsulated viral DNA without affecting the levels of intracellular viral DNA in B-cells . Loss of KSHV and of RRV PAN RNA expression both reduce production of encapsulated progeny virions , despite lacking nucleotide sequence conservation . Therefore , we asked whether one PAN RNA homolog could substitute for the other . Initially , we attempted to transiently express either RRV PAN RNA in KSHV-infected BCBL-1 cells or KSHV PAN RNA in RRV-infected BJAB cells after knockdown of endogenous PAN RNA . However , titrating the amount of transfected rescue plasmid ( from 2 to 15 μg ) revealed that the reciprocal PAN RNA homolog failed to rescue the loss of released encapsulated virus observed after PAN RNA knockdown ( S1A , S1D , S1F and S1I Fig ) . Yet , levels of intracellular viral DNA replication were unchanged ( S1E and S1J Fig ) . In KSHV-infected BCBL-1 cells , we similarly failed to observe rescue of several late lytic proteins upon transfection of an RRV PAN RNA rescue plasmid ( S1A Fig ) . Using RT-qPCR , we quantified the expression level of each PAN RNA homolog relative to the average expression level of five viral transcripts . Surprisingly , neither PAN RNA was expressed at greater than 1% of the endogenous species-matched PAN RNA in BJAB RRV or BCBL-1 PAN-knockdown cells ( S1B and S1G Fig ) . We conclude that the expression level of transiently transfected PAN RNA is insufficient to rescue the knockdown phenotype in BCBL-1 or BJAB B-cells . Unfortunately , we are unable to confirm this conclusion directly . Unlike the situation where protein-coding genes containing silent codon mutations are made , it is impossible to reintroduce a target RNA in the presence of ASOs without making point mutations in the rescue construct , which may have unanticipated functional effects . Consequently , we performed rescue experiments with PAN RNA knockout bacmids ( Fig 2 ) . We attempted to perform bacmid rescues in several physiologically relevant B-cell lines; however , the low transfection efficiency typical of B-cells in culture prevented expression of sufficient PAN RNA . As a result , knockout bacmid experiments were performed in HEK293T cells , as previously reported [21–23] . A bacmid containing a partial deletion of KSHV PAN RNA ( BAC36CRΔPAN ) was previously studied and , similar to the results of knockdown experiments , release of encapsulated virus into the media was diminished [10] . In the KSHV genome , 31% of the PAN locus overlaps the K7 open reading frame [24] ( Fig 3A ) , making an exclusive PAN RNA deletion impossible . In contrast , the RRV genome contains no known overlapping open reading frame , allowing a full RRV PAN RNA deletion [25] . To generate an RRVΔPAN bacmid , we eliminated 1300 bp from the wild-type RRV bacmid [26] encompassing the entire RRV PAN RNA transcript , 140 bp of upstream promoter sequence and 22 bp of downstream polyadenylation sequence ( S2 Fig ) . In lieu of the PAN RNA gene , a 1641-bp cassette was inserted that contains a kanamycin/neomycin resistance open reading frame to facilitate bacmid selection during recombineering ( S2 Fig ) . Next , we tested the ability of KSHV versus RRV PAN RNA to rescue the corresponding PAN RNA deletion bacmids . HEK293T cells were first transiently transfected with KSHV BAC36CRΔPAN bacmid [10] , the KSHV lytic transcriptional activator ORF50/RTA and either a KSHV PAN RNA rescue plasmid or a vector control . The vector control showed that the levels of encapsulated viral DNA released into the media were reduced ( ~40% ) and less K8 . 1 late lytic protein expressed 72 h after lytic induction with 600 μM valproic acid ( Fig 2A and 2D ) . Both phenotypes were rescued not only by expressing KSHV PAN RNA , but also by expressing RRV PAN RNA under the control of the KSHV PAN RNA promoter from a rescue plasmid ( Fig 2A and 2D ) . Using qPCR , we quantified the expression level of each PAN RNA homolog relative to the average expression level of five viral transcripts ( Fig 2B ) , revealing that KSHV and RRV PAN RNAs were each robustly expressed at approximately 2-fold the abundance of PAN RNA in BCBL-1 cells ( Figs 2B and S1B ) . Moreover , in contrast to what is observed in BCBL-1 cells , the levels of intracellular viral DNA replication were reduced in the vector control samples lacking PAN RNA ( Fig 2E ) . This is likely due to the reduced expression of early and late viral mRNAs ( Fig 2C ) , observed previously when the KSHVΔPAN bacmid was introduced into HEK293 cells [10 , 11] . Importantly , expression of RRV PAN RNA in the presence of the KSHV BAC36CRΔPAN bacmid rescued late lytic KSHV protein expression and yielded a similar number of extracellular progeny virions , compared to rescue with KSHV PAN RNA ( Fig 2C to 2E ) . We then performed the converse PAN RNA rescue in the context of the RRVΔPAN genome . HEK293T cells transiently transfected with RRVΔPAN bacmid were induced into the lytic phase by transiently transfecting rORF50/RTA in the presence of 100 nM TSA . The complete knockout of RRV PAN RNA from the bacmid reduced the level of encapsulated virus released into the medium in the presence of a vector control to about 10% relative to a RRV PAN RNA expression vector ( Fig 2I ) . Similar to what was observed for KSHV BAC36CRΔPAN , this phenotype was rescued ( 6- to 10-fold ) by expression of either KSHV or RRV PAN RNA ( Fig 2F to 2J ) . Together the data indicate that KSHV and RRV PAN RNAs , although very different in sequence , can substitute for each other’s function during the lytic phase of herpesviral infection . Previous studies using ChIRP suggested that KSHV PAN RNA associates broadly with host and viral chromatin [10 , 11] . We revisited this hypothesis using an alternative chromatin mapping technique known as CHART [20] to identify association sites for KSHV and RRV PAN RNA . Because these two lncRNAs can functionally substitute for one another , we reasoned that any chromatin interaction sites essential for production and release of progeny viral particles should be conserved . To reduce the high background common to both CHART and ChIRP technologies , we applied three strategies: ( 1 ) testing both KSHV and RRV PAN RNA homologs for similar sites of chromatin association; ( 2 ) exploring PAN RNA association as a function of time after inducing the lytic phase; and ( 3 ) using two independent capture oligonucleotide sets for each PAN RNA to limit background and selection of non-specific binding sites . Expression profiles of PAN RNA during the lytic phase were analyzed after induction of BJAB RRV cells or KSHV-carrying BCBL-1 cells with 500 nM TSA or 600 μM valproic acid , respectively; RNA was collected at multiple time points . For the CHART studies , we assessed PAN RNA-chromatin association when the lncRNAs were at 25% , 75% and 100% of the maximal expression level . Northern blot analysis of KSHV and RRV PAN RNA revealed that these values were achieved at 18 , 24 and 48 h for KSHV ( Fig 3B ) and 16 , 24 and 36 h for RRV ( Fig 3F ) . Candidate CHART capture oligonucleotides were designed according to specifications of the CHART protocol [20] and their ability to bind PAN RNA was determined by RNase H sensitivity assays . RNase H cleaves the RNA strand of RNA-DNA duplexes and thus a DNA oligonucleotide capable of hybridizing to PAN RNA produces a cleavage product that can be detected on a Northern blot . Schematics of the tested CHART PAN RNA oligonucleotides in Fig 3A and 3E illustrate the oligonucleotide binding sites on each of the two PAN RNAs . RRV PAN RNA was only moderately accessible ( Fig 3G ) , while KSHV PAN RNA was fully accessible to antisense oligonucleotide binding and cleavage by RNase H ( Fig 3C ) at all time points tested . This may reflect a difference in the complement of proteins bound to each PAN RNA . The pulldown efficiencies of RNase H-candidate CHART capture oligonucleotides were determined by incubating biotin-labeled oligonucleotides with streptavidin beads and nuclear lysate pre-cleared with unbound beads . After stringent high-salt washing , RNAs purified on the streptavidin beads were analyzed by Northern blot ( Fig 3D and 3H ) . To reduce background , two sets of pulldown oligonucleotides were chosen based on the criterion that pulldown efficiency be greater than 10% of the input sample . The exception to the use of two oligonucleotides was the KT493 oligonucleotide , which binds a repeat sequence in RRV PAN RNA and alone has an excellent pulldown efficiency ( Fig 3E and 3H , KT493 ) . The second oligonucleotide set for RRV PAN RNA consisted of tkv440 and tkv539 . The KSHV CHART oligonucleotide set 1 ( oligonucleotides E and K ) and set 2 ( oligonucleotides F and L ) each contained oligonucleotides that target regions near the middle and 3′ end of the transcript , avoiding the ENE stabilization element and not overlapping the K7 transcript ( Fig 3A ) . The pulldown efficiency was ~19% for oligonucleotide set 1 and ~22% for oligonucleotide set 2 ( Fig 3D ) . PAN RNA is observed as two bands on a Northern blot following the sonication step of the CHART procedure . The small fraction of PAN RNA isolated from the latent sample was generated by the 1–3% of cells undergoing spontaneous lytic reactivation . We evaluated PAN RNA association with chromatin for both KSHV and RRV using two separate antisense capture oligonucleotide sets and several time points representing different levels of PAN RNA expression . BCBL-1 cells containing KSHV and BJAB cells containing RRV were induced into the lytic phase with 600 μM valproic acid or 500 nM TSA , respectively , for up to 48 h prior to performing CHART enrichment of PAN RNA-associated genomic DNA . CHART peaks were called for each time point , relative to input , using MACS2 software ( see Materials and Methods ) . Any genomic locus called as a peak was then assessed for its enrichment score at the other three time points , regardless of whether the MACS2 software independently identified it as a peak in the dataset . Sites of chromatin association at each time point were identified as those genomic DNA loci that associate with PAN RNA in both oligonucleotide sets in at least one of two biological replicates and whose enrichment increased with respect to latent samples , which are largely devoid of PAN RNA ( Fig 3B and 3F ) . PAN RNA CHART peaks were then further separated into three categories: those with an enrichment that peaked at 16 , 24 or 48 h for KSHV , and at 18 , 24 or 36 h for RRV , respectively . We identified 1034 KSHV and 228 RRV PAN RNA CHART peaks ( Fig 4A and S1 Appendix ) . Only two peaks overlapped between the two PAN RNAs , chr5:1960322–1961985 and chr21:10730325–10731108 , albeit with different temporal patterns of enrichment ( Fig 4A ) . Neither of these genomic locations reside within a gene or genomic regulatory feature reported in the UCSC genome annotation . The lack of overlap between binding sites on host chromatin for the two PAN RNA homologs was not limited by the high-throughput CHART procedure because after isolating genomic DNA associated with KSHV PAN RNA , putative KSHV CHART peaks , but not RRV CHART peaks , could be verified as enriched by qPCR ( Fig 4B ) . Published ChIRP analyses of KSHV PAN RNA identified the viral ORF50 promoter as a specific site of chromatin interaction . Using CHART analysis , we also–to some extent–isolated this region; however , the peaks generated by the two KSHV oligonucleotide sets did not overlap and qPCR of CHART-isolated DNA showed enrichment only with one capture oligonucleotide set ( S3 Fig ) . This suggests that the KSHV ORF50 promoter is likely a region of open chromatin that is readily accessible for non-specific binding events–not a specific site of PAN RNA interaction . To evaluate whether the same protein factor might bind both PAN RNA homologs , but recruit the lncRNAs to separate , non-overlapping sites on the genome , we compared the PAN RNA CHART peaks to 275 ENCODE eCLIP ( enhanced crosslinking and immunoprecipitation ) datasets and 162 ENCODE ChIP ( chromatin immunoprecipitation ) datasets . Of the 276 factors represented in these two analyses , none overlapped with greater than 15% of either the KSHV or RRV CHART peaks ( Fig 4C and S2 Appendix ) . Additionally , no single factor significantly overlapped with both PAN RNA CHART datasets . Finally , we fractionationated lytic BCBL-1 cells and analyzed the distribution of PAN RNA between the nucleoplasm and chromatin by Northern blot ( Fig 4D ) . Surprisingly , KSHV PAN RNA could not be detected in the chromatin fraction . Three salts ( NaCl , LiCl and NH4Cl ) were used in the same fractionation protocol to minimize the chance that the absence of PAN RNA signal in the chromatin was due to the fractionation buffer . These salts lie on different points along the Hofmeister series–a classification that orders ions by their ability to stabilize and solubilize proteins [27] . The quality of fractionation was verified both by Western blotting and qRT-PCR ( Fig 4D and 4E ) . Although PAN RNA was undetectable in chromatin , unspliced transcripts and the ncRNA Kcnq1ot1 were chromatin-enriched as expected ( Fig 4E ) [28 , 29] , indicating that our fractionation protocol is capable of isolating chromatin-associated RNA . These data suggest that the majority of PAN RNA does not associate with chromatin . Loss of KSHV PAN RNA results in reduced K8 . 1 and ORF65 late lytic protein expression ( [2] , Figs 2 and S1 ) . Transcription of KSHV late lytic genes is facilitated by a viral pre-initiation complex ( vPIC ) encoded by the viral genome [30] . This six-component complex includes a TATA-binding protein ORF24 and five proteins of unknown function ( ORF18 , ORF30 , ORF31 , ORF34 , ORF66 ) . Viral genes transcribed by this complex contain a unique TATA box typified by the consensus sequence TATT [31 , 32] . PAN RNA does not associate with chromatin and is therefore unlikely to be an integral component of the vPIC . However , post-transcriptional or post-translational modification of vPIC components mediated by PAN RNA could alter the activity of the transcription factor complex . To test whether KSHV PAN RNA affects vPIC transcriptional activity , we ( 1 ) assessed whether knockdown of PAN RNA alters the steady-state mRNA level of any vPIC component; ( 2 ) analyzed vPIC-regulated transcript levels upon PAN RNA knockdown; and ( 3 ) monitored the influence of PAN RNA expression on a dual luciferase reporter system that expresses firefly luciferase from a vPIC-dependent promoter . At 48 h post lytic induction of BCBL-1 cells , when late lytic transcription is underway , qPCR analyses revealed that the steady-state levels of neither vPIC component mRNAs nor vPIC-regulated transcripts were significantly altered upon knockdown of PAN RNA , as compared to control knockdown samples ( Fig 5A and 5B ) . The approximate fold-change in mRNA levels of viral transcripts upon knockout of the vPIC component , ORF31 , are indicated by a dotted line ( Fig 5B ) [33] . If PAN RNA cooperates with the vPIC to facilitate transcription of the indicated target genes , we would expect a change in the levels of these transcripts to be similar in the absence of PAN RNA to that in the absence of ORF31 . To directly test whether PAN RNA expression affects the activity of the vPIC , we took advantage of a luciferase reporter system whose expression is activated only in the presence of all six viral transcription factors [30] . Upon transient transfection of this vPIC reporter and constructs encoding the six vPIC components in HEK293T cells , a 5-fold increase in luciferase activity was observed when the firefly luciferase gene was present downstream of the late lytic K8 . 1 promoter , but not downstream of the early ORF57 promoter ( Fig 5C ) . ORF50/RTA expression is required to transcribe PAN RNA from its endogenous promoter [34] . Control samples containing an ORF50/RTA or PAN RNA expression plasmid alone did not affect luciferase levels . When the ORF50/RTA and PAN RNA expression plasmids were both present , and therefore PAN RNA was expressed , the extent of luciferase activation by the vPIC remained unchanged ( Fig 5C; RTA+PAN ) . Although the components of the vPIC complex are conserved in RRV , a late lytic viral transcription factor complex has not yet been described . Together , these data suggest that the reduction in late lytic viral gene expression observed upon PAN RNA knockdown does not occur via the concerted action of this lncRNA and the vPIC . To determine how loss of PAN RNA results in misregulation of late lytic viral genes , we induced BCBL-1 TREx cells into the lytic phase for 24 h with 1 μg/mL doxycycline following knockdown of PAN RNA or , as a control , GFP mRNA . Doxycycline regulates expression of the inducible gene encoding the lytic activator RTA/ORF50 . As observed previously , knockdown of KSHV PAN RNA using two separate antisense oligonucleotides resulted in a specific reduction in late lytic viral protein expression ( K8 . 1 and ORF65 ) ( Fig 6A ) and in progeny virions released into the media ( Fig 6B ) , but no change in the intracellular viral genomic DNA copy number ( Fig 6C ) . However , when we isolated total RNA from the same cells and assessed mRNA levels by qRT-PCR , we did not observe a significant reduction in the mRNA encoding late lytic viral proteins upon loss of KSHV PAN RNA expression ( Figs 6D and S5 ) . This suggests that the decrease in late lytic protein expression observed upon PAN RNA knockdown is not the result of changes in transcription or RNA degradation rates . Effects on protein levels , without a corresponding change in RNA levels could be a consequence of reduced export of mRNA from the nucleus , reduced protein production , accelerated protein decay or any combination thereof . PAN RNA is predominantly , if not exclusively nuclear . Therefore direct effects on the cytoplasmic processes of protein production or decay are unlikely . Nonetheless , we assessed whether protein decay rates are altered in the absence of PAN RNA expression . After 48 h of lytic induction , BCBL-1 cells were treated with the translation elongation inhibitor cycloheximide . At 0 , 8 , 24 and 30 h of translation inhibition , equivalent numbers of cells were harvested and analyzed by Western blot for the early viral protein ORF6 , late viral proteins K8 . 1 and ORF65 , as well as the host protein GAPDH ( Fig 6G ) . Little difference in protein decay was observed upon knockdown of PAN RNA , as compared to control knockdown samples ( Fig 6H ) . This suggests that PAN RNA is not influencing protein degradation of late lytic viral proteins . To evaluate whether PAN RNA affects export of viral mRNA from the nucleus , we fractionated lytically-induced BCBL-1 cells following knockdown of either PAN RNA or , as a control , GFP mRNA . The quality of fractionation was assessed by Western blot and by qRT-PCR ( Fig 6E and 6F ) for a cytoplasmic marker ( GAPDH protein ) and nuclear markers ( FUS protein , unspliced actin RNA , unspliced GAPDH RNA ) . The fraction of both host and viral mRNAs decreased in the cytoplasm and increased in the nucleoplasm after knockdown of PAN RNA ( Fig 6F ) . This was true of both late and early lytic viral transcripts , as well as host GAPDH and actin mRNAs . Importantly , transcripts that are exported from the nucleus by alternative pathways , such as 5S rRNA [35 , 36] , SRP RNA [37] and HIST2AC mRNA [38] , are not affected by knockdown of PAN RNA . Fractionation of BJAB-RRV cells , despite evidence for a clean fractionation as assessed by Western blot ( S6A Fig ) , was prone to RNA leakage from isolated nuclei , which complicates analysis of these samples . Nonetheless , a reduction in nuclear mRNA export of viral mRNAs is also observed in BJAB-RRV cells following knockdown of RRV PAN RNA ( S6B Fig ) . We conclude that PAN RNA affects nuclear mRNA export during the lytic phase of viral infection .
We have shown that two herpesviral PAN RNA homologs are functionally interchangeable , despite lacking appreciable nucleotide sequence conservation . Knockdown or knockout of KSHV PAN RNA was previously shown to result in loss of late lytic protein expression , and consequently , a reduction in release of new virions into the surrounding media [2 , 10] . In this study , we demonstrate that knockdown or knockout of RRV PAN RNA likewise causes a reduction in release of encapsulated viral DNA ( Figs 1 and 2 ) . Furthermore , by expressing either RRV or KSHV PAN RNA from the appropriate herpesvirus species-matched PAN RNA promoter , both PAN RNAs are capable of restoring progeny virion release from HEK293T cells when co-expressed with either a KSHV or RRV PAN RNA knockout bacmid in HEK293T cells ( Fig 2 ) . RRV PAN RNA is also capable of rescuing the deficiency in late lytic protein expression observed with the KSHVΔPAN bacmid . In contrast to BCBL-1 and BJAB cells , the level of intracellular viral DNA produced from a PAN RNA knockout bacmid expressed in HEK293T cells is reduced . This suggests that in addition to lacking late lytic viral proteins required for packaging viral DNA , the viral DNA itself is not available for incorporation into progeny virions . This is likely due to a deficiency in robust lytic induction , as reported [10]; we also observed reduced expression of early mRNA , as well as late mRNA ( Fig 2 ) . Such subtle differences in phenotype associated with the loss of PAN RNA could be attributable either to cell type differences–a phenomenon that has been observed for other herpesviral gene knockouts and viral gene expression analyses [39–41]–or to the residual level of PAN RNA present after knockdown in B-cells , which may be sufficient to promote robust viral DNA replication . The herpesviral noncoding PAN RNA is essential to the viral life cycle , but the molecular mechanism by which this lncRNA acts remains unclear . Seven hypothetical models for PAN RNA function have been proposed [42–44] . Three of these posit that PAN RNA associates with chromatin [9–11] . We conclude that the amount of PAN RNA that purportedly associates with chromatin must be very small because we were unable to detect any KSHV PAN RNA in the chromatin following fractionation of BCBL-1 cells , despite detecting other known chromatin-associated RNAs ( Fig 4A and 4B ) . Nonetheless , because KSHV and RRV PAN RNA homologs can functionally substitute for each other , we tested whether direct association with chromatin loci pertinent for producing progeny virions is indeed conserved between these two herpesviruses . We developed a CHART scheme that involved two different capture oligonucleotide sets and three time points during the lytic phase for both herpesviruses ( Fig 3 ) . When the CHART peaks representing sites of PAN RNA chromatin association were compared between KSHV and RRV , we observed only two host chromatin loci in common between the two homologs ( Fig 4 ) . Neither of these chromatin loci lie within or near annotated genes , chromatin marks or repetitive elements , which makes their relevance questionable . The dearth of peaks that overlap between the KSHV and RRV PAN RNA CHART datasets was confirmed by qPCR analyses demonstrating that KSHV PAN RNA does not associate with chromatin loci identified as RRV CHART peaks ( Fig 4B ) . We tested whether the two PAN RNA homologs might be localized on separate , non-overlapping regions in the genome either by a DNA binding protein or an RNA binding protein associated with nascent RNA . We compared our PAN RNA CHART peak datasets to all ENCODE CHIP and eCLIP datasets available at the time of our analysis . Of the 264 proteins included in this analysis , none exhibited greater than 15% overlap with either of our datasets . The protein that showed the greatest overlap ( 14 . 3% ) with the KSHV PAN RNA dataset , but was nearly absent from that of RRV , was FOXK1 . The binding site for a different forkhead protein , FOXD3 , was enriched in PAN RNA ChIRP peaks [11] . Forkhead transcription factors are the key effectors of many essential signaling pathways [45] and bind the DNA consensus sequence GTAAACA , but flanking sequences or cofactors appear to influence binding of the different family members [46] . However , the correlation of KSHV PAN RNA with forkhead protein binding sites does not appear to be conserved for RRV PAN RNA . Previous work studying PAN RNA association with chromatin used a technique similar to CHART , known as ChIRP [13] . The ChIRP scheme used would identify any PAN RNA chromatin association sites , but genuine sites might be obscured by the high number of non-specific and off-target background peaks . The raw KSHV PAN RNA ChIRP data [11] are not available for direct comparison with the CHART data presented here , but from the published PAN RNA ChIRP gene list , we do not observe any overlap in chromatin loci between the two datasets . This lack of reproducibility reinforces our conclusion that PAN RNA does not associate with specific sites on host or viral chromatin . The KSHV ORF50 promoter was validated as a site of PAN RNA chromatin association as evidenced by the ability of the demethylases JMJD3 and UTX to interact with the ORF50 promoter only when PAN RNA is expressed [10]; however , we hypothesize that this result is a false positive . The ChIRP study reported that expression of ORF50 is severely reduced in the absence of PAN RNA [10] . The lack of association between PAN RNA-associated factors JMJD3 and UTX and the ORF50 promoter observed on the KSHVΔPAN bacmid could be attributable to reduced ORF50 expression , and hence the extent of open chromatin at this locus . The chromatin sites observed in both the ChIRP and CHART data likely represent non-specific enrichment of open chromatin regions that readily interact with the highly abundant nuclear PAN RNA . This is supported by the poor overlap of PAN RNA CHART peaks between datasets; for each time point , approximately 1% or fewer of the CHART peaks were called in all four datasets ( S4 Appendix ) . At this point we cannot rule out the possibility that PAN RNA globally reorganizes the nucleus through molecular crowding , resulting in occasional , transient , non-specific DNA interactions . It is well established that herpesviruses reorganize the host nucleus and concentrate viral DNA into replication compartments [47] . Together , our results suggest that association of PAN RNA with specific host or viral chromatin loci is neither the primary nor pertinent function of this lncRNA during the viral life cycle . Campbell and colleagues found that the knockdown of PAN RNA leads to the recruitment of LANA protein at viral promoters , an activity normally associated with latency , suggesting a role for PAN RNA in latent-lytic transitions and chromatin association [12] . However , our study indicates that PAN RNA does not associate with the viral chromatin above background levels . We interpret the effect of PAN RNA on LANA , despite in vitro binding studies , to be due to inhibition of virion production . The unpackaged state of newly synthesized viral genomes in the absence of PAN RNA may lead to increased accessibility for LANA binding . Although PAN RNA is considered exclusively nuclear , ribosome footprint profiling detected ribosome-protected regions of KSHV PAN RNA within three open reading frames ( ORFs ) at the 5′ end of the transcript , which overlaps with K7 [48] . Due to the extreme abundance of PAN RNA , if less than 1% of the PAN RNA transcripts escape the nucleus and are translated , a non-trivial amount of PAN protein could be produced [42] . Comparison of the sequence of KSHV and RRV PAN RNAs indicate that the ORFs are not conserved in nucleotide sequence , peptide sequence or peptide length ( S4A and S4B Fig ) . KSHV PAN ORF1 . 1 contains a putative signal peptide , which might permit the peptide to traverse the secretory pathway [48] . A peptide of different length and sequence could be expressed from RRV PAN RNA ( nts 682–786 ) that likewise contains a putative signal peptide ( S4C Fig ) . The coding potential of PAN RNA homologs remains a topic for further study . We suggest that the absence of PAN RNA during the viral lytic phase perturbs efficient nuclear export of mRNAs . A modest reduction in nuclear export of several viral and host mRNAs was observed when either KSHV or RRV PAN RNA was eliminated indicating that this is a conserved phenotype of PAN RNAs . Despite the change in the nuclear-cytoplasmic distribution of early viral transcripts , and host GAPDH and actin mRNA , we did not observe a change in corresponding protein levels . A reduction in mRNA export likely has a more pronounced effect on diminishing the level of late lytic viral proteins due to the absence of late lytic transcription and translation prior to PAN RNA expression . In contrast , expression of host transcripts and early viral transcripts may generate a protein population that remains unchanged for the duration of the PAN RNA knockdown experiments . Consistent with this hypothesis , the half-life of GAPDH has been estimated to be ~38 h–significantly longer than the time frame of our assays [49] . Whether PAN RNA directly interfaces with mRNA nuclear export machinery and why late lytic protein expression , specifically , is affected remains a topic of future research . We hypothesize that PAN RNA’s function is to associate with RNA binding proteins , such as PABPC [50] , that are relocalized from the cytoplasm to the nucleus during viral infection . In the absence of PAN RNA , such as in the case of a knockout or knockdown , an abundance of unsequestered RNA binding proteins mislocalized in the nucleus could interfere with nuclear mRNA export , viral replication and gene expression . A study of PAN RNA secondary structure in the cell indicates that some regions of KSHV PAN RNA are protected from chemical probing and may be sites of protein interaction [44] . PAN RNA represents a cautionary tale for the investigation of RNA association with chromatin , whereby sequence reads of DNA cross-linked to an abundant RNA give the appearance of specific interactions . Rather than the latest technological advances , well-controlled , unbiased approaches should be applied to determine the true biological role of the ncRNA under investigation . Protein-RNA co-purification studies may also be subject to false positive interactions . This has been especially true of PAN RNA , which–by virtue of its abundance and diffuse nuclear distribution–can transiently and non-specifically associate with many factors . Molecular tools available for the study of lncRNAs such as PAN RNA are still under development . Most lncRNAs are of low abundance and lack obvious nucleotide sequence conservation that aids in identification of essential sequence elements . Here , we have exploited the use of a ncRNA from a related virus to provide a genetically-tractable model system . Viruses thus offer a unique advantage over host ncRNAs for interrogating questions in RNA biology . PAN RNAs may be multifunctional and fulfill different roles at each stage of the lytic phase . Moving forward , carefully designed and innovative approaches are needed to expand insights into the multifaceted functions of gammaherpesviral PAN RNAs .
Body cavity based lymphoma-1 ( BCBL-1 ) [51] , BCBL-1 TREx-RTA ( gift from Jae Jung , USC ) and BJAB RRV cells [19] were maintained in RPMI supplemented with 1% L-glutamine and 20% FBS . Human embryonic kidney 293T ( HEK293T ) cells ( ATCC CRL-3216 ) were maintained in DMEM supplemented with 1% L-glutamine and 10% FBS . BCBL-1 cells were induced with 600 μM valproic acid , BCBL-1 TREx-RTA were induced with 1 . 5 μg/mL doxycycline , BJAB RRV cells were induced with 100 or 500 nM TSA , where indicated . For transfection of BCBL-1 and BCBL-1 TREx-RTA cells , 10 million cells were pelleted and washed once with media lacking serum . Either 2 nmoles of RNaseH targeting oligonucleotide or 15 μg of plasmid DNA were electroporated into 10 million cells in a 0 . 4 cm cuvette at 975 μF/210 mV . HEK293T cells were transfected with Mirus TransIT-293 reagent according to the manufacturer’s directions . Antibodies used are 1:1000 anti-FUS ( Proteintech Group , 11570-1-AP ) , 1:1000 anti-Histone H4 ( Upstate Cell Signaling Solutions , 07–108 ) , 1:2000 anti-GAPDH ( Sigma G8795 ) , 1:500 anti-KSHV K8 . 1 ( Advanced Biochemicals Incorporated ) and 1:1000 anti-KSHV ORF6 ( gift from G . Hayward at The Johns Hopkins University ) . Cells were treated with 100 μg/mL cycloheximide for the indicated times . The RRV wild-type bacmid was a kind gift from the Desrosiers lab ( University of Miami ) [26] . This bacmid was used by Genebridges to construct the RRVΔPAN bacmid , which was confirmed by direct sequencing and enzymatic digestion ( S2B Fig ) . Luciferase reporters and vPIC transcription factor expression plasmids were previously described [30] . KSHV PAN RNA and RRV PAN RNA fragments were assembled downstream of the KSHV PAN promoter using Gibson PCR assembly and cloned into the BamHI and SpeI sites of p4030-16TR [52] . KSHV PAN RNA was cloned downstream of the RRV PAN promoter using Gibson PCR assembly into pCDEF4-RRV PAN RNA [8] . Endogenous RNaseH cleavage was assayed as described [53] . Briefly , 1–5 x 107 cells were resuspended in 100 μL/107 cells of Sucrose Buffer I ( 0 . 32 M sucrose , 3 mM CaCl2 , 2 mM MgAc2 , 0 . 1 mM EDTA , 10 mM Tris-HCl pH 8 . 0 , 1 mM DTT , 0 . 5 mM PMSF and 0 . 5% ( v/v ) NP-40 ) . After centrifuging the lysate at 500 x g for 5 min at 4°C , the supernatant was removed . Nuclei were washed in 1 mL Sucrose Buffer I lacking NP-40 . After removing the supernatant , the nuclei pellet was resuspended in 20 μL/107 cells of Low Salt RNA Buffer ( 20 mM HEPES , pH 7 . 9 , 25% glycerol , 1 . 5 mM MgCl2 , 0 . 02 M KCl , 0 . 2 mM EDTA , 0 . 5 mM DTT , 0 . 5 mM PMSF ) . 1/5 volume of High Salt RNA Buffer ( 20 mM HEPES , pH 7 . 9 , 25% glycerol , 1 . 5 mM MgCl2 , 0 . 8 M KCl , 0 . 2 mM EDTA , 0 . 5 mM DTT , 0 . 5 mM PMSF , 1% NP-40 ) was added five times . The nuclear extract was then incubated at 4°C on a rotary platform for 20 min and diluted 1:2 . 5 with RNA Nuclear Diluent ( 25 mM HEPES pH 7 . 6 , 25% glycerol , 0 . 1 mM EDTA , 0 . 5 mM DTT , 0 . 5 mM PMSF ) . 10 μL 100 μM DNA oligonucleotides were added to 90 μL of nuclear extract and incubated for 30 min at 37°C . Reactions were stopped by adding 0 . 5 mL TRIzol and purified according to the manufacturer’s protocol . Samples were run on a 1 . 2% formaldehyde agarose for a northern blot analysis . CHART was carried out as described [20] . CHART capture oligonucleotides were designed according to instructions in [20] . The pulldown oligonucleotides consisted of OLIGO SEQ-3′ end-TEG and were ordered from IDT . Sequencing libraries were constructed by the Yale Center for Genomic Analysis ( YCGA ) and Yale Stem Cell Genomics Core using Illumina CHIP-Seq Sample Prep Kit . Sequencing was performed on Illumina HISeq 2500 or 2000 instruments . Each CHART time point was prepared in two biological replicates . CHART-seq data were deposited in the Gene Expression Omnibus ( GEO ) under the accession number GSE121268 . 50-bp deep sequencing reads were mapped using Bowtie2 default settings to the custom indices containing both the appropriate viral genome ( RRV accession #AF210726; KSHV accession #GQ994935 ) and the human genome . Reads that mapped to multiple genomic loci were removed using samtools ( samtools view input . bam | grep AS:i:0 | grep -v XS:i:0 ) . MACS2 ( https://pypi . python . org/pypi/MACS2 ) [54] was used to call peaks and determine the fold enrichment for each dataset . For each time point , peaks were required to be ( 1 ) called by the MACS2 software in at least one biological replicate for both CHART capture oligonucleotide datasets , ( 2 ) have an enrichment score at least 1 . 5-fold greater than the latent sample , and ( 3 ) be absent from the no-oligo control pulldown . For each peak that met these criteria , the average MACS2 enrichment score at that locus was determined for both biological replicates for each of the three lytic phase time points . Each CHART peak was categorized by the time point at which this average enrichment score was highest . CHART peaks from KSHV and RRV were designated as overlapping if the peak regions called by MACS2 overlapped by at least 100 bp . ENCODE [55] CHIP and CLIP datasets were downloaded from www . encodeproject . org and compared to the KSHV and RRV peak genomic loci using the BEDTools suite intersect sub-command [56] . Fractionation was performed on ice with pre-chilled buffers using either 36 h-induced KSHV-infected BCBL-1 cells or 43 h-induced BJAB-RRV cells . The washed and pelleted cells were resuspended in 100 μL of RLB buffer ( 10 mM Tris pH 7 . 5 , 140 mM NaCl , 1 . 5 mM MgCl2 , 10 mM beta-glycerophosphate , 0 . 5% Nonidet P-40 ) . Digitonin ( Sigma , D-1407 ) was added to a final concentration of 25 μg/mL while wheat germ agglutinin ( Sigma , L9640 ) was added to a final concentration of 1 μg/mL ( BCBL-1 ) or 1 mg/mL ( BJAB-RRV ) and the cells were incubated on ice for 5 min to allow cell permeation . For lysis , the cells were carefully layered over 300 μL of RLB buffer containing sucrose ( 10 mM Tris pH 7 . 5 , 140 mM NaCl , 1 . 5 mM MgCl2 , 24% ( wt/vol ) sucrose , 10 mM beta-glycerophosphate , 0 . 5% NP-40 ) and then centrifuged at 13 , 000 x g for 10 min at 4°C . After centrifugation , supernatants were transferred into a fresh microcentrifuge tube and designated as cytoplasmic fractions . The nuclear pellets were washed three times in 100 μL of RLB buffer and spun down at 400 x g for 4 min at 4°C; the first of which contained digitonin at a final concentration of 25 μg/mL . The pelleted nuclei were resuspended in 30 μL of NUN1 buffer ( 20 mM Tris pH 7 . 9 , 75 mM NaCl , 0 . 5 mM EDTA , 0 . 125 mM PMSF , 50% glycerol , 10 mM beta-glycerophosphate , 0 . 1 mg/ml tRNA , 1x protease inhibitor ) and incubated on ice for 5 min . The nuclei were then lysed by adding 300 μL of NUN2 buffer ( 20 mM Hepes pH 7 . 6 , 7 . 5 mM MgCl2 , 0 . 2 mM EDTA , 0 . 1 mg/ml tRNA , 0 . 3 M NaCl , 1 M urea , 10 mM beta-glycerophosphate , 1% Nonidet P-40 , 1 mM DTT , 1x protease inhibitor ) . After incubation on ice for 15 min with occasional vortexing and centrifugation at 15 , 000 x g for 15 min at 4°C , the supernatants were transferred to fresh microcentrifuge tubes as the nucleoplasmic fractions . The chromatin-associated pellets were washed three times with 50 μL of NUN2 buffer , spun at 15 , 000 x g for 4 min at 4°C , and 100 μL of chromatin buffer ( 50 mM HEPES pH 7 . 4 , 100 mM NaCl , 0 . 1% SDS , 0 . 5% sodium deoxycholate , 1% NP-40 , 1 mM DTT , 10 mM beta-glycerophosphate , 1x protease inhibitor ) added . The chromatin-associated pellet was then sonicated at 4°C for 10 cycles of 30 sec on , 30 sec off , using a Diagenode Bioruptor Pico sonication device . Input samples were sonicated using the same conditions . NaCl was replaced with LiCl or NH4Cl in all buffers for indicated fractionations . One half of each fraction was prepared for RNA and protein analysis . RNA was extracted with TRIzol per the manufacturer’s instructions . The purity of the resulting fractions was assessed by qPCR and Western blot . 9 x 104 HEK293T cells were seeded per well in a 12-well plate . Transfections were conducted using Mirus TransIT-293 transfection reagent with OPTI-MEM media 24 h post plating . pGL4 . 16 constructs ( pGL4 . 16 ORF57 and pGL4 . 16 K8 . 1; [30] ) and a Renilla control vector were used at a 1:1 molar ratio ( 300 ng /sample ) . Total DNA was kept constant in each sample by adding pBluescript SK+ . Viral transcription factor plasmids were transfected in equimolar amounts . Dual luciferase assays were conducted 24 h post transfection with the Dual-Luciferase Reporter Assay System ( Promega , #E1910 ) . The reagents were prepared according to the manufacturer’s instructions and measurements were conducted with a GloMax-Multi Detection System ( Promega ) . Luciferase measurements were recorded as the ratio of firefly to Renilla luciferase activity and normalized to sample 7 ( pGL4 . 16 K8 . 1 vPIC ) . Data presented were from four independent experiments , each using mean values of technical triplicate samples . RNA was purified with TRIzol and treated with RQ1 DNase ( Promega ) according to the manufacturer’s protocols . 1 μg of RNA was used to generate cDNA with random hexamer primers and Superscript III ( Invitrogen ) using the recommended protocol . cDNA was diluted 3-fold and 0 . 75 μL was analyzed in a 15-μL qPCR reaction using FastStart Essential DNA Green Master ( Roche ) SYBR reagent on a Roche Lightcycler 96 . RNA levels were normalized to RNaseP RNA levels , which do not change during lytic induction . To determine the extent of PAN RNA overexpression in HEK293T cells relative to BCBL-1 cells , PAN RNA levels were normalized to the average CT value of five viral transcripts ( ORF18 , ORF26 , ORF4 , ORF62 and ORF67A ) , which accounts for variances in viral genome copy number and induction efficiency between samples . For samples obtained from the same cell line , the fold-change in PAN RNA levels calculated using RNaseP RNA was comparable to the calculation using the five viral transcripts . Seven days after lytic phase induction , 3 million cells were pelleted , washed with PBS and resuspended in genomic DNA isolation buffer ( 100 mM NaCl , 10 mM Tris-Cl pH 8 , 25 mM EDTA , 0 . 5% SDS ) with 0 . 1 mg/mL proteinase K . The solution was incubated overnight at 40°C , phenol extracted , ethanol precipitated and diluted to 30 ng/μL prior to analysis by qPCR . The average signal from two primer pairs specific to the viral genome was normalized to the average signal from two primer pairs specific to the human genome . Seven days after lytic phase induction , 1 . 5 mL of supernatant was collected , passed through a 0 . 45 micron filter and incubated with 20 units/mL DNase One ( New England Biolabs ) for 1 h at 37°C . Proteinase K lysis buffer ( 0 . 75% SDS , 0 . 1 M NaCl; 50 mM Tris , pH 7 . 5; 10 mM EDTA , 0 . 1 mg/mL proteinase K ) was added to a final volume of 2 mL and then incubated at 40°C for 1 h . 1 ng/mL of a control plasmid ( psiCHECK-2 ) was added to each sample as a normalization control for loss of DNA during subsequent phenol chloroform extraction and ethanol extraction . DNA was resuspended in 15 μL of ddH2O ( resulting in 100-fold concentration ) and analyzed by qPCR . The average signal from two primer pairs specific to the viral genome was normalized to the signal from the control plasmid .
|
Herpesviruses produce noncoding RNAs , some of which are essential to the viral life cycle . One such noncoding RNA from Kaposi’s sarcoma-associated herpesvirus is the polyadenylated , nuclear ( PAN ) RNA , which is required for production and release of progeny virions from infected cells . In this study , we demonstrate that although lacking nucleotide sequence conservation , PAN RNAs from two related viruses–when knocked down–exhibit the same phenotype , loss of late lytic viral gene expression and progeny virion production . Moreover , they can functionally substitute for each other to rescue this phenotype . We demonstrate that , in contrast to published literature , the reduction in viral gene expression upon PAN RNA knockdown is not due to loss of PAN RNA association with conserved , specific chromatin loci , nor does PAN RNA expression affect the viral transcription factor complex required for generation of the late lytic viral mRNAs . We present data suggesting that PAN RNA instead serves as a binding platform to sequester cellular proteins that are mislocalized to the nucleoplasm . These herpesviral noncoding RNAs can serve as models for the mechanistic study of human noncoding RNAs .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"infographics",
"medicine",
"and",
"health",
"sciences",
"non-coding",
"rna",
"sequences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"messenger",
"rna",
"microbiology",
"nucleotides",
"charts",
"viruses",
"dna",
"viruses",
"rna",
"isolation",
"epigenetics",
"molecular",
"biology",
"techniques",
"chromatin",
"herpesviruses",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"genome",
"complexity",
"genomics",
"antisense",
"rna",
"chromosome",
"biology",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"kaposi's",
"sarcoma-associated",
"herpesvirus",
"molecular",
"biology",
"biochemistry",
"rna",
"biomolecular",
"isolation",
"data",
"visualization",
"cell",
"biology",
"nucleic",
"acids",
"oligonucleotides",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"organisms"
] |
2018
|
Two herpesviral noncoding PAN RNAs are functionally homologous but do not associate with common chromatin loci
|
Successive division events in the spherically shaped bacterium Staphylococcus aureus are oriented in three alternating perpendicular planes . The mechanisms that underlie this relatively unique pattern of division and coordinate it with chromosome segregation remain largely unknown . Thus far , the only known spatial regulator of division in this organism is the nucleoid occlusion protein Noc that inhibits assembly of the cytokinetic ring over the chromosome . However , Noc is not essential in S . aureus , indicating that additional regulators are likely to exist . To search for these factors , we screened for mutants that are synthetic lethal with Noc inactivation . Our characterization of these mutants led to the discovery that S . aureus Noc also controls the initiation of DNA replication . We show that cells lacking Noc over-initiate and mutations in the initiator gene dnaA suppress this defect . Importantly , these dnaA mutations also partially suppress the division problems associated with Δnoc . Reciprocally , we show that over-expression of DnaA enhances the over-initiation and cell division phenotypes of the Δnoc mutant . Thus , a single factor both blocks cell division over chromosomes and helps to ensure that new rounds of DNA replication are not initiated prematurely . This degree of economy in coordinating key cell biological processes has not been observed in rod-shaped bacteria and may reflect the challenges posed by the reduced cell volume and complicated division pattern of this spherical pathogen .
Binary fission in bacteria is carried out by a multiprotein complex called the divisome [1] . Assembly of this complex initiates with the formation of a ring-like structure at mid-cell called the Z-ring composed of the tubulin homolog FtsZ [2] . The Z-ring functions as a dynamic scaffold for the recruitment and assembly of the division machinery . In rod-shaped bacteria like E . coli and B . subtilis , midcell localization of the Z-ring is controlled by two negative regulatory systems: Min and nucleoid occlusion [1 , 3] . The Min system prevents Z-ring assembly at the poles while nucleoid occlusion blocks its assembly over of the nucleoid; together they ensure that division occurs at mid-cell . In E . coli , nucleoid occlusion is mediated by SlmA , a TetR family DNA binding protein [4] . SlmA directly inhibits FtsZ polymerization and this inhibition is stimulated by binding to specific DNA sequences [5–7] . In B . subtilis , nucleoid occlusion requires a ParB homolog called Noc [8] . The mechanism by which Noc inhibits Z-ring formation is less clear . Noc forms higher order nucleoprotein complexes [9] that associate with the cell membrane via an amino-terminal amphipathic helix and these complexes have been suggested to physically occlude the assembly of the division machinery [10] . Neither the Min system nor the nucleoid occlusion proteins are essential in E . coli or B . subtilis [4 , 8 , 11 , 12] . However , defects in both regulatory systems results in a lethal division block [4 , 8] . It was this synthetic lethal relationship that initially facilitated the identification of Noc and SlmA . Importantly , the specific binding sites for SlmA and Noc are distributed throughout the origin-proximal portions of their respective chromosomes and are largely absent in the region surrounding the terminus [6 , 9 , 13] . During the chromosome replication-segregation cycle in both organisms , the origin regions are segregated toward the cell poles and the terminus is brought to mid-cell , where it remains for the majority of the cell cycle [14 , 15] . As DNA replication nears completion , chromosome-bound nucleoid occlusion factors are depleted from mid-cell , relieving division inhibition at this site . Accordingly , SlmA and Noc are not only thought to function in division site selection , but also in coordinating cell division with chromosome replication and segregation [6 , 9] . Unlike rod-shaped bacteria , where the same perpendicular plane is used for binary fission , cell division alternates between three consecutive perpendicular planes in the spherical bacterium Staphylococcus aureus [16] . The mechanisms governing division site selection in this major human pathogen are poorly understood . S . aureus does not encode the Min system , but does have a homolog of B . subtilis Noc ( 48% identity , 65% similarity ) . A previous study showed that S . aureus cells lacking Noc are impaired in division and on average are larger than wild-type [17] . Furthermore , Δnoc mutants displayed an elevated frequency of cells with multiple Z-rings and division events bisecting the chromosome . Thus , a role for Noc in the spatial regulation of division is conserved in S . aureus . Even though S . aureus cells lack a Min system , Noc is not essential for viability [17] . Additional division regulators are thus likely to operate in this bacterium . By analogy with the Min- Noc- phenotype in rod-shaped organisms , we reasoned that defects in such regulators would result in a synthetically lethal or sick phenotype in combination with a noc deletion . To identify these factors we used transposon-sequencing ( Tn-Seq ) [18] to screen for genes that when inactivated result in a synthetically lethal with noc ( Sln ) phenotype . Two sln genes were identified in this screen , comEB and rhomboid ( rbd ) . To further characterize the nature of the synthetic lethal defect of Δnoc Δrbd and Δnoc ΔcomEB mutants , suppressors that allowed growth of the double mutants were selected . Surprisingly , in both cases , suppressor mutations that mapped to the dnaA gene encoding the initiator of DNA replication were isolated , suggesting a role for Noc in controlling origin firing . Indeed , S . aureus cells lacking Noc were found to over-initiate replication , and the dnaA suppressor mutations ameliorated the over-replication phenotype and significantly reduced the severity of the division defects displayed by Δnoc cells . Reciprocally , we found that over-expression of DnaA enhanced the over-initiation and cell division phenotypes of cells lacking Noc . Altogether , our data reveal that S . aureus Noc coordinates division with chromosome replication and segregation by controlling replication initiation as well as preventing divisome formation over the origin-proximal region of the chromosome . This degree of economy in coordinating key cell biological processes may reflect the challenges posed by the smaller size and complicated division pattern of this spherical bacterium .
To identify potential division site regulators , we screened for genes that become essential in the absence of S . aureus Noc ( SaNoc ) using Transposon-sequencing ( Tn-seq ) [18] . A wild-type S . aureus RN4220 strain and a Δnoc derivative were mutagenized with a Mariner-based transposon [19] . The resulting libraries were separately pooled and the transposon-genomic DNA junctions of all insertions in the libraries were mapped by massively parallel DNA sequencing . For each library , we detected insertions in ~30% of the potential Mariner insertion sites at TA dinucleotides , and each insertion site was detected by an average of ~50 reads . Transposon insertions in several genes were statistically underrepresented in the Δnoc library compared to the wild-type library as assessed by the Mann-Whitney U test ( Fig 1A and S1 Table ) . To validate the candidate synthetic lethal with Δnoc ( sln ) genes , we constructed a SaNoc depletion strain in which noc was expressed under the control of an anhydrotetracycline ( aTc ) -inducible promoter ( Ptet ) in the S . aureus HG003 background . Transposon-insertion mutants in the candidate genes were obtained from the Nebraska Transposon Mutant Library [20] and transduced into the Noc-depletion strain grown in the presence of inducer and then tested for their ability to form colonies in its absence . Transposon insertions in comEB and rbd were the only mutants that resulted in a strong dependence on noc expression for viability ( S1 Table ) . We therefore focused our analysis on these two factors . Insertion-deletion mutations in comEB and rbd were generated and the synthetic lethality upon depletion of Noc was confirmed ( Fig 1B ) . The rbd gene encodes a member of the Rhomboid family of membrane-embedded serine proteases [21 , 22] . ComEB is homologous to deoxycytidylate deaminase , an enzyme in one of the dTTP synthesis pathways [23] . Next , we investigated the cytological phenotypes of the Noc- ComEB- and Noc- Rbd- mutants . Cells lacking Rbd or ComEB but with functional SaNoc grew at rates similar to wild-type ( Fig 2A ) and appeared similar in size and morphology by fluorescence microscopy ( Fig 2B ) . By contrast and as expected , depletion of SaNoc in the mutant backgrounds resulted in severe growth defects ( Fig 2A ) . Examination of these SaNoc-depleted cells by fluorescence microscopy after 6 hours of growth in the absence of inducer revealed an increase in cell size , indicative of a defect in cell division . Specifically , the average volume of cells lacking Noc was ~69% greater than wild-type , while cells depleted of SaNoc in the absence of Rbd or ComEB were even larger ( ~86% and ~107% greater than wild-type ) ( Fig 2B ) . To investigate the defects in cell division at higher resolution , we visualized stained thin-sections by transmission electron microscopy . Cells lacking SaNoc or Rbd alone displayed a low frequency ( 4–6% ) of abnormal and incomplete septa that were not oriented orthogonal to each other ( Fig 2C and S1 Fig ) . Such aberrant divisions were even more prevalent ( ~27% ) in cells lacking both proteins , and the frequency of lysed cells also increased ( ~24% ) ( Fig 2C , S1 Fig and S2 Table ) . Collectively , these results indicate that the loss of ComEB or Rbd exacerbates the division defects of Δnoc cells , ultimately resulting in lysis . In an attempt to elucidate the potential roles of ComEB and Rbd in cell division we sought to identify suppressors of the observed synthetic lethal phenotypes . To do so , we first identified conditions that support growth of the ΔcomEB or Δrbd mutant upon depletion of SaNoc ( Fig 3A and S2 Fig ) ( see Materials and Methods ) . The permissive plating conditions were then used to generate Δnoc ΔcomEB and Δnoc Δrbd double deletion mutants . To isolate suppressors , the double mutants were grown in liquid medium under permissive conditions and then plated on solid agar under restrictive conditions . Nine suppressors of each double mutant pair were mapped using whole genome sequencing and confirmed by Sanger sequencing ( S3 Table ) . Two of the Δnoc Δrbd suppressors and one of the Δnoc ΔcomEB suppressors mapped to the dnaA gene encoding the replication initiator protein ( Fig 3B ) [24 , 25] . One of the suppressors of Δnoc Δrbd had a 56 bp deletion in the 5’ UTR of dnaA ( sup 1 ) , while another encoded a DnaA variant with a V141L substitution within the ATPase domain ( Domain III ) ( sup 2 ) . Similarly , a variant with an R254Q substitution in Domain III ( sup 3 ) suppressed the Δnoc ΔcomEB phenotype . Although we had hoped that the suppressors would shed light on a potential role for ComEB and Rbd in cell division regulation , the fact that mutations in dnaA suppressed both the Δnoc ΔcomEB and Δnoc Δrbd growth defects suggested that it was some aspect of the Noc- phenotype that was most likely being corrected by the suppressors . These findings raised the intriguing possibility that SaNoc function is linked to DNA replication . Accordingly , we decided to focus the remainder of this study on this potential connection . The possible roles of ComEB and Rbd in cell division and the cause of the synthetic lethal with Δnoc phenotype will be the subject of a separate report ( see Discussion ) . DnaA is the replication initiation protein in most bacteria [24 , 25] . It is an AAA+ ATPase that exists in both ATP- and ADP-bound forms . ATP-bound DnaA binds cooperatively to sequence elements ( called DnaA boxes ) in the replication origin , and catalyzes the unwinding of double stranded DNA . DNA unwinding leads to the ordered loading of helicase and the rest of the replication machinery followed by the initiation of bi-directional replication . To investigate whether cells lacking Noc have defects in the initiation of DNA replication we performed marker-frequency analysis using quantitative PCR ( qPCR ) and whole genome sequencing . This analysis revealed that the Δnoc mutant had a higher origin to terminus ( ori:ter ) ratio than wild-type cells and this increase could be suppressed by the dnaA suppressor mutations ( Fig 4A–4C ) , raising the possibility that the absence of SaNoc causes over-initiation of DNA replication . An increase in the ori:ter ratio can be caused by over-initiation of DNA replication or a reduction in replication elongation rates or fork stalling and/or collapse . The latter situation could be relevant in this case because cells lacking Noc have an increased frequency of chromosome guillotining [17] , which would lead to fork stalling . To investigate the nature of the defect in the Δnoc mutant , we quantitatively assessed DNA content relative to cell volume as a proxy for cell mass in single cells [26 , 27] . Fixed permeabilized cells were stained with propidium iodide and analyzed by phase contrast and fluorescence microscopy . Consistent with the idea that cells lacking SaNoc over-initiate , the Δnoc mutant had a higher DNA content relative to cell volume compared to wild-type and the dnaA ( sup1 ) mutation restored this ratio to wild-type levels ( S3 Fig ) . Further support for a role for Noc in controlling replication initiation comes from an experiment in which we attempted to correct the increase in the ori:ter ratio by expressing noc in trans from an aTc-inducible promoter . As anticipated , modest ( ~2 . 3-fold ) over-production of Noc in the Δnoc mutant ( Fig 4F ) led to a reduction in the ori:ter ratio ( Fig 4B ) . However , comparison of the genomic profile relative to wild-type ( Fig 4C and S3 Fig ) revealed that the ori:ter ratio was even lower than wild-type , consistent with the idea that Noc acts as a negative regulator of replication initiation . Finally , to address whether the increase in the ori:ter ratio in the Δnoc mutant resulted from fork stalling due to division on top of the chromosomes , we performed genome-wide marker frequency analysis on wild-type and Δnoc strains before and after inhibition of cell division using the FtsZ inhibitor PC190723 [28] . We carried out whole genome sequencing of wild-type and the Δnoc mutant prior to the addition of the drug and approximately 2 and 4 mass doublings after its addition . Despite the strong block to cell division , the increase in the ori:ter ratio in the Δnoc mutant persisted ( S4A Fig ) . Importantly , we found that DNA content increased after drug addition ( S4B Fig and S5 Fig ) , indicating that the division-inhibited cells continued DNA replication . Thus , chromosome guillotining cannot account for the increase in the ori:ter ratio in the Δnoc mutant . Altogether , these experiments are most consistent with a model in which SaNoc functions to limit replication initiation and in its absence cells over-initiate . Marker-frequency analysis in cells lacking Rbd or ComEB revealed that these mutants were not impaired in replication initiation ( Fig 4D ) . Furthermore , the ori:ter ratio in the Δnoc Δrbd double mutant was similar to the Δnoc single mutant ( Fig 4D ) , indicating that the Rbd- defect did not exacerbate the over replication phenotype of Noc- cells . Because SaNoc shares 48% identity with its B . subtilis counterpart ( BsNoc ) , we wondered whether B . subtilis cells lacking Noc also over-initiate . Marker-frequency analysis revealed that the B . subtilis Δnoc mutant had an ori:ter ratio similar to wild-type ( Fig 4E ) . As a control , we examined the ori:ter ratios of B . subtilis cells lacking the ParB homolog Spo0J and the ParA homolog Soj . Previous studies indicate that Soj ( ParA ) functions as both an activator and inhibitor of DnaA , while Spo0J ( ParB ) promotes the inhibitory activity of Soj [29 , 30] . As reported previously , both mutants had over-initiation phenotypes ( Fig 4E ) , but neither was as pronounced as in S . aureus Δnoc cells . In addition to SaNoc , S . aureus cells encode a second ParB-like protein ( SaParB ) that is 47% identical to Spo0J ( BsParB ) , although they lack a ParA counterpart . A mutant lacking SaParB had an ori:ter ratio similar to wild-type cells ( Fig 4E ) , indicating that among the ParB related proteins in S . aureus , a role in replication control is restricted to SaNoc . In B . subtilis , BsNoc binds multiple sites throughout the chromosome that are enriched around the origin-proximal two-thirds of the chromosome and largely absent from the replication terminus region [9] . This distribution of binding sites is thought to ensure that Z-ring formation is coordinated with chromosome segregation [9] . Because our data suggest that in S . aureus Noc regulates replication initiation , we wondered whether it might have a distribution of DNA binding sites distinct from BsNoc and different from that predicted based on mapping the sequence of the BsNoc binding sites on the S . aureus chromosome [31] . To investigate this we determined the S . aureus Noc binding sites ( NBS ) by chromatin-immunoprecipitation coupled to Illumina sequencing ( ChIP-seq ) . We used antibodies raised against BsNoc , which cross-reacted with the S . aureus homolog . A total of 41 chromosomal loci were identified that co-precipitated with SaNoc from wild-type cells compared to the Δnoc mutant ( Fig 5A and S7 Table ) . Importantly , all but one of these loci were located in the origin-proximal half of the chromosome ( between -700kb and +700kb ) , resulting in a distribution of binding sites that resembles the B . subtilis pattern [31] . Furthermore , as is the case in B . subtilis [9] , SaNoc binding spanned a region of 8–10 kb surrounding the peak of enrichment ( Fig 5B and 5C ) , suggesting that S . aureus Noc forms large nucleoprotein complexes at its binding sites . Finally , using motif discovery algorithms from the MEME suite [32] , the consensus Noc binding site was determined , the core of which was found to be identical to the B . subtilis NBS consensus sequence [31] ( Fig 5D ) . The similar distribution of Noc binding sites , the identical consensus binding sequence , and the shared ability of the Noc homologs to spread from these sites are all consistent with SaNoc functioning as a nucleoid occlusion factor in addition to its role in replication control identified here . Since the Noc proteins from B . subtilis and S . aureus are 48% identical and they have the same consensus binding site , we investigated whether they could substitute for each other in vivo . B . subtilis cells lacking the Min system and Noc are viable at 30°C but fail to grow at 42°C [8] . Accordingly , we generated his6 fusions to both B . subtilis and S . aureus noc under the control of an IPTG-inducible promoter ( Pspank ) and introduced these into a ΔminD Δnoc double mutant under permissive conditions . Surprisingly , only B . subtilis noc was able to complement the growth defect of the double mutant at 42°C ( Fig 6A ) . Similarly , we took advantage of the synthetic lethal phenotype of Δnoc Δrbd in S . aureus . We introduced the same his6 fusions under the control of an aTc-inducible ( Ptet ) promoter into the double mutant at the L54a phage attachment site . Again , only the native noc allele supported growth under the non-permissive condition ( Fig 6B ) . Furthermore , not only was B . subtilis noc incapable of supporting growth of the S . aureus Δnoc Δrbd mutant , it was unable to correct the over-initiation defect of the S . aureus Δnoc mutant ( Fig 6C ) . Importantly , immunoblot analysis using anti-6xHis antibodies indicated that in both complementation tests , the heterologous Noc proteins were produced at comparable or even higher levels than the native proteins ( Fig 6D ) . Furthermore , as has been observed previously with BsNoc [9] , a YFP fusion to SaNoc expressed in B . subtilis localized in foci along the cell periphery indicative of it associating with membranes surrounding the nucleoid in addition to binding DNA ( Fig 6E and S6 Fig ) . Finally , ChIP-seq experiments using anti-6xHis antibodies revealed that both His6-Noc fusions bound to virtually all the same chromosomal sites in B . subtilis and S . aureus ( Fig 6F , S7 Fig , S8 Fig and S7 Table ) . Thus , merely binding to the NBS sites throughout the chromosome is not sufficient for Noc function , suggesting that the Noc orthologs require species-specific contacts with replication and/or division factors and that these interaction interfaces are not conserved between the two proteins ( See Discussion ) . The SaNoc enrichment profile from the ChIP-seq experiment contained a small peak of enrichment within the promoter region of the dnaA gene ( Figs 5C and 6F ) . Although BsNoc expressed in S . aureus contained a similar enrichment peak , we wondered whether SaNoc functions in replication control by regulating dnaA expression . Accordingly , we compared the relative levels of DnaA and the house-keeping sigma factor SigA by immunoblot analysis using anti-DnaA and anti-SigA antibodies . Both antibodies were raised against the B . subtilis proteins but cross-reacted with their S . aureus orthologs . Intriguingly , we detected a 1 . 7-fold increase in DnaA levels in cells lacking SaNoc compared to wild-type ( Fig 7A ) . Furthermore , the dnaA suppressor mutations reduced the levels of DnaA in the Δnoc mutant to a level slightly below wild-type . To directly test whether the over-initiation phenotype in cells lacking SaNoc was due to this increase in DnaA levels , we compared the ori:ter ratio in a strain harboring an aTc-inducible promoter fusion to dnaA . Consistent with previous results , we did not observe a correlation between mild DnaA overproduction and over-initiation of DNA replication in S . aureus [33] . In the absence of inducer , the leaky expression from this plasmid-borne dnaA allele resulted in DnaA levels comparable to the Δnoc mutant ( Fig 7A ) . However and importantly , the ori:ter ratio in this strain was similar to wild-type and a matched empty vector control strain . Furthermore , even higher levels of DnaA in the presence of inducer did not cause over-initiation compared to the control strain ( Fig 7A ) . As a positive control for these experiments we used a strain harboring the same aTc-inducible promoter fused to an ATPase defective allele of dnaA ( dnaAR318H ) . As reported previously [33] , induction of DnaAR318H caused a severe growth defect and led to dramatic over-initiation of replication ( Fig 7A ) . We conclude that Noc directly or indirectly affects DnaA levels but these changes are not the major determinant of the observed effects on DNA replication . If , as our data suggest , Noc negatively regulates replication initiation at the post-translational level , then overproduction of DnaA in a Δnoc mutant should exacerbate the over-initiation phenotype . As can been seen in Fig 7B , overexpression of dnaA in cells lacking Noc dramatically increased replication initiation with an ori:ter ratio approaching the levels reached upon DnaAR318H production in wild-type cells . Furthermore , similar to what has been reported previously for the ATPase deficient DnaA variant [33] , this level of over-replication led to a dramatic growth defect ( Fig 7C ) and a significant increase in cell size ( S9 Fig ) ( see below ) . Altogether these data argue that Noc exerts its control on replication initiation through a post-translational mechanism . Previous studies indicate that S . aureus cells lacking SaNoc display division defects including oblique FtsZ-rings , an increase in cell size , and the production of anucleate cells [17] ( Fig 2B and 2C ) . Here , we have shown that the Δnoc mutant also over-initiates DNA replication . To investigate the contribution of over-initiation to the division defects in cells lacking SaNoc , we took advantage of the suppressor mutations isolated in dnaA . We analyzed cell size and FtsZ localization by fluorescence microscopy in the Δnoc mutant and in the Δnoc mutant harboring dnaA ( sup1 ) or dnaA ( sup2 ) . Consistent with our observation that over-initiation leads to an increase in cell size ( S9 Fig ) , both dnaA mutations partially suppressed the large cell phenotype of the Δnoc mutant ( Fig 8C and S10 Fig ) . To monitor FtsZ localization we used an FtsZ-GFP fusion as a merodiploid [34] in the same set of strains . Interestingly , the suppressors reduced the percentage of cells with abnormal Z-rings ( Fig 8A and 8C ) . In particular , we observed a reduction in oblique Z-rings and in the percentage of cells with diffuse FtsZ-GFP fluorescence . Furthermore , we also found that ~10% of cells lacking SaNoc exhibited extremely strong DAPI staining ( S10 Fig ) . We suspected this phenotype was due to increased cell envelope permeability in the mutant , and staining cells with the membrane-impermeable dye propidium iodide ( PI ) confirmed this ( Fig 8B ) . Importantly , the dnaA suppressors alleviated this hyper-staining/membrane permeability phenotype such that it was comparable to wild-type cells ( Fig 8B and 8C ) . Collectively , these data suggest that the cell division problems experienced by S . aureus cells lacking SaNoc are partially due to the over-initiation of DNA replication .
The elevated ori:ter ratio observed in Δnoc cells and its suppression by mutations in dnaA are consistent with a model in which SaNoc controls replication initiation . Further support for this idea comes from our findings that cells lacking Noc have elevated DNA content relative to cell volume and cells with increased Noc protein levels have a reduction in the ori:ter ratio relative to wild-type . Finally , the synthetic growth phenotype and increased ori:ter ratio in Δnoc mutant cells that over-expresses DnaA also argue for a role in replication initiation . However , because of the limited tools available to study replication in S . aureus , we cannot rule out the possibility that Noc also functions in replication elongation [35] . How SaNoc controls replication initiation remains unclear . Our finding that overexpression of dnaA does not lead to increased replication initiation unless Noc is inactivated indicates that SaNoc is unlikely to be regulating origin firing by affecting dnaA expression . Furthermore , the failure of BsNoc to functionally substitute for SaNoc even though it binds to almost all the same chromosomal sites as SaNoc suggests that site-specific DNA-binding activity alone is unlikely to account for replication control . However , we note that our ChIP-seq analysis identified two small enrichment peaks for SaNoc surrounding the dnaA gene while expression of BsNoc in S . aureus only had one of these small peaks ( Fig 6F ) . Intriguingly , the predicted DnaA binding sites required for origin unwinding are located in these two regions [33 , 36] . Accordingly , it is formally possible that SaNoc binding to these sites occludes DnaA and inhibits origin firing . Alternatively , this second enrichment peak could be reporting on an interaction between Noc and DnaA bound to the origin ( see below ) . Because this intergenic region is critical for origin firing [37] , we were unable to investigate whether or not Noc binding at this position is important for its role in replication control . We hypothesize that SaNoc recognizes a specific molecular target to mediate its effect on replication initiation and that the interaction interface with this target is not conserved in BsNoc . DnaA is an obvious candidate as a direct regulatory target of SaNoc , but so far we have not been able to detect a SaNoc-DnaA interaction . Further work will therefore be required to determine the precise mechanism by which SaNoc controls initiation . Nevertheless , the discovery that SaNoc controls replication initiation in addition to its role in nucleoid occlusion provides a striking example of a single protein that can coordinate cell division and DNA replication by affecting both major cell-cycle processes . It is noteworthy that mutations in dnaC encoding the replicative helicase were also identified as suppressors of Δnoc Δrbd and Δnoc ΔcomEB ( S3 Table ) . Furthermore , at least one of these mutations ( dnaCA352V ) was able to suppress the over-initiation phenotype of the Δnoc Δrbd mutant ( S11 Fig ) . Since DnaC is required for both replication initiation and elongation [38] the molecular basis of suppression remains unclear . In the context of the model that Noc negatively regulates initiation , we hypothesize that these DnaC variants are specifically impaired in initiation function , however , further analysis will be required to establish the defects in these mutants . The mechanism by which Noc proteins block cell division over the nucleoid has remained unclear for many years . Attempts at identifying a specific component of the divisome targeted by BsNoc have been unsuccessful [10] . However , this protein was recently shown to have an amino-terminal amphipathic helix that recruits DNA-bound Noc complexes to the membrane [10] . Based on this result , it has been proposed that , rather than interacting with a specific component of the division apparatus , BsNoc nucleoprotein complexes at the membrane physically occlude the assembly of the division machinery in areas occupied by the origin-proximal portion of the chromosome [10] . If true , the only activities of Noc required for its control of cell division are its association with Noc-binding sites on the chromosome and their recruitment to the membrane . Given that SaNoc binds to the same sites as BsNoc , forms higher order nucleoprotein complexes and , like BsNoc , appears to associate with the membrane in addition to DNA , the physical occlusion model would predict that SaNoc should be able to substitute for BsNoc . However , we find that SaNoc fails to support growth of Min- Noc- B . subtilis cells under non-permissive conditions , suggesting that physical occlusion of divisome assembly is unlikely to be the sole mechanism by which Noc proteins mediate division inhibition . Our results suggest that an additional species-specific interaction between Noc and a divisome protein is also likely to be required for its activity . The search for this elusive interaction partner should therefore continue . To ensure faithful inheritance of genetic material , cell division must be coordinated with DNA replication and segregation . Noc is thought to contribute to this coordination by blocking cell division over the origin-proximal region of the chromosome where its binding sites are distributed [9] . This division inhibition activity is the only known function of B . subtilis Noc . However , as we have shown here , S . aureus Noc has the additional activity of restricting the initiation of DNA replication , thus further integrating the processes of division and DNA replication/segregation in this organism . To our knowledge , this activity for SaNoc is the first example of a ParB-like protein controlling replication initiation . ParB proteins typically bind centromeric parS sequences and work with ParA partner proteins to promote origin segregation [39] . In B . subtilis , the ParA protein Soj has also been shown to both inhibit and activate DNA replication initiation by DnaA in a manner modulated by the ParB protein Spo0J [29] . S . aureus lacks a ParA homolog , indicating that the effects on replication observed in mutants defective for SaNoc are not an indirect consequence of a misregulated ParA protein partner as is the case for B . subtilis mutants inactivated for Spo0J ( ParB ) . Although we favor the idea that SaNoc directly modulates DnaA activity in S . aureus additional work is required to establish the mechanism of replication control . Irrespective of the precise mechanism , our results raise the question of why S . aureus utilizes a Noc protein with the added activity of controlling the initiation of DNA replication whereas B . subtilis does not . Differences in cell morphology and the geometries of chromosome segregation and cell division in these organisms might explain the distinct Noc regulatory activities used by these organisms . In rod-shaped cells , chromosome segregation proceeds along the long cell axis . The spatial cues do not change from one cell cycle to the next . Therefore , even if replication were to outpace division , the pattern of chromosome segregation would remain the same . The situation in spherical cells like S . aureus is very different . Just as division alternates between three perpendicular planes , so too must the orientation of chromosome segregation . Thus , if replication were initiated prematurely relative to the progress of division , the appropriate spatial cues might not be in place to direct segregation in an orientation compatible with that of the next division plane . Furthermore , Noc-dependent inhibition of division due to the segregation of these prematurely replicated origins would be more difficult to overcome given the reduced spatial resolution in these small spherical cells . Consistent with this idea , we note that a significantly larger portion of the terminus region in S . aureus lacks Noc binding sites compared to B . subtilis . The dual function of SaNoc in nucleoid occlusion and replication control may therefore serve to maintain a tighter coupling between division and replication to keep the geometries of chromosome segregation in step with that of the division plane and/or to ensure that division is not impaired by prematurely segregated origins . The challenge for the future will be to determine how SaNoc restrains replication and how its apparent inhibitory effects are relieved in response to cell cycle signals to facilitate proper coordination between replication initiation and cell division .
S . aureus strains were grown in tryptic soy broth ( TSB ) at 30°C or 37°C with aeration , unless otherwise indicated . The medium was supplemented with erythromycin ( 5 μg/ml for chromosomal insertions and 10 μg/ml for plasmids ) , chloramphenicol ( 5 μg/ml for chromosomal insertions , 10 μg/ml for plasmid ) , spectinomycin ( 100 μg/ml ) , kanamycin and neomycin ( 25 μg/ml each ) , 5-bromo-4-chloro-3-indolyl ß-D-galactopyranoside ( X-Gal , 250 μg/ml ) , or anhydrotetracycline ( aTc , 100 ng/ml unless otherwise indicated ) . B . subtilis strains were derived from the prototrophic strain PY79 [40] . Cells were grown in Luria Broth ( LB ) or Casein Hydrolysate ( CH ) medium [41] at 37°C with aeration , unless otherwise indicated . The medium was supplemented with tetracycline ( 10 μg/ml ) , spectinomycin ( 100 μg/ml ) , kanamycin 10 μg/ml ) , chloramphenicol ( 5 μg/ml ) , xylose ( 0 . 5% w/v ) or isopropyl ß-d-thiogalactopyranoside ( IPTG , 0 . 5 mM ) . Lists of oligonucleotide primers , strains , plasmids , and descriptions of their construction can be found in Supplemental Material . Transposon libraries of S . aureus strains TM18 ( wt ) and Δnoc::spec were constructed according to Wang et al . [19] , except that the donor strain TM17 contained the transposon plasmid pTP77 . Each library had >70 , 000 insertions . Transposon-sequencing ( Tn-seq ) was adapted from Zhang et al . [42] . Briefly , genomic DNA isolated from each library was sonicated ( using Q800R2 QSONICA ) to 200–400 bp fragments . Nicked ends were repaired ( Quick Blunt kit , NEB ) , and dA-tails were added with Taq polymerase ( NEB ) , followed by ligation of adapters using T4 ligase ( NEB ) . Transposon-junctions were PCR-amplified using primers oTP182 and oTP183 . PCR products were purified and incubated with the endonuclease NotI ( NEB ) to remove any contaminating DNA derived from the transposon plasmid . Finally , a second round of PCR was performed on 100 ng of each sample , using primers oTP184-187 and oTP188-191 . A 200–400 bp product was gel-purified and sequenced on the Illumina HiSeq 2500 platform ( Biopolymers Facility at Harvard Medical School ) . The sequencing reads were mapped to each TA site on the S . aureus NCTC8325 genome ( NCBI NC_007795 . 1 ) . Genes in which reads were statistically underrepresented in Δnoc compared to wild-type were identified by the Mann Whitney U test . Visual inspection of transposon insertion profiles was performed with the Sanger Artemis Genome Browser and Annotation tool . Tn-seq datasets are accessible through NCBI's Gene Expression Omnibus under accession number GSE101335 . Noc depletion strains harboring Δrbd or ΔcomEB ( strains aTP431 and aTP508 ) were streaked on plates containing 0 . 5 X LB No NaCl , LB No NaCl , LB 0 . 5% NaCl , or LB 1% NaCl , and incubated at 30˚ , 37˚ or 42˚C . The colony size of each strain was compared with the Noc depletion strain aTP359 grown under the same conditions . The growth conditions in which colonies from aTP431 and aTP508 were comparable in size to aTP359 were considered permissive . LB 0 . 5% NaCl at 37°C did not allow growth for either aTP431 or aTP508 and was considered non-permissive . ChIP was performed as described previously [43] with modification . Briefly , cells were grown in TSB medium to OD600 = 0 . 4 for S . aureus , and in CH medium to OD600 = 0 . 25 for B . subtilis , and incubated with 3% formaldehyde for 30 min at room temperature . Formaldehyde was quenched and the cells washed , and lysed . Lysates were sonicated using a Q800R2 sonicator ( QSONICA ) to generate 300–500 bp randomly sheared chromosomal DNA fragments . The lysates were then incubated with anti-BsNoc or anti-6xHis antibodies ( Genscript ) overnight at 4°C . The sample was then processed as described previously and sequenced using the Illumina MiSeq platform . The sequencing reads were mapped to the S . aureus NCTC8325 genome ( NCBI NC_007795 . 1 ) or B . subtilis PY79 genome ( NCBI NC_022898 . 1 ) using CLC Genomics Workbench software ( Qiagen ) . ChIP-seq datasets are accessible through Gene Expression Omnibus under accession number GSE93264 . Whole genome sequencing was performed according to Baym et al . [44] . 1 . 5 ng genomic DNA from each strain was used in a 2 . 5 μl Nextera Tagmentation reaction ( Illumina ) . After incubation at 37°C for 5 minutes , the reaction was immediately placed on ice . Tagmented DNA was then mixed with 11 μl KAPA Library Amplification Kit master mix ( Kapa Biosystems ) , 4 . 4 μl of each indexing primer ( 5 μΜ ) followed by PCR amplification . Amplified DNA was then purified from 15 μl of each reaction with 12 μl of Agencourt AMPure XP beads ( Beckman Coulter ) and resuspended in 30 μl buffer containing 10 mM Tris-Cl ( pH 8 . 0 ) , 1 mM EDTA ( pH 8 . 0 ) , and 0 . 05% Tween-20 . The concentration of each sample was measured by Qubit ( Thermo Fisher Scientific Inc . ) and equivalent amounts of DNA were pooled , analyzed on TapeStation ( Agilent Technologies ) for quality control , and sequenced using the Miseq platform ( Illumina ) . Sequencing reads were analyzed on CLC Genomics Workbench . Single nucleotide polymorphisms ( SNP ) and deletions were identified by comparing the sequence of the suppressors to the parental strains . Raw data are accessible through NCBI's Sequence Read Archive under accession number SRP111468 . Cells at OD600 ~ 0 . 4 were pelleted by centrifugation at 3 , 300 xg for 2 min , and immobilized on 2% ( wt/vol ) agarose pads containing M9 salts solution . Cell walls were stained with 0 . 5 μg/ml BODIPY FL vancomycin ( Van-FL , ThermoFisher ) and 0 . 5 μg/ml Vancomycin-HCl . Cell membranes were stained with 1 μg/ml Nile Red ( ThermoFisher ) . DNA was stained with DAPI ( Molecular Probes ) at 2 μg/ml , or with Propidium Iodide ( Molecular Probes ) at 5 μM . To quantitatively characterize DNA content , cells were immediately fixed with ethanol followed by staining with Propidium iodide . Fluorescence microscopy was performed using a Nikon TE2000 inverted microscope with a Nikon CFI Plan Apo VC 100X objective lens . Images were cropped and adjusted using MetaMorph software ( Molecular Devices ) . Final figures were prepared using Adobe Illustrator ( Adobe Systems ) . Image analyses were performed using MATLAB . All cells were segmented using phase-contrast images . Well-separated cells that were not undergoing cytokinesis were manually selected and used for analysis . Cell volume and intensity of fluorescence signals were determined using the built-in functions in MATLAB . Cells were harvested at OD600 ~ 0 . 4 , resuspended into LB medium , and fixed overnight at 4°C with 2 . 5% Glutaraldehyde , 1 . 25% Paraformaldehyde and 0 . 03% picric acid in 0 . 1 M sodium cacodylate buffer ( pH 7 . 4 ) . Fixed cells were washed in 0 . 1M cacodylate buffer and incubated with 1% Osmium Tetroxide ( OsO4 ) /1 . 5% Potassium Ferrocyanide ( KFeCN6 ) for 1 hour . Samples were then washed twice with ddH2O , washed once with Maleate buffer ( MB ) , and incubated in 1% uranyl acetate in MB for 1 hour . Samples were washed twice with ddH2O , followed by dehydration in increasing concentration of alcohol ( 10 min each with 50% , 70% , 90% , twice with 100% ) . The samples were then incubated in propyleneoxide for 1 hour and infiltrated overnight in a 1:1 mixture of propyleneoxide and TAAB Epon ( Marivac Canada Inc . St . Laurent , Canada ) . The samples were then embedded in TAAB Epon and polymerized at 60°C for 48 hrs . Ultrathin sections ( ~60 nm ) were cut on a Reichert Ultracut-S microtome , mounted onto copper grids stained with lead citrate , and examined in a JEOL 1200EX Transmission electron microscope . Images were recorded with an AMT 2k CCD camera . S . aureus cells were grown in LB 0 . 5% NaCl medium at 37°C , except Δnoc Δrbd , which was first grown under permissive conditions ( 0 . 5X LB no NaCl , 30°C ) to OD600 ~0 . 4 , then back diluted into nonpermissive conditions ( LB 0 . 5% NaCl medium , 37°C ) at an OD600 ~0 . 04 . Sodium azide was added to cells at OD600 ~0 . 4 to prevent further growth and chromosomal DNA was then isolated using DNeasy Blood and Tissue Kit ( Qiagen ) . To obtain cells with an ori:ter ratio of 1 , exponentially growing cells ( OD600 = 0 . 15 ) were treated with 50 μg/ml rifampicin for 1 hour at 37°C before harvesting . Quantitative PCR was performed using the SYBR FAST qPCR Kits ( Kapa Biosystems ) on a StepOne Plus Real-Time PCR System ( Applied Biosystems ) . Primer sets oTP478/oTP479 and oTP480/oTP481 were used to amplify the S . aureus oriC and ter regions , respectively . The 2-ΔΔCt method was used to quantify the relative ori:ter ratio , where chromosomal DNA sample from rifampicin-treated wild-type cells was used for normalization . Marker-frequency analysis of B . subtilis strains was the same as S . aureus , except that cells were grown in CH medium at 30°C until OD600 ~ 0 . 25 . Primer sets oTP482/oTP483 and oTP484/oTP485 were used to amplify the B . subtilis oriC and ter regions , respectively . Cell growth and chromosomal DNA preparation were the same for genome-wide marker frequency analysis . Whole genome sequencing was performed as described above . The sequencing reads from all strains were normalized to 51 million and the data were plotted in 30 kb bins . The smoothed conditional means were determined by LOESS method using ggplot2 in R [45] . The fraction of points used to fit each local regression ( spanS ) was 0 . 08 . WGS data for marker frequency analysis are accessible through Gene Expression Omnibus under accession number GSE100776 . Whole cell lysates from exponentially growing cultures ( OD600 = 0 . 4 ) were prepared as previously described [46] . Samples were heated at 100°C for 5 min prior to loading . Equivalent amount of cells based on OD600 at the time of harvest were separated by SDS-PAGE on Tris-Glycine gels , and transferred to PVDF membrane ( EMD Millipore ) . Membranes were blocked in 5% nonfat milk with 0 . 5% Tween-20 for 1 h . Blocked membranes were probed with anti-6xHis ( 1:4000 ) ( Genscript ) , anti-Noc ( B . subtilis ) ( 1:10 , 000 ) , anti-DnaA ( B . subtilis ) ( 1:10 , 000 ) [47] , or anti-σA ( B . subtilis ) ( 1:10 , 000 ) [48] antibodies . Primary antibodies were detected with horseradish-peroxidase conjugated anti-mouse ( for anti-His ) or anti-rabbit ( for anti-Noc , anti-DnaA and anti-σA ) antibodies ( BioRad ) and detected with Western Lightning ECL reagent as described by the manufacturer . Images were obtained by using the Molecular Imager Gel Doc XR system ( Bio-Rad ) and analyzed by ImageJ [49] . S . aureus immunoblots were performed in a strain lacking the spa gene encoding Surface Protein A .
|
The mechanisms by which bacteria coordinate cell division with chromosome replication and segregation remain poorly understood . This coordination is particularly challenging in the spherical bacterium Staphylococcus aureus , which unlike rod-shaped bacteria , divides in three consecutive perpendicular division planes . The only known spatial regulator of division in S . aurues is the nucleoid occlusion protein Noc . In Bacillus subtilis , Noc has been shown to bind specific DNA sequences on the chromosome and block the assembly of the cell division apparatus over these sites . Because these binding sites are enriched in the origin-proximal portion of the chromosome and absent in the terminus region , Noc is thought to help coordinate cell division with chromosome segregation . Here , we report that S . aureus Noc protein not only plays a similar role in nucleoid occlusion , but also controls the initiation of DNA replication , thus providing an even tighter connection between cell division and chromosome biology than previously appreciated . This degree of economy in coordinating key cell biological processes may reflect the challenges posed by the small cell size and complicated division pattern of this spherical pathogen .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"bacillus",
"microbiology",
"light",
"microscopy",
"staphylococcus",
"aureus",
"prokaryotic",
"models",
"dna",
"replication",
"microscopy",
"experimental",
"organism",
"systems",
"dna",
"bacteria",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"chromosome",
"biology",
"staphylococcus",
"medical",
"microbiology",
"fluorescence",
"microscopy",
"microbial",
"pathogens",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"phenotypes",
"bacillus",
"subtilis",
"genetics",
"electron",
"microscopy",
"biology",
"and",
"life",
"sciences",
"phase",
"contrast",
"microscopy",
"organisms",
"chromosomes"
] |
2017
|
The nucleoid occlusion factor Noc controls DNA replication initiation in Staphylococcus aureus
|
Pleiotropy refers to the phenomenon in which a single gene controls several distinct , and seemingly unrelated , phenotypic effects . We use C . elegans early embryogenesis as a model to conduct systematic studies of pleiotropy . We analyze high-throughput RNA interference ( RNAi ) data from C . elegans and identify “phenotypic signatures” , which are sets of cellular defects indicative of certain biological functions . By matching phenotypic profiles to our identified signatures , we assign genes with complex phenotypic profiles to multiple functional classes . Overall , we observe that pleiotropy occurs extensively among genes involved in early embryogenesis , and a small proportion of these genes are highly pleiotropic . We hypothesize that genes involved in early embryogenesis are organized into partially overlapping functional modules , and that pleiotropic genes represent “connectors” between these modules . In support of this hypothesis , we find that highly pleiotropic genes tend to reside in central positions in protein-protein interaction networks , suggesting that pleiotropic genes act as connecting points between different protein complexes or pathways .
The phenomenon of pleiotropy highlights the fact that some genes in the genome perform multiple biological functions . Although individual examples of pleiotropic genes have been discovered [1]–[4] , pleiotropy remains a poorly understood genetic phenomenon and there have been very few systematic studies . In S . cerevisiae , the collection of mutant strains for nearly all genes has enabled high-throughput tests of growth fitness under a variety of environmental conditions [5] , [6] . The degree of pleiotropy has been estimated based on the number of conditions under which mutant strains showed abnormal fitness [6] . In multi-cellular organisms , the availability of high-throughput RNAi techniques may lead to the opportunity for systematic analysis of pleiotropic genes . However , when multiple phenotypic effects are present , it is not obvious whether the phenotypic effects should be attributed to the loss of a single function or to multiple functions . For example , a phenotypic effect at earlier stages of animal development may accumulate during cell divisions and migrations , resulting in many defects at later stages of development . In this case , although many defects are observed , they can all be accounted for by the loss of a uniform gene function . Therefore , it is not clear how pleiotropic genes should be identified in practice and what mechanisms lie behind pleiotropy . C . elegans is especially amenable to genome-wide loss-of-function analyses because of well-characterized anatomy , short life cycle , and the convenience of RNAi techniques . The C . elegans early embryo is a model system for studying mitotic cell divisions . Piano et al screened a set of ovary-enriched genes by RNAi and systematically described early embryonic defects for 161 genes in terms of RNAi-associated phenotypes [7] . Using the RNAi data , they grouped these genes into “phenoclusters” , which correlated well with functional annotations of these genes . Sonnichsen et al . performed whole-genome RNAi experiments to search for genes involved in early embryogenesis [8] . They defined a series of cellular defects occurring in the first two cell divisions , and identified 661 genes that showed at least one of these defects . These genes were manually grouped into functional classes . For example , genes involved in cell polarity were grouped together since the RNAi of these genes resulted in symmetric cell divisions; genes involved in DNA damage checkpoints were grouped together since the RNAi of these genes resulted in delayed P1 cell division . Multiple defects during early cell divisions can be scored when a single gene is perturbed . All the scored defects happen in the first approximately 50 minutes of embryonic development , up to a four-cell stage embryo . This short time window ensures that most observed defects are direct rather than secondary . These data and information provide an excellent biological context to systematically explore the phenomenon of pleiotropy . In this paper , we address several open questions regarding pleiotropy using C . elegans early embryogenesis as the model system . First , how can complex phenotypes be decomposed and be linked to the loss of specific biological functions ? Second , how can we systematically identify pleiotropic genes ? Third , does pleiotropy exist commonly in a biological system ? Finally , what potential mechanisms underlie pleiotropy ? We find that sets of cellular defects ( or “signatures” ) are well correlated with losses of certain biological functions , and these signatures can be used to decompose complex phenotypic profiles so as to provide functional annotations . Approximately half of the genes involved in early embryogenesis are found to be pleiotropic , suggesting the prevalence of pleiotropy in biological systems . By integrating phenotypic profiles with protein-protein interaction networks , we observe that highly pleiotropic genes tend to show a higher network “betweenness” [9] than other genes involved in early embryogenesis , suggesting that pleiotropic genes play an important role in connecting various biological pathways .
Systematic RNAi screens have identified genes involved in early embryogenesis and have characterized their phenotypic profiles , which are composed of a series of cellular defects [8] . As has been described previously [8] , phenotypic data can be visualized in a matrix where rows index genes and columns index defects . A gene is given a score of either zero ( absence ) or a positive value ( presence ) for each of the 45 defects [8] . We plotted the distribution of the percentage of genes involved in early embryogenesis against the number of defects for which the genes have positive scores ( Figure 1 ) . By randomly permuting the values among genes while keeping each column sum fixed ( i . e . , fixing the total number of genes each defect is associated with ) , we generated random control datasets and observed that significantly more genes in the real data set exhibit a large number of loss-of-function defects than those in random control sets . In the real dataset , 57 out of 661 genes show 15 or more defects , whereas on average only 1 gene is expected to show this number of defects in a randomly permuted dataset ( P-value<0 . 001 , see Methods ) . Genes exhibiting a large number of defects in their phenotypic profiles may be candidates for pleiotropic genes . However , should the degree of pleiotropy be solely determined by the number of defects ? It is possible that occurrences of some cellular defects are highly correlated with one another . The highly correlated defects are likely caused by the perturbation of a single-function gene rather than a pleiotropic gene . In order to investigate how strongly cellular defects correlate with each other , we analyzed the occurrence of each individual defect and the co-occurrence of each pair of defects . We then computed the ratio of the observed co-occurrence of each defect pair to the expected co-occurrence as if the two defects occurred independently ( see Methods ) . We plotted the ratios as a correlation map ( Figure 2 ) and found that some defects co-occur much more frequently than expected , while some never co-occur in the same phenotypic profile , suggesting that not all defects occur independently from each other . For example , P1/AB nuclear separation—cross-eyed ( Defect 23 ) and four-cell stage nuclei—size/shape ( Defect 34 ) co-occur at very high frequency , suggesting that embryos showing defects in nuclear separation at the two-cell stage are very likely to be abnormal in nuclear size and shape at the four-cell stage . P1/AB nuclear separation—cross-eyed also co-occurs with P0 cytokinesis—furrow specification ( Defect 20 ) and several other defects , and four-cell stage nuclei—size/shape also co-occurs with P0 spindle rocking ( Defect 17 ) and several other defects . We also analyzed the occurrence of cellular defects by both linear principal component analysis ( PCA ) and logistic principal component analysis ( LPCA ) [10] . Although LPCA appears to be more appropriate for 0-1 type of data , PCA is more appealing in terms of its interpretability because the dimensions of LPCA are not orthogonal and the eigenvalues of LPCA cannot be used to rank the importance of principle components . As dimensional reduction tools , both PCA and LPCA gave similar results for this dataset–the projection of the defects onto the plane spanned by the first and second principal components ( PCs ) reveals very similar pattern ( Figure 3 , for LPCA ) . For example , P0 cytokinesis—furrow specification ( Defect 20 ) , P1/AB nuclear separation—cross-eyed ( Defect 23 ) , four-cell stage nuclei—size/shape ( Defect 34 ) , and P0 spindle rocking ( Defect 17 ) show high co-occurrence in the correlation map , and they are positioned close to one another in the LPCA plot as well . The observation of closely related defects suggests that the degree of pleiotropy cannot be readily measured by simply counting the number of defects . In order to study pleiotropy , we need to identify combinations of defects , or “phenotypic signatures , ” which describe the effects of losing individual biological functions . Cell divisions in early embryogenesis involve a number of biological functions such as chromosome segregation , cytokinesis , and cell polarity . Sonnichsen et al . manually grouped genes identified in the RNAi screen into 23 mutually exclusive classes according to their phenotypic profiles [8] . Among these , 22 classes have functional annotations and the remaining one is composed of genes whose phenotypic profiles contain a large number of defects and do not resemble profiles of any functionally characterized genes . We designed a computational approach to determine phenotypic signatures for each of the 22 functional classes and to identify additional genes potentially belonging to the given class ( Figure 4 ) . The phenotypic signature of a class is defined as a collection of cellular defects significantly enriched in that class as compared to the whole dataset . More specifically , for each class as defined in [8] , we computed the P-value for the enrichment of each defect according to the hypergeometric distribution . This class' phenotypic signature is then composed of all defects whose enrichment P-values are no greater than 0 . 05 after correcting for multiple comparisons . As a result , we found phenotypic signatures for 18 of the 22 functional classes . For the remaining 4 classes , no significantly enriched defects could be identified , because these classes all contained too few genes ( 5 or fewer ) for any defect to pass our statistical threshold . The above procedure can be illustrated for the cell polarity class ( Figure 5 ) . Originally , a total of 12 genes , including some genes previously known to be involved in cell polarity , were assigned to this class . We identified 7 defects significantly enriched in this class as its phenotypic signature . Among those defects , P1/AB asynchrony of division and four-cell stage configuration are the characteristic defects of asymmetric cell divisions . Defects in P0 pronuclear meeting , P0 spindle positioning , P0 spindle poles , P1 nuclear migration/rotation , and AB spindle orientation are the ones that are likely to accompany the loss of asymmetry . We searched the rest of the dataset for additional genes with phenotypic profiles matching the signature ( see Methods ) and identified RGA-3 , a putative Rho GTPase activating protein . This gene was originally classified as involved in cortical structure . Our search for phenotypic signatures did not rule out its functional involvement in cortical structure , but suggested its additional roles in cell polarity . A recent paper reported that knocking down RGA-3 along with its paralog RGA-4 resulted in changes in the boundary of anterior and posterior domains of PAR proteins in the early embryo [11] . This experiment confirmed our prediction for RGA-3's involvement in cell polarity . Such functional assignment of genes based on phenotypes may seem obvious , since genes sharing similar phenotypes should share similar functions . However , without the in-depth analysis of phenotypic signatures , additional roles of the genes are often neglected . Another example of phenotypic signature is shown for the chromosome function class , which is a relatively large class consisting of 64 genes originally . Its phenotypic signature included P1/AB nuclear separation—cross-eyed , P1/AB nuclei—size/shape , four-cell stage cross-eyed , four-cell stage nuclei—size/shape , and so on ( Figure 6 ) . Using the phenotypic signature , we identified 8 additional genes for this class . The phenotypic profiles of these 8 genes all contain defects other than those included in the chromosome function signature , and thus were originally assigned to other classes . Interestingly , 5 of these 8 genes are known to be involved in nuclear transport functions , suggesting potential connections between nuclear transport and chromosome functions . Evidence supporting their roles in chromosome function has been reported in recent literature . NPP-8 , which is part of the nuclear pore complex , was found to be recruited to the chromatin after anaphase onset in the early embryo [12] . NPP-19 , another nuclear pore complex protein , along with F10C2 . 4 , an uncharacterized gene , were both found to be tightly co-expressed with a group of genes involved in chromosome maintenance [13] . By determining phenotypic signatures and identifying additional genes as belonging to each functional class , we allow genes playing multiple roles in early embryogenesis to be assigned to multiple classes . We define Pleiotropy Index as the number of classes a gene is assigned to . More than half of the genes involved in early embryogenesis are pleiotropic ( i . e . , with Pleiotropy Index ≥2 ) , suggesting that pleiotropy occurs extensively ( Figure 7 ) . Genes that were not assigned to a functional class in the original screen are mostly pleiotropic ( Table S1 ) . Although the profiles of these genes do not resemble those of any other known genes , they now can be decomposed into several phenotypic signatures that lead to functional discoveries . For example , F25H2 . 4 , an uncharacterized gene , is assigned to the classes of cytoplasmic structure , mitochondrial function , meiotic cell cycle progression , and meiosis chromosome segregation . Although pleiotropy is relatively common , only 3% of the genes involved in early embryogenesis are highly pleiotropic ( i . e . , with Pleiotropy Index ≥5 ) . Many signaling proteins show a very high Pleiotropy Index ( Table S2 ) , probably because signaling proteins can be part of various molecular machines functioning in early embryogenesis . For example , of all the 19 kinases involved in early embryogenesis , 18 are pleiotropic ( 95% compared to 59% of all genes involved in early embryogenesis ) , and 5 are highly pleiotropic ( 26% compared to 3% of all genes ) . The biochemical reaction that kinases catalyze is phosphorylation , and a single kinase can catalyze phosphorylation in multiple contexts and with different protein targets . Eliminating a kinase may thus result in multiple sets of defects because a variety of protein targets in different contexts cannot be phosphorylated properly . Since the defects in consideration are not independent of each other , it is possible that the foregoing definition of Pleiotropy Index , although biologically meaningful , can be biased . To resolve this issue , we take the top 33 principal components ( PCs ) of the data matrix , which can account for 90% of the total variation , and regard them as “mega-defects . ” Then , for a gene G , we define its influence from a functional class K as the average of the correlations of this gene's loading vector with those of all the genes in this class ( see Methods ) . A gene G's Relative Pleiotropy Score is the sum of its influences from all functional classes . The Relative Pleiotropy Score does not have direct functional implications as Pleiotropy Index does , but it gives a relative value of how complex a phenotypic profile is and avoids over-counting highly correlated defects . We observe that the Relative Pleiotropy Score such defined is highly correlated with Pleiotropy Index ( Figure S1 ) , indicating that both are reasonable proxies to the concept of pleiotropy . Recent work has revealed a modular organization of genes and proteins in model organisms [13]–[18] . Here a module refers to a group of genes or proteins acting in concert to achieve a certain biological function . However , it is not yet clear how these modules are connected and coordinated . An immediate implication from our finding of pleiotropic genes is that gene modules overlap instead of being separate from one another . We hypothesized that pleiotropic genes act as “connectors” between different modules . The few most highly pleiotropic kinases , for instance , connect most of the major modules in early embryogenesis ( Figure 8 ) . Many cellular events in early development are mediated by protein-protein interactions ( PPIs ) . Complexes or pathways in PPI networks can be the molecular identities of modules . According to our hypothesis , the highly pleiotropic proteins we have identified should reside in central positions in the C . elegans PPI network [13] , [19] . We tested our hypothesis by studying the relationship between a protein's “betweenness” and its Relative Pleiotropy Score or Pleiotropy Index . The betweenness of a given node is defined as the number of times that node is on the shortest paths connecting any two nodes in a network [9] ( see Methods ) . It is a network property that measures the extent to which a node is topologically in a central position between sub-graphs of a network [9] , and it has been applied to characterize modularity of biological networks [20] , [21] . We ranked the betweenness values for early embryogenesis genes that involve two or more interactions in the network , and found that the rank of betweenness is significantly correlated with the Relative Pleiotropy Score ( P-value = 0 . 004 ) ( Figure 9 ) . Furthermore , this statistical significance of the correlation appears to be contributed mostly by a few genes with the highest Relative Pleiotropy Scores . For example , the sum of betweenness ranks for the 12 genes with the highest Relative Pleiotropy Scores is 1123 , whereas the sum of betweenness ranks for 12 randomly sampled early embryogenesis genes is 1794 on average ( P-value = 0 . 01 ) . Similarly , we found that the sum of betweenness values for the 11 genes with the highest Pleiotropy Indices ( Pleiotropy Index≥5 ) is significantly higher than that for 11 early embryogenesis genes chosen at random ( 454701 vs . an average of 179400 , P-value = 0 . 03 ) ( see Methods ) . The betweenness property of highly pleiotropic genes presents supporting evidence to our hypothesis that pleiotropic genes act more as connectors between gene modules .
In this paper , we presented the first systematic investigation of pleiotropic genes in a multi-cellular organism . Using pre-defined functional classes as seeds , we identified phenotypic signatures associated with these classes , and then assigned genes based on their matches to the signatures . We annotated many uncharacterized genes with complex phenotypic profiles by decomposing their profiles into signatures that are indicative of biological functions . We also identified additional functions which were previously unknown for some characterized genes . Our approach can potentially be generalized and applied to many other phenotypic datasets . For example , Gene Ontology categories can be used in place of pre-defined functional classes in order to obtain phenotypic signatures . Furthermore , the reproducibility of detecting defects in RNAi experiments may also be used to define signatures from large amount of phenotypic profiles . Although each gene identified as required for early embryogenesis was assigned to only one class in the original RNAi screen , we found that nearly half of these genes are pleiotropic . Some genes , in particular those encoding signaling molecules , are highly pleiotropic . We examined evolutionary rates of highly pleiotropic genes by comparing sequences from C . elegans and C . briggsae . We found that highly pleiotropic genes evolved at similar rates to other early embryogenesis genes ( data not shown ) , suggesting that pleiotropy may not constitute severe constraints for protein evolution . Our finding is consistent with a previous report that pleiotropic and non-pleiotropic genes evolve at similar rates in yeast [22] . We also assessed the possibility that abundantly expressed genes are more likely to be highly pleiotropic . We retrieved the expression levels of early embryogenesis genes from a SAGE ( Serial Analysis of Gene Expression ) dataset [23] , and correlated with Pleiotropy Index . By performing linear regression we found a significant negative correlation between expression level and Pleiotropy Index ( P-value<0 . 01 ) ( Figure S2 ) . The highly pleiotropic genes tend to be less abundantly expressed than genes assigned with only one or two phenotypic signatures . This is consistent with our observation that signaling molecules such as kinases are enriched in the set of highly pleiotropic genes . The genes involved in cell signaling are often only expressed at a low level but play very important regulatory roles . Finally , we proposed a mechanistic interpretation of pleiotropy from the perspective of functional modules in cellular networks . Since pleiotropic genes are multi-functional , we reasoned that they are likely to coordinate distinct functions involved in early embryogenesis . Consistent with this notion , we found that highly pleiotropic genes exhibit higher betweenness in PPI networks than randomly selected genes . However , there are examples of non-pleiotropic genes showing high betweenness and high pleiotropic genes showing low betweenness . A potential reason is that current PPI data is neither comprehensive nor precise . False positives and false negatives exist in the datasets of genome-wide yeast two-hybrid screens . Consequently , the estimation of centrality based on betweenness may not accurate for every protein in the network . Another possible reason is that mechanisms other than centrality in PPI networks may contribute to pleiotropy . Hodgkin discussed possible underlying mechanisms of pleiotropy and classified them into seven different types [24] . “Combinatorial pleiotropy” , the situation that a protein plays various roles through its various binding partners , is only one type of mechanism . This mechanism is important for the pleiotropy in early embryogenesis , probably because many protein complexes mediate this process . It is not clear yet what mechanisms underlie pleiotropy in other biological processes in multi-cellular organisms . We combined results from two genome-wide RNAi screens [25] , [26] which scored maternal sterility , embryonic lethality , and a limited number of post-embryonic defects with the C . elegans PPI networks . We found 7 genes that exhibited 8 or more of the scored defects and had 2 or more interactions . These 7 genes had a higher sum of betweenness values than that of 7 randomly selected genes , though the P-value of the difference is marginal ( P-value = 0 . 09 ) . This result indicates that PPI networks may contribute to pleiotropy in a broader context , but other mechanisms of pleiotropy probably apply as well . Currently , few datasets that score a large number of phenotypes in detail are available for multi-cellular organisms . The mechanisms underlying pleiotropy are worth further investigations once we have more comprehensive and accurate phenotypic profiles as well as other types of functional genomic data .
Phenotypic profiles were represented as a binary matrix where rows indexed genes and columns indexed defects . Each entry in the matrix was either zero or a positive number , indicating the absence or presence of defects . We obtained control datasets by randomly permuting values among genes for each column while keeping the number of positive cells in each column fixed . We calculated the frequency of occurrence for each individual defect ( F ( i ) ) and the frequency of co-occurrence for each pair-wise combination of defects ( F ( i , j ) ) . For each pair of defects , we calculated the ratio ( R ( i , j ) ) of the observed co-occurrence frequency over the expected frequency as if the two defects occurred independently: R ( i , j ) = F ( i , j ) / ( F ( i ) ×F ( j ) ) . We generated a map of R ( i , j ) using the heatmap function in the statistical language R . There were 22 manually assigned functional classes in the phenotypic dataset . We used genes originally assigned in a class as seeds to identify defects enriched in that class . The collection of enriched defects was defined as the phenotypic signature of the given class . We used the cumulative hypergeometric distribution to determine whether a defect was significantly enriched in a class compared to the whole dataset . In a given class , if the phenotypic profiles of x genes contained a given defect , the P-value was calculated as the following:In this formula , N represents the total number of genes in the dataset; K represents the total number of genes for which phenotypic profiles contain the given defects; n represents the number of genes in the given class . For each functional class , we examined whether any additional genes can be assigned to the given class by matching phenotypic profiles to the identified signature of that class . First , we obtained phenotypic profiles of genes originally assigned to the given class and calculated the average number ( “A” ) of defects matching the signature of that class . Second , we obtained phenotypic profiles of genes not originally belonging to that class and scored them by the number of defects matching the signature . If a gene scored equal to or higher than A , this gene was assigned to the given class . This procedure does not require a perfect match , but it does make the enrichment of defects in the signatures even more enriched in each individual class . In the procedure , we allowed genes to be assigned to multiple classes besides their original assignment , since some genes might play more than one role in early embryogenesis . Phenotypic signatures of different classes contain different sets of defects . In a few cases , the signature of one class ( X ) contains all the defects from the signature of another class ( Y ) . In other words , the defects in the signature of class Y are a subset of that of class X . Thus , a phenotypic profile containing all the defects of the signature for class X automatically contains all the defects of the signature for class Y . In order not to overestimate the degree of pleiotropy , genes with phenotypic profiles matching the signature of X are only assigned to class X , instead of both X and Y . For example , the signature of the protein synthesis class contains all of the defects from the signatures of the cytoplasmic structure , meiosis chromosome segregation , chromosome segregation , and mitochondrial function classes . It can be speculated that blocking protein synthesis results in a number of deleterious effects that resemble perturbing cytoplasmic structure , meiosis chromosome segregation , chromosome segregation , and mitochondrial functions . Thus genes assigned to the protein synthesis class were not considered for assignment to any of the above classes . LPCA is a dimensionality reduction method for binary data [10] . We applied LPCA to the phenotypic profiles of early embryogenesis genes and projected all the defects onto the first two principal components for visualization . The MATLAB code of LPCA was downloaded from www . cis . upenn . edu/~ais/software/lpca_code . tar . We applied PCA to the phenotypic profiles which consist of 661 genes in rows and 45 defects in columns . Eigenvalue diagnosis indicated that 33 principle components accounted for 90% of the variation in the dataset . We calculated an average of Pearson correlation coefficients between the gene of interest and any genes from a given functional class . The relative pleiotropy score is defined as the sum of average Pearson correlation coefficients of all the functional classes . The betweenness of a node is defined as the number of shortest paths running through the node of interest [9] . We computed the shortest paths between all pairs of nodes in the largest component of C . elegans PPI networks [13] , [19] . For each pair of nodes , we enumerated all possible paths in between the chosen pair and increased the betweenness score of the nodes on the shortest paths by one . If there were N alternative shortest paths on route , we split the credit and assigned partial score 1/N to the nodes on the shortest paths . We computed betweenness values for proteins that interact with at least two other proteins , because a protein with only one interacting partner could not be on any shortest paths except for the paths involving the protein itself . We calculated the sum of betweenness values for the early embryogenesis proteins with Pleiotropy Index of 5 or higher . The P-value of significance was estimated by randomly selecting the same number of early embryogenesis genes that had betweenness values and by calculating the sum of their betweenness values . The simulation was repeated 1 , 000 , 000 times .
|
In a biological system , some genes play single roles while others perform multiple functions . How can we determine which genes are multi-functional ? An informative way for probing gene functions is to eliminate the expression of a given gene and observe the phenotypic consequences . RNAi techniques have enabled the generation of genome-wide phenotypic data . Conventionally , genes are clustered into mutually exclusive categories according to the observed defects following RNAi . However , assigning genes that may play multiple roles exclusively into a single category is arbitrary . This paper works out a computational approach that categorizes genes while allowing assignment of genes with complex phenotypes into multiple categories . We apply this approach to genes involved in cell divisions of C . elegans early embryos , and find that about half of these genes can be assigned to more than one functional category . This approach has allowed the identification of previously undiscovered gene functions . We also find that genes playing many roles in early embryos tend to reside in central positions in protein networks . Our approach can be used to perform functional annotations based on phenotypic data in other systems and to identify genes that coordinate multiple biological functions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"developmental",
"biology",
"computational",
"biology/systems",
"biology"
] |
2008
|
Systematic Analysis of Pleiotropy in C. elegans Early Embryogenesis
|
Melioidosis is a neglected tropical disease endemic across South East Asia and Northern Australia . The etiological agent , Burkholderia pseudomallei ( B . pseudomallei ) , is a Gram-negative , rod-shaped , motile bacterium residing in the soil and muddy water across endemic regions of the tropical world . The bacterium is known to cause persistent infections by remaining latent within host cells for prolonged duration . Reactivation of the recrudescent disease often occurs in elders whose immunity wanes . Moreover , recurrence rates in melioidosis patients can be up to ~13% despite appropriate antibiotic therapy , suggestive of bacterial persistence and inefficacy of antibiotic regimens . The mechanisms behind bacterial persistence in the host remain unclear , and hence understanding host immunity during persistent B . pseudomallei infections may help designing potential immunotherapy . A persistent infection was generated using a small-colony variant ( SCV ) and a wild-type ( WT ) B . pseudomallei in BALB/c mice via intranasal administration . Infected mice that survived for >60 days were sacrificed . Lungs , livers , spleens , and peripheral blood mononuclear cells were harvested for experimental investigations . Histopathological changes of organs were observed in the infected mice , suggestive of successful establishment of persistent infections . Moreover , natural killer ( NK ) cell frequency was increased in SCV- and WT-infected mice . We observed programmed death-1 ( PD-1 ) upregulation on B cells of SCV- and WT-infected mice . Interestingly , PD-1 upregulation was only observed on NK cells and monocytes of SCV-infected mice . In contrast , cytotoxic T-lymphocyte-associated antigen-4 ( CTLA-4 ) downregulation was seen on NK cells of WT-infected mice , and on monocytes of SCV- and WT-infected mice . The SCV and the WT of B . pseudomallei distinctly upregulated PD-1 expression on B cells , NK cells , and monocytes to dampen host immunity , which likely facilitates bacterial persistence . PD-1/PD-L1 pathway appears to play an important role in the persistence of B . pseudomallei in the host .
Burkholderia pseudomallei ( B . pseudomallei ) is the causative agent of melioidosis , an infectious disease , endemic across parts of South East Asia and Northern Australia [1] . Despite causing an estimated 89 , 000 deaths worldwide annually , melioidosis still remains a neglected tropical disease [2] . Being a major cause of community-acquired sepsis , melioidosis has a high mortality rate up to 40% [3] . Common routes of infection include percutaneous inoculation , inhalation , and/or ingestion of contaminated particles or aerosols [4] . Although melioidosis can manifest diverse symptoms such as pneumonia and abscesses in various organs including the brain , bacteremic melioidosis with pneumonia commonly leads to early mortality [5 , 6] . Apart from acute infection , B . pseudomallei can cause persistent disease with little or no clinical symptoms over a prolonged period of latency in the host , and only reactivate after years [7–9] . This suggests the likelihood of B . pseudomallei to reactivate only when the host immunity wanes . Indeed , B . pseudomallei can be considered also as an opportunistic pathogen , as melioidosis patients are commonly individuals with at least one or more underlying diseases ( ~80% ) and the elderly [3] . Moreover , recurrence rates in patients can be up to ~13% despite appropriate antibiotic treatments[10] , suggestive of bacterial persistence and inefficacy of antibiotic regimens . The mechanisms behind bacterial persistence in the host remain unclear . Small-colony variants ( SCVs ) representing a sub-population of bacteria have been frequently associated with persistent infections [11–15] . As the name implies , SCVs are slow-growing and form pin-point colonies after 24–72 hours of incubation on agar medium [16] . Although the SCVs of Staphylococcus aureus ( S . aureus ) remain the most extensively studied variant , the morphotypes have also been described in many other bacteria including B . pseudomallei . SCVs are known to be relatively more resistant to antibiotics compared with their wild-type ( WT ) counterparts [17] . In B . pseudomallei , SCVs were reported to display a greater degree of drug resistance [18] . To the best of our knowledge , we are the only group till date that attempted to study WT and SCVs of B . pseudomallei . Our proteomic studies revealed that SCVs and WT pre- and post-infection of A549 lung epithelial cells showed distinct expressions of proteins involved in adhesion , invasion , and virulence ( Al-Maleki et al . , 2014; Al-Maleki et al . , 2015 ) . More importantly , our previous study also demonstrated that SCVs and WT triggered distinct host immune responses during persistent B . pseudomallei infections . Another study also demonstrated that B . pseudomallei can switch to different morphotypes during stress , and have distinct abilities to persist in vitro and in vivo [19] . Hence , these pieces of evidence together suggest that SCVs and WT could play different roles in persistent clinical melioidosis . Programmed death-1 ( PD-1 ) negatively regulates T cell functions , as its engagement with its ligand PD-L1 and PD-L2 arrest T cell proliferation , cytokine secretion , and cytolytic functions [20] . PD-1 is by far the best characterized co-inhibitory molecule associated with T-cell exhaustion in chronic viral infections [21 , 22] . Apart from chronically-infecting viruses [23–25] , many bacteria that cause persistent infections , such as Mycobacterium tuberculosis and Helicobacter pylori ( H . pylori ) , are known to upregulate PD-1 and PD-L1 [26–30] . Persistent B . pseudomallei infections in BALB/c mice also led to PD-1 upregulation on CD4+ and CD8+ T cells , suggestive of T cell exhaustion . This is in line with a previous study that reported on PD-L1 upregulation in polymorphonuclear neutrophils infected with B . pseudomallei , which consequently inhibited CD4+ T-cell functions as well [31] . These results suggest an important role of PD-1/PD-L1 pathway that might potentially be exploited by B . pseudomallei to facilitate persistence in the host . While the role of PD-1 in functional exhaustion is clearly established in T cells , accumulating lines of evidence indicate that PD-1 negatively regulates the functions of B cells , natural killer ( NK ) cells , and monocytes [32–37] . Cytotoxic T-lymphocyte-associated antigen-4 ( CTLA-4 ) represents another co-inhibitory molecule that is inducibly expressed on T cells . CTLA-4 is homologous to CD28 ( the co-stimulatory molecule that provides second signal for T cell activation ) , and inhibits T cell activation [38] . Both CTLA-4 and CD28 engage with two cognate ligands , B7-1 ( CD80 ) and B7-2 ( CD86 ) , although CTLA-4 binds with a greater affinity [39] . Similar to PD-1 , the role of CTLA-4 has been extensively studied in T cells . CTLA-4 upregulation on T cells has been well documented in hepatitis B ( HBV ) and human immunodeficiency virus ( HIV ) infections [38 , 40] . In bacterial infections , CTLA-4 has been reported to cause T cell anergy especially in H . pylori infections in mice , and pathogen clearance was improved following the blockade of CTLA-4 [41] . Notwithstanding , the role of CTLA-4 is well-studied in T cells , its role in other immune cells rather remains ambiguous . To date , very few studies have demonstrated that CTLA-4 inhibits the functions of B cells , NK cells , and monocytes [42–45] . Therefore , it is conceivable to hypothesize that PD-1 and CTLA-4 could dampen host immune responses leading to establishment of persistent infections . We demonstrated previously that persistent B . pseudomallei infections can lead to an increased expression of PD-1 on CD4+ and CD8+ T cells [46] . Here , we aimed to investigate into B cell and innate cell responses , including PD-1 and CTLA-4 expressions , during experimental persistent B . pseudomallei infections . We proposed that persistent B . pseudomallei infections can lead to upregulation of PD-1 on B cells , NK cells , and monocytes , resulting in suboptimal host immune responses . B . pseudomallei employs PD-1/PD-L1 pathway as an immune exhaustion strategy to persist in the host .
All mouse experiments were conducted according to the guidelines of the University of Malaya Animal Care and Use Policy ( UM ACUP ) , and the protocols were reviewed and approved by the Animal Experimental Unit of University of Malaya , Kuala Lumpur , Malaysia ( Ref . No . : 2014-08-05/MMB/R/JSV ) . The Animal Experimental Unit of University of Malaya is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) , and conforms to all government laws and regulations . It provides for approved research and teaching activities , and safeguards the health and welfare of staff and students involved in scholarly activities using animals or animal parts derived from animals . Animals were maintained with controlled temperature , 12h light/dark cycles and given water and feed ad libitum . All efforts were made in order to minimize animal suffering . In addition , all bacterial isolates used in this study were analyzed anonymously . A clinical isolate of B . pseudomallei from a melioidosis case in University Malaya Medical Center ( UMMC ) isolated as previously described was used in the study [47] . The isolate , when cultured on agar medium at 37°C , was found to differentiate into two colony morphotypes , OB ( WT , INSDC: APLK00000000 . 1 ) and OS ( SCV , INSDC: APLL00000000 . 1 ) . OB and OS were characterized using a commercial analytical profile index API 20NE ( BioMėrieux ) test and an in house PCR assay specific for B . pseudomallei [48] . The two morphotypes were cultured on nutrient agar and a single colony of each morphotype was inoculated into Luria-Bertani ( LB ) broth ( Becton Dickinson , Franklin Lakes , New Jersey , USA ) at 37°C overnight in a shaker incubator at 200 revolutions per minute ( rpm ) . Following culture , glycerol ( Acros Organics , Geel , Belgium ) with a final concentration of 30% ( v/v ) was added to the LB cultures and stored at -80°C as a stock culture for the entire duration of the study . Bacterial inoculum of OB and OS was prepared as previously described [46] . Briefly , a single colony of OB and OS from nutrient agar was cultured in LB broth and incubated at 37°C overnight at 200rpm . Later , overnight cultures were adjusted to an OD600 of 0 . 05 with LB broth and incubated under similar conditions . Cultures that reached the mid-logarithmic phase ( OD600 0 . 5–0 . 7 ) were harvested , washed , and re-suspended in phosphate-buffered saline ( PBS ) . Subsequently , the bacterial suspensions were ten-fold serially diluted with PBS until the desired inoculum was obtained . The inoculum was plated on nutrient agar to enumerate colony-forming units ( CFUs ) . Seven to eight-week-old female BALB/c mice obtained from University Putra Malaysia were used in the experiments . All mice were acclimatized for two weeks prior to infection . Mice were under ad libitum feeding conditions . Mice were anaesthetized with isoflurane ( Piramal Healthcare Ltd , India ) , and 10μL of bacterial inoculum was administered via the intranasal route . A persistent B . pseudomallei infection was generated as described previously with minor modifications [49] . Sub-lethal bacterial dose ( ~2–8% of LD50 ) was determined as suggested by Goodyear et al . that used ~5% of LD50 to generate persistent infections in BALB/c mice . Recently , we confirmed that persistent infections with the sub-lethal bacterial dose led to bacterial colonization in the lungs , livers or spleens for ≥60 days by CFU enumeration of these organs , and development of macroscopic hepatic and splenic abscesses in infected mice [46] . Groups of six mice were infected with OB and OS morphotypes , respectively . Only mice that survived for ≥60 days were sacrificed for use in the downstream experiments . One experiment was performed to collect organ samples ( n = 4 per group ) from infected ( OB or OS ) and uninfected mice , respectively for histopathological analysis . Two independent experiments were performed in order to collect adequate sample size ( n = 6 per group ) from OB-infected , OS-infected , and uninfected mice for analysis of immune cells . Mice inoculated with PBS were used as controls , and will be referred to as uninfected mice for simplicity . Mice with persistent B . pseudomallei infections were anaesthetized with isoflurane , and blood was drawn via terminal cardiac puncture . Heparinized blood samples were diluted with PBS at a 1:1 ratio . PBMCs were isolated as described [50 , 51] . Briefly , PBMCs were prepared by density-gradient centrifugation over Ficoll-Paque ( Sigma Aldrich ) . PMBC layer was obtained and washed twice with PBS . Cell viability was determined by 0 . 4% trypan blue ( Life Technologies ) staining . Lungs , livers , and spleens from mice were harvested after withdrawal of blood , and fixed immediately in 10% neutral buffered formalin for 24 hours . Organs were processed as parafilm blocks , followed by the H & E staining , and examined using a microscope . Representative images for each visceral organ were captured . PBMCs ( 1x106 cells in each tube ) were stained with Alexa Fluor 488 hamster anti-mouse CD3e ( BD Biosciences , clone 145-2C11 ) , Pe-Cy7 rat anti-mouse CD4 ( BD Biosciences , clone GK1 . 5 ) , APC-H7 rat anti-mouse CD8a ( BD Biosciences , clone 53–6 . 7 ) , APC hamster anti-mouse PD-1 ( BD Biosciences , clone J43 ) , and PE hamster anti-mouse CTLA ( BD Biosciences , clone UC10-4F10-11 ) , Fixable Viability Stain 510 ( BD Biosciences , cloneR35-95 ) . Corresponding isotype control for each antibody was prepared for appropriate setting of gates during multicolor flow cytometry analysis . All antibodies were pre-titrated for optimal working concentration . Data were acquired on an 8-color FACSCanto II immunocytometry system ( BD Biosciences ) with BD FACSDiva software ( BD Bioscience ) . Data were exported from BD FACSDiva and analyzed using Flowjo software version 10 ( Tree Star , Oregon , USA ) . We used a combination of positive and negative selection strategies to identify B cells , NK cells , and monocytes . NK cells express CD8 , and monocytes express low levels of CD4 [52 , 53] . Therefore , we defined B cells as the lymphocyte population ( FSC-A vs SSC-A ) that was CD3- , CD4- , and CD8- , NK cells as the lymphocyte population that is CD3- and CD8+ , and monocytes as the monocyte population ( FSC-A vs SSC-A ) that was CD3- and CD4dim . Two-tailed Mann-Whitney U test was used to determine statistical significance among different groups , due to the assumption that samples might not follow Gaussian distribution . Results were illustrated using Box-Whisker Plots . All statistical analyses were done using GraphPad Prism 6 software ( La Jolla , California , USA ) . The level of significance was first set at *P<0 . 05 , ** P<0 . 01 , ***P<0 . 001 , and adjusted with appropriate Bonferroni correction .
Morphological differences between OB and OS morphotypes on Ashdown’s agar , which is a selective agar for B . pseudomallei [54] were compared following 24 and 48 hours incubation at 37°C under aerobic conditions ( Fig 1A & 1B ) . OB is the WT , whereas OS is the SCV of B . pseudomallei isolated from the same melioidosis patient . OB developed clear and visible colonies on Ashdown’s agar , while OS could hardly be observed after 24 hours of incubation . Over 48 hours , OB continued to grow larger , whereas OS appeared as small pin-point colonies . This suggests that OS can be differentiated from OB by its slow-growth rate and morphology on Ashdown agar under aerobic conditions . In addition , OB and OS showed distinct morphologies after 72 hours of incubation ( Fig 1C ) . OB appeared as pale purple , rough , wrinkled , and irregular colonies , whereas OS appeared as dark purple , smooth , round and ≥ 2mm diameter colony . Lungs , livers , and spleens harvested from OS-infected or OB-infected mice ( n = 4 per group ) were processed and stained with H & E to investigate the histopathological changes in a persistent B . pseudomallei infection . Non-necrotic solid lung lesions characterized by a discrete focus consisting primarily of mononuclear cells ( Fig 2F ) were observed in infected mice . Livers of infected mice showed lesions with predominantly mononuclear cells ( Fig 2G ) . Cytoplasmic vacuolation was also observed in the hepatocytes of surrounding lesions , characterized by swelling of hepatocytes and clearing of cytoplasm ( Fig 2G ) . Cytoplasmic vacuolation in hepatocytes suggests the likelihood of mild-acute and sub-acute liver injury due to persistent B . pseudomallei infections . Several mice showed splenomegaly with large encapsulated abscess cavities containing neutrophils , mononuclear cells , bacteria , and necrotic cellular debris surrounded by a layer of foamy macrophages , followed by epithelioid macrophages and lymphocytes ( Fig 2H–2J ) . On the other hand , some mice ( n = 2 ) after 60 days of persistent B . pseudomallei infections appeared to have normal red and white pulps with no lesion . Together , these results indicate that intranasal infection of sub-lethal dose B . pseudomallei causes persistent infections that can lead to histopathological changes and systemic spread of the bacteria from the lungs into the livers and spleens . Next , we sought to compare the frequencies of B cells , NK cells , and monocytes in the PBMCs of OB-infected , OS-infected , and uninfected mice . Gating strategy for selection of cell population was illustrated ( Fig 3 ) . Our results revealed that both OS-infected and OB-infected mice had a higher NK cell frequency relative to uninfected mice ( Fig 4B ) . Interestingly , OS-infected mice had a higher NK cell frequency relative to the uninfected mice . No significant differences were found in monocyte and B cell frequencies among OS-infected , OB-infected , and uninfected mice ( Fig 4A & 4C ) . Together , our results suggest a potential role of NK cell in persistent B . pseudomallei infections . PD-1 and CTLA-4 belong to the B7-CD28 superfamily , and their expressions inhibit B cell functions [32 , 35 , 39 , 42 , 43] . Adaptive immune responses play a paramount role against persistent infections . However , the expression levels of PD-1 and CTLA-4 on B cells have seldom been investigated in B . pseudomallei infections . We found that the percentage of B cells that expressed PD-1 was increased in OS-infected and OB-infected compared with uninfected mice ( Fig 5A & 5B ) . Nevertheless , no significant changes were found on B cells that expressed CTLA-4 among the three groups studied ( Fig 5A & 5C ) . Taken together , our results indicate that expression of PD-1 , but not CTLA-4 , could attenuate optimal B cell functions during persistent B . pseudomallei infections . Next , we looked into the innate immunity . PD-1 and CTLA-4 expressions arrest IFN-γ secretion capability of NK cells [34 , 45] . Thus , we examined the profile of PD-1 and CTLA-4 expressions on NK cells . Interestingly , OS-infected mice had a remarkable increase of NK cells expressing PD-1 relative to WT-infected and uninfected mice ( Fig 6A & 6B ) . However , no changes in NK cells expressing PD-1 were observed between the OB-infected and uninfected mice . Interestingly , OB-infected mice had a lower NK cell frequency that expressed CTLA-4 as compared with OS-infected and uninfected mice ( Fig 6A & 6C ) . Our findings demonstrate that SCV B . pseudomallei upregulates PD-1 expression on NK cells in mice during persistent infections , suggestive of NK cell exhaustion . Finally , we investigated the expressions of PD-1 and CTLA-4 on monocytes , as these two co-inhibitory molecules both negatively regulate monocyte functions [36 , 37 , 44] . We observed that OS-infected mice had a higher frequency of monocytes expressing PD-1 compared with OB-infected and uninfected mice ( Fig 7A & 7B ) . No significant differences were observed between OB-infected and uninfected mice . Interestingly , we noticed a remarkable decrease in monocytes expressing CTLA-4 in OS-infected and OB-infected relative to uninfected mice ( Fig 7A & 7C ) . OS-infected mice had a lower frequency of monocytes expressing CTLA-4 compared with OB-infected mice . Together , our results suggest that PD-1 and CTLA-4 are implicated in monocyte functions during persistent B . pseudomallei infections .
Our OS morphotype was stable and reproducible throughout the experiment , as it did not revert back to WT morphology when cultured from glycerol stock , during growth kinetic study and preparation of inoculum for infection . Our previous study showed that OS grew slower on nutrient agar compared to OB [46] . Moreover , our growth kinetic study demonstrated that OS had a defect in growth in vitro , as it grew much slower in Lunia-Bertani ( LB ) broth and reached a much lower OD600 density compared to OB . In this study , OS also grew slower on Ashdown’s agar and had different morphology when compared to OB . We categorized OB as type I ( pale purple , irregular and rough colonies ) and OS as type III ( dark purple and smooth colonies ) morphotype according to Chantratita et al . [19] . Despite previous studies on SCV and WT B . pseudomallei using both in vitro and in vivo model , literature on the pathogenesis of persistent infections due to SCVs and WT still remains scarce [46 , 55 , 56] . Tuchsherr et al . [57] revealed that intracellular infection of endothelial cells with SCVs of S . aureus did not cause any dramatic change in the genes that regulate innate immune responses compared to WT morphotypes . Nevertheless , the study by Tuchsherr et al . could only explain an acute-like intracellular infection in vitro as the infection assay was only conducted for a few hours . Hence , our experimental mouse model serves the purpose of comparing the pathogenesis between SCVs and WT during persistent B . pseudomallei infections . We previously showed that a higher bacterial burden was observed in spleens , but not in the lungs and livers of mice infected with the SCV compared with the WT morphotype [46] . Besides , mice that survived an infection with the SCV of B . pseudomallei for two months were more likely to develop macroscopic liver or splenic abscesses compared with the WT morphotype . Significant changes in lungs , despite using the intranasal route for infection were not observed . This observation can be supported by a previous study on the “persistence model” that demonstrated higher bacterial recovery percentage from livers and spleens compared with lungs after intranasal challenge [49] . To date , only Conejero et al . [58] attempted to characterize the histopathological changes in lungs , livers , and spleens in a chronic B . pseudomallei infection using C57BL/6 mice . In their study , four different types of lung lesions were observed , with two types of them forming granulomas . Additionally , the study also observed pyogranuloma with a necrotic center containing neutrophils surrounded by macrophages , plasma cells , and lymphocytes in the liver . Small pyogranulomas containing neutrophils and macrophages were also common in the liver . Moreover , multifocal to coalescent pyogranulomatous splenitis containing a necrotic center and nonnecrotic microgranulomas consisting epithelioid macrophages were also observed . In our study , contrary to Conejero et al . ’s findings , only one type of lung lesion was observed , which was characterized by a discrete focus of lymphocyte infiltration , with no granuloma ( Fig 2F ) . However , similar findings were observed in many of the lungs of infected mice that had few to no significant lesions . In contrast to Conejero et al . , minor hepatic lesions characterized by infiltration of predominantly mononuclear cells in most of the infected mice were observed . Notably , cytoplasmic vacuolation in hepatocytes of several infected mice was observed , suggesting mild-acute and subacute liver injury due to persistent B . pseudomallei infections ( Fig 2G ) . In contrast to Conejero et al . , normal spleen histology for several infected mice was observed in the present study . Several mice with macroscopic abscesses showed splenomegaly and necrotic pyogranulomas containing neutrophils , which was surrounded by a layer of foamy macrophages , followed by epithelioid macrophages and lymphocytes ( Fig 2H and 2I ) . Different observations compared with Conejero et al . were possibly caused by different strains of B . pseudomallei and mice used in the study . More importantly , Conejero et al . sacrificed mice with chronic B . pseudomallei infections after 20 to 60 days of infections , while our study sacrificed mice only after 60 days of persistent infections . The inconsistent duration of sacrificing mice for histopathological investigation in their study might have contributed significantly to different observations . Mice that were sacrificed on the 20th day could have not survived for a longer period due to a more serious B . pseudomallei infection . This would have led to a more severe histopathological changes and a biased observation . The present study sacrificed mice only after 60 days , which is the period considered as chronic melioidosis [59] . This duration leads to a more accurate and consistent histopathological investigation for persistent B . pseudomallei infections . We speculate that the SCV is more likely to cause severe persistent disease , which was reflected by the higher bacterial load in spleens and more abscess formation in livers and spleens compared with the WT [46] . It is illogical that SCVs result in a more severe pathology than WT , as this process will not benefit the bacteria to persist for a longer duration due to massive host immune responses . However , a recent study by Dietrich et al . [60] demonstrated that non-replicating M . tuberculosis caused a higher CFU and an increased number of granulomas in mouse lungs compared with WT after six weeks . Dietrich et al . suggest that non-replicating M . tuberculosis might undermine host immunity leading to higher bacterial replication and severe pathology . This might be part of the resuscitation process for non-replicating or dormant bacteria to eventually facilitate its transmission . SCVs are similar to dormant bacteria in certain ways , including resistance to antibiotics and slow/zero growth rate . Our results demonstrate that SCVs might employ the same strategy as dormant M . tuberculosis by causing a greater degree of pathology in order to facilitate its transmission in persistent B . pseudomallei infections . This explanation is reasonable as one would anticipate SCVs to relapse in some time in future to transmit the disease . There are only limited studies on the role of B cells , NK cells , and monocytes in melioidosis . Antibodies against B . pseudomallei appears to play a less significant role against melioidosis despite that many individuals still show high seropositivity across endemic regions [61] . B cells were found to play a lesser role in protecting against B . pseudomallei as evident from experiments conducted on B cell-deficient ( μMT ) mice . Nevertheless , it is now clear that μMT mice still produce B cells that could produce other isotypes , raising doubt of using this model to investigate the role of B cells [62–66] . Moreover , immunized mice which produced a high IgG level after lethal challenge had a survival rate of >80% after 40 days [67] , suggestive of a protective role of B cells in experimental B . pseudomallei infections . Here , we showed an increase in B cells expressing PD-1 in the SCV-infected and WT-infected , compared with uninfected mice . This suggests that B . pseudomallei could upregulate PD-1 on B cells to limit optimal B cell functions , which likely affect antibody production , and their interaction with follicular Th cells ( Tfh ) however , may require more investigations . NK cells are a unique population , as many studies have demonstrated that NK cells capture hallmarks of adaptive immunity including antigen specificity and memory responses [68 , 69] . In addition , NK cells can be functionally exhausted similar to T cells during chronic diseases [70 , 71] . NK cells have been identified as the major producer of interferon-γ ( IFN-γ ) in experimental and human melioidosis [66 , 72] . In experimental melioidosis , NK cell-derived IFN-γ showed functional redundancy with IFN-γ released by other immune cells in the first two days of infection [66] . Nevertheless , NK cells could still be playing an essential protective role over prolonged periods of B . pseudomallei infections [66] . Accordingly , our results demonstrated that persistent B . pseudomallei infections can lead to an increase in NK cell frequency regardless of bacterial morphotype differences . In this study , SCV-infected mice had a higher NK cell frequency as compared with WT-infected mice . Interestingly , only SCV-infected mice showed a higher percentage of PD-1+ NK cells , suggesting NK cell exhaustion . WT-infected mice had a lower frequency of CTLA-4+ NK cells relative to SCV-infected and uninfected mice . However , the role of differential expressions of PD-1 and CTLA-4 on the regulation of NK cell activities warrants further investigation . Several findings have indicated that monocytes could possibly play a role in B . pseudomallei infections [73–75] . A recent study on human primary monocytes demonstrated that B . pseudomallei infections stimulated IL-23 production in these cells [75] . Interestingly , in our earlier study , it was demonstrated that persistent infections with SCV B . pseudomallei led to an increase in plasma IL-17A . Briefly , IL-23 is essential for inducing the production of IL-17 , as well as expanding and stabilizing Th17 cells [76] . Thus , activated monocytes could be the major source of IL-23 in maintaining Th17 cells during persistent B . pseudomallei infections . Moreover , high mRNA expression of inflammatory genes in monocytes positively correlated with mortality in patients with sepsis due to B . pseudomallei [74] . Our findings affirmed that SCV-infected mice had a higher PD-1 expression on monocytes compared with WT-infected and uninfected mice . Strikingly , SCV-infected and WT-infected mice had a lower frequency of monocytes expressing CTLA-4 relative to uninfected mice , with the SCV resulting in a lower CTLA-4 expression on monocytes compared with the WT . The observed changes in PD-1 and CTLA-4 expression on monocyte functions would be an interesting future consideration . In this study , we were not able to use definitive approaches including CD19 , CD56 , CD14/CD16 to identify B cells , NK cells , and monocytes . However , the markers used in the current study were based on previous studies conducted on HIV infection [52 , 53] . In addition , we were only able to characterize the expression of PD-1 and CTLA-4 levels on B cells , NK cells , and monocytes without dissecting the functional role of these molecules . Nevertheless , we demonstrated for the first time that persistent B . pseudomallei infections with SCVs can concurrently lead to PD-1 expression on B cells , NK cells , and monocytes in mice , clearly suggesting host immune exhaustion . Remarkably , SCVs caused a higher PD-1 upregulation on NK cells and monocytes compared with WT . Together with our previous work , we could conclude that SCVs caused PD-1 upregulation on adaptive ( T and B cells ) and innate immune cells ( NK cells or monocytes ) , while WT caused PD-1 upregulation only on adaptive immune cells . These observations might be due to the more efficient ability of SCVs in causing host immune exhaustion , or in causing a greater pathology as compared with WT . We speculate that SCVs initiate a higher expression of PD-1 to suppress the host immune responses and facilitate their persistence , causing increased bacterial burden . Interestingly , SCVs and WT were shown to cause CTLA-4 downregulation on NK cells and monocytes . It is unclear whether PD-1 upregulation and/or CTLA-4 downregulation are playing the dominant role over the functions of these immune cells . Future studies should be aimed to investigate the functional role of PD-1 and CTLA-4 on various immune cells in B . pseudomallei infections using murine knockout models and checkpoint inhibitors using in vivo experiments .
|
B . pseudomallei is a bacterium that causes melioidosis , a disease endemic in Southeastern Asia and Northern Australia . It is estimated that melioidosis leads to 89 , 000 deaths worldwide each year . Nevertheless , melioidosis continues to remain a neglected tropical disease that is not even on the list of neglected tropical diseases of the World Health Organization . Furthermore , the disease has a high mortality and recurrence rate , which can be up to 40% and 13% , respectively . It has also been well documented that B . pseudomallei causes latent/persistent infections for a prolonged period without showing apparent symptoms in the infected individual . The mechanisms that are responsible for bacterial persistence in the host remain unclear . Our results demonstrated that B . pseudomallei were able to upregulate PD-1 expression on B cells , NK cells , and/or monocytes during persistent diseases , which likely diminish optimal host immunity . The weakened host immunity in turns facilitates persistence of the bacterium . Interestingly , the SCV had a higher PD-1 expression on distinct immune cells compared to the WT , which might explain its frequent association with persistent infections . Immunotherapies by targeting PD-1/PD-L1 pathway could serve as a better treatment than the conventional antibiotic regimens , which cause a high rate of recurrence in melioidosis patients .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"spleen",
"melioidosis",
"immunology",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"abscesses",
"infectious",
"diseases",
"white",
"blood",
"cells",
"animal",
"cells",
"t",
"cells",
"antibody-producing",
"cells",
"diagnostic",
"medicine",
"cell",
"biology",
"monocytes",
"nk",
"cells",
"b",
"cells",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] |
2017
|
Persistent infection due to a small-colony variant of Burkholderia pseudomallei leads to PD-1 upregulation on circulating immune cells and mononuclear infiltration in viscera of experimental BALB/c mice
|
A persistent question in epigenetics is how heterochromatin is targeted for assembly at specific domains , and how that chromatin state is faithfully transmitted . Stable heterochromatin is necessary to silence transposable elements ( TEs ) and maintain genome integrity . Both the RNAi system and heterochromatin components HP1 ( Swi6 ) and H3K9me2/3 are required for initial establishment of heterochromatin structures in S . pombe . Here we utilize both loss of function alleles and the newly developed Drosophila melanogaster transgenic shRNA lines to deplete proteins of interest at specific development stages to dissect their roles in heterochromatin assembly in early zygotes and in maintenance of the silencing chromatin state during development . Using reporters subject to Position Effect Variegation ( PEV ) , we find that depletion of key proteins in the early embryo can lead to loss of silencing assayed at adult stages . The piRNA component Piwi is required in the early embryo for reporter silencing in non-gonadal somatic cells , but knock-down during larval stages has no impact . This implies that Piwi is involved in targeting HP1a when heterochromatin is established at the late blastoderm stage and possibly also during embryogenesis , but that the silent chromatin state created is transmitted through cell division independent of the piRNA system . In contrast , heterochromatin structural protein HP1a is required for both initial heterochromatin assembly and the following mitotic inheritance . HP1a profiles in piwi mutant animals confirm that Piwi depletion leads to decreased HP1a levels in pericentric heterochromatin , particularly in TEs . The results suggest that the major role of the piRNA system in assembly of heterochromatin in non-gonadal somatic cells occurs in the early embryo during heterochromatin formation , and further demonstrate that failure of heterochromatin formation in the early embryo impacts the phenotype of the adult .
Eukaryotic genomes are packaged into chromatin , which can broadly be characterized as having two alternative forms , euchromatin and heterochromatin . Heterochromatin was first distinguished as dense ( darkly staining ) chromosomal material , seen by microscopy [1] . Since that time , heterochromatin has been investigated extensively in systems from yeast to human to understand its characteristics and its biological significance . Euchromatin is gene-rich , and generally more accessible for transcription , while heterochromatin is gene poor , more condensed and exhibits highly regular nucleosome arrays [2] , [3] . More recently , genome-wide mapping has shown that chromatin consists of numerous different states characterized by different patterns of histone modifications and associated chromosomal proteins [4]–[6] . Euchromatin is enriched in histone acetylation and H3K4me2/3 , marks associated with active transcription; heterochromatin is typically enriched in silencing marks such as H3K9me2/3 , and in heterochromatin protein HP1a [5] . While it is transcriptionally inert compared to euchromatin , heterochromatin plays an important role in a variety of biological processes , including regulation of DNA repair [7] , [8] , maintaining silencing of transposable elements ( TEs ) , and maintaining the integrity of the genome [9] . Mis-regulation of the constituent proteins or regulators of heterochromatin formation will lead to malfunction of the cell , including development of cancers [10] . How the cell decides which regions of the genome to package as heterochromatin , with concomitant gene silencing , is an important question . Studies from diverse systems have indicated a role for non-coding RNA ( ncRNA ) in heterochromatin assembly . In Schizosaccharomyces pombe , the heterochromatic region surrounding centromeres contains dg/dh repeats , which are actively transcribed during S phase and believed to be the source of siRNAs [11] , [12] . The siRNAs produced guide the RNA-induced transcriptional silencing ( RITS ) complex to the regions to be heterochromatized , resulting in localization of histone methyltransferase Clr4 to create methylated histone 3 lysine 9 ( H3K9me2/3 ) . This further stabilizes the RITS complex and leads to binding of the HP1a homolog Swi6 via its chromodomain , resulting in the spread of H3K9me2/3 enriched heterochromatin [12] . Similar RNA-associated heterochromatin targeting mechanisms have also been observed in plants , ciliates , worms ( C . elegans ) , mammals and flies ( Drosophila ) [13]–[19] . In Drosophila , Piwi is the only one of five argonaute proteins ( capable of binding small RNAs ) that both enters the nucleus and plays a major role in silencing TEs . Hence it is considered a likely candidate to play a role in heterochromatin formation [3] , [19]–[21] . Epigenetic signals are responsible for enabling the cells to “remember” the past stimulus , to sustain the chromatin states and transcriptional status that results [21]–[23] . Once established , whether the chromatin states ( specifically histone modification patterns ) can be inherited following mitosis , or whether they depend on a recurrent stimulus from the cis-element , is largely unknown . Some studies suggest that sequence specific elements and the RNA systems involved are required for maintenance of heterochromatin status [24] . However , recent work in yeast and worms suggests that heritable gene expression states and structural heterochromatin ( H3K9me2/3 ) can be maintained in the absence of cis nucleating sequences or the original stimulus [25]–[27] . Drosophila melanogaster , a multicellular system with complex development , is a good choice for investigating this question . D . melanogaster , a model organism used for genetic studies for over 100 years , has been an excellent system for the study of chromatin . Of its 180 Mb genome , about one third is packaged into heterochromatin [28] . A euchromatic gene juxtaposed to a heterochromatin mass , by rearrangement or transposition , exhibits stochastic expression in different cells , so called position effect variegation ( PEV ) [29] , [30] . The silencing of the target gene ( the reporter ) in some cells in which it is normally active is believed to reflect the spreading of the silencing components from the adjacent heterochromatin mass; thus PEV is a sensitive reporter of the heterochromatic environment [19] , [31] , [32] . Screens for suppressors or enhancers of PEV have identified a variety of histone modifiers , chromatin structural components and other chromatin regulators [33] , [34] . The central heterochromatin components , such as Heterochromatin Protein 1 ( HP1a ) and the histone H3K9 methyltransferase ( HMTase ) SU ( VAR ) 3–9 , are found in various species from yeast , to fruit fly , to mammals [35] , [36] , demonstrating the well-conserved mechanisms and evolutionary significance of heterochromatin in eukaryotes . In Drosophila melanogaster , constitutive pericentric heterochromatin is not observed cytologically in the initial zygote , but emerges during blastoderm formation ( ∼2 hour embryo ) [31] , [37] . At this stage , how heterochromatin is established , whether the RNAi system is involved etc , is not clear . Chromatin assembly during this period ( prior to significant zygotic transcription ) is dependent on maternally loaded RNA and protein products , complicating genetic analysis . Analysis using an inducible lacZ reporter has found that silencing occurs at the onset of gastrulation , ∼1 hour after heterochromatin is visible cytologically [38] . Further investigation is needed to determine how the heterochromatin state is transmitted during development . To test whether the RNAi system and other heterochromatin components participate in initiation of heterochromatin formation in the early embryo and in the maintenance of heterochromatin during development , we knocked down the expression of those proteins of interest in different stages to dissect their possible roles . Using a combination of PEV and Chromatin Immuno-Precipitation ( ChIP ) assays , we find that maternal depletion of piRNA component Piwi , or of heterochromatin structural protein HP1a , results in a loss of reporter silencing in the adult , indicating that these proteins are actively involved in heterochromatin assembly in the early zygote; depletion of Piwi is associated with depletion of HP1a at the reporter site . In contrast , in post-gastrulation developing mitotic cells , only depletion of HP1a or H3K9 histone methyltransferase EGG , but not of RNAi components Piwi or AGO2 , leads to loss of reporter silencing . ChIP-array data for HP1a profiles in the piwi mutant confirm that the HP1a level decreases at TE sequences in response to the depletion of Piwi . Our results suggest that the RNAi system plays a critical role in heterochromatin establishment and associated reporter silencing at the heterochromatin/euchromatin border , but plays a minor role in sustaining the chromatin states during development in Drosophila .
HP1a is a critical protein for heterochromatin , thought to be required both for initial assembly and maintenance of heterochromatin structure . Silencing of PEV reporters is known to be sensitive to the dosage of HP1a [39] . While very high levels of expression of the HP1a gene [Su ( var ) 205] are seen in the ovary and in the 0–12 hr embryo , the gene is expressed at moderately high levels throughout the life of the organism [40] . To test the effect of HP1a loss of function in early zygotes on heterochromatin structure in mature stages , we utilized HP1a mutant allele Su ( var ) 20502 , coupled with PEV reporters . The Su ( var ) 20502 loss of function allele has a V26M substitution that alters the chromo domain binding pocket , resulting in a loss of HP1a binding to H3K9me2/3 . Given a wild type father , those offspring inheriting the Su ( var ) 205 wild type allele from Su ( var ) 20502/+ heterozygous mothers experience only maternal depletion of HP1a ( denoted as M in Figure 1 ) , while the Su ( var ) 20502/+ offspring experience both maternal and zygotic depletion ( denoted as M+Z , Figure 1 ) . Note that maternal depletion will impact both events in the female germ line and in the early zygote , based on maternal loading of chromosomal proteins and mRNAs . The reciprocal cross allows us to assay flies with depletion of HP1a in the paternal germ cells ( thought to have minimal impact , so denoted C in Figure 1 ) and in the developing zygote only ( Z , Figure 1 ) . This experiment makes it possible to differentiate between the effects of maternal and zygotic functional HP1a depletion . Using PEV reporter line BL1 ( an inversion allele of the hsp70-lacZ transgenic reporter , with the reporter gene positioned adjacent to a 3L pericentric heterochromatin mass [41] ) , we find that this HP1a mutation leads to a significant increase in hsp70-lacZ expression in the adult fly . In M and Z flies , the β-galactosidase activity is increased by about 2- and 4-fold , respectively ( Figure 1A ) ; this suppression of PEV is additive , as shown by the ∼6-fold increase in M+Z flies ( Figure 1A ) . Similarly , using the hsp70-w reporter 118E10 on chromosome 4 or the classic PEV reporter wm4 ( an inversion on the X chromosome ) , we find that the pigment levels of M , Z and M+Z flies successively increase , reflecting a more accessible chromatin structure at this locus when functional HP1a is reduced either in early zygotes or in developing animals ( Figure 1B ) . This maternal effect is observed in both male and female flies ( Figure 1 , Figure S2 ) . ChIP-qPCR ( Chromatin Immuno-Precipitation – quantitative PCR ) experiments in adult flies were performed to measure the enrichment levels of silencing marks HP1a and H3K9me2 at the 118E10 reporter to explore the perturbation of local chromatin structure . Indeed , both HP1a and H3K9me2 enrichment are decreased at the hsp70 promoter sequence of the reporter in M , Z , and M+Z animals ( Figure 1C ) , consistent with the expectations based on the eye pigment levels . In particular , the H3K9me2 level in C flies is significantly higher than that in M flies , and that in Z flies is significantly higher than that in M+Z flies ( P<0 . 05 , Figure 1C ) , indicating that maternal depletion of functional HP1a , which impacts the early zygote at the time of heterochromatin formation , has a persistent impact; a small but measurable PEV phenotype coupled with some depletion in silencing marks HP1a and H3K9me2 is observed in the adult , even when the developing zygote has two wild type alleles of Su ( var ) 205 . As anticipated , the HP1a mutation also shows an impact on later heterochromatin assembly ( required after each mitosis ) , demonstrating an active role in heterochromatin formation and/or maintenance during later development . While the maternal effects per se are small , one consistently observes greater loss of silencing in the M+Z flies compared to the Z flies , arguing that a deficit during heterochromatin formation cannot be entirely overcome by supplying the required protein later during development . It is of interest to examine the roles of both heterochromatin-specific chromosomal proteins and of the components of the RNAi system , given the potential of the latter to target assembly of the former . Piwi , the piRNA binding protein , carries small RNAs into the nucleus and is reported to be an HP1a-interacting protein; thus it has been suggested to be essential for the accumulation of local silent marks to silence targets transcriptionally [20] , [23] . To ask whether Piwi functions in early embryos and/or later stages during development for the establishment of heterochromatin , we first used the piwi2 null allele to reduce Piwi levels in all cell types . The ovaries of heterozygous piwi2/+ females show a ∼2-fold decrease in levels of Piwi protein compared to wild type ( Figure 2A ) . This presumably leads to a ∼2-fold decrease in maternal loading of Piwi into the eggs from piwi2/+ mothers . Those offspring inheriting the piwi wild type allele from both parents experience maternal depletion of Piwi only ( denoted as M in Figure 2 ) , while the piwi2/+ offspring experience both maternal and zygotic depletion ( denoted as M+Z , Figure 2 ) . Piwi is expressed primarily in gonads in adults . To see whether maternal depletion of Piwi has any effect on silencing in the non-gonadal somatic cells , we assayed the β-galactosidase activity in carcasses and ovaries separately . In M and M+Z adult flies , where Piwi is depleted in early embryos by a maternal effect , the hsp70-lacZ PEV reporter located at the 3L pericentric heterochromatin ( reporter BL1 ) shows increased expression in carcasses relative to the relevant controls ( C and Z ) ( Figure 2B ) . This demonstrates that maternal depletion of Piwi leads to less silencing of the BL1 reporter expression in non-gonadal somatic tissues assayed in adults . In ovaries , while the expression levels of β-galactosidase from the BL1 reporter in M and M+Z flies again are higher than the relevant controls C and Z , we observe that the M and Z levels are comparable , higher than that in C flies and lower than that in M+Z flies ( Figure 2C ) . This result is also clearly shown by X-gal staining of the ovaries ( Figure 2D ) . This pattern suggests that Piwi is important in both embryos and developing female gonadal cells for determining the silent state of the reporter . This early effect in gonadal somatic cells is in agreement with work demonstrating that maternal piRNAs are required to silence retrotransposons in ovarian somatic cells [42] . The zygotic effect observed in gonads is in congruence with the observations that Piwi is required in OSC ( a cell line derived from ovarian somatic cells ) and follicle cells to maintain the transcriptional silencing of transposable elements [21] , [43] , [44] . To further study Piwi's role in heterochromatin silencing in non-gonadal somatic cells , we used the 118E10 and wm4 reporter lines to determine whether maternal depletion of Piwi has any effect in the eye lineages . We consistently observe a weak but significant increase in eye pigment levels in M and M+Z flies compared to control C and Z flies , indicating a maternal effect ( Figure 2E ) . It is interesting that genetically wild type female flies subjected to Piwi depletion as early embryos exhibit a loss of silencing of the PEV reporter in both somatic and gonadal tissues ( M in Figures 2B–E ) , indicating that the impact of the maternal material ( proteins and RNAs ) involved in heterochromatin formation is transmitted to the pole cells as well as somatic cells of the offspring at the early embryo stage [45] . Most likely this impact occurs during the initiation of heterochromatin assembly in early ( 1–3 hr ) zygotes , a stage when Piwi is enriched in the whole embryo , including both somatic cells and germ line cells [46] . To confirm that Piwi has an early effect on heterochromatin establishment , we applied a recently developed approach to knock down ( KD ) its expression in the early zygote by using the female germline-specific nanos-GAL4-tubulin ( NGT ) drivers [47] with the UAS-shRNA transgenic lines produced by the TRiP project [48] , [49] ( Figure 3A ) . The efficacy of this strategy has been assessed by measuring RNA and protein levels of the KD target gene . When females with NGT drivers are crossed to males with an shRNA hairpin targeting Piwi , Piwi mRNA is observed to decrease by 2 fold in 1 . 5–3 hour F1 embryos , presumably as a consequence of the maternally loaded GAL4 driving the paternally provided shRNA hairpin ( Figure 3A ) . The level of Piwi protein is decreased by ∼2-fold in these 1–2 h embryos ( Figure S3 ) . We used the PEV reporter hsp70-lacZ located on the Y chromosome ( BL2 , a translocation of the transgenic reporter from the 3L tip to the Y chromosome , which shows expression in various somatic cells as well as in gonads [41] ) to assay whether knocking down Piwi in the early zygote will lead to any perturbation in heterochromatin structure in later stage animals . Indeed , early zygotic depletion of Piwi ( using the KD strategy in Figure 3A and analyzing the F1 progeny ) leads to suppression of variegation of hsp70-lacZ in both larvae and adult animals . The expression level of hsp70-lacZ , measured by assaying β-galactosidase activity , was elevated by 2-fold in the whole animal ( Figure 3B , male adults , and Figure S4 , male larvae ) . As the NGT driver is highly specific , being expressed in female germ line and silent in male gonads ( Figure S5 , tested by the UAS-mCD8::GFP construct [50] ) , this change must be attributed to events in the early embryo . In addition , the β-galactosidase activity was assayed in larval imaginal discs using X-gal staining; this assay clearly shows that the number of cells with active lacZ expression has been increased following Piwi KD ( Figure 3B , bottom panel ) . The results obtained using the β-galactosidase assay to measure the expression levels of hsp70-lacZ demonstrate that Piwi KD in early zygotes , where heterochromatin formation is initiated , leads to suppression of variegation of this PEV reporter . These findings argue for a role for Piwi in heterochromatin establishment , with the consequences maintained through subsequent mitotic inheritance of this chromatin state during development . To test whether this occurs through a chromatin-based mechanism , we assayed the HP1a levels in the hsp70-lacZ promoter region in adult animals by ChIP-qPCR . The ChIP-qPCR result shows that the HP1a level at the hsp70-lacZ promoter region is decreased by about 2 fold in adults following KD of Piwi in early zygotes ( Figure 3C ) . Thus , the suppression of PEV caused by the depletion of Piwi in the early zygote is a chromatin-based mechanism , in that the increased reporter expression is associated with perturbation of HP1a levels at this site . HP1a levels were tested at other heterochromatic loci to see whether there is a widespread change in heterochromatic regions . Previous work examining the impact of Piwi depletion ( or depletion of another piRNA component , spn-E ) in female germ line nuclei has shown that Piwi plays a role in transcriptional silencing in the germ line for some ( but not all ) transposable elements ( TEs ) , including HeT-A , bari and blood , by an HP1a-dependent mechanism [20] , [23] , [51] . Assaying HP1a enrichment by ChIP-qPCR , we observe that only HeT-A shows a 2-fold decrease at the promoter region on depletion in the early embryo ( Figure 3D ) , while levels of HP1a associated with blood and bari appear unchanged . In addition , neither classic heterochromatin genes light and rolled , nor chromosome 4 genes ci and bt , show a significant alteration in HP1a levels ( Figure 3D ) . This suggests that PEV reporters , at the border of heterochromatin and euchromatin , are particular targets of the mechanism involved and/or particularly sensitive reporters , and that Piwi depletion in early zygotes at this level does not lead to a dramatic impact on the heterochromatin structure as a whole . This result should be anticipated given the survival of the mutant animals . To contrast the roles of those proteins involved in the maintenance of heterochromatin in somatic cells with those not required , we examined the impact of KD of different genes specifically in the eye lineage by using the eye-specific ey-GAL4 driver and corresponding hairpins , assaying the effects on various PEV white reporters . The ey-GAL4 transgene , under the control of the eyeless promoter , is active in the developing eye disc from late embryogenesis until shut off before the last cell division , following the progression of the morphogenetic furrow ( MF ) across the eye disc in third instar larvae ( [52]; Figure 4A ) . The inducible feature of this method allows us to manipulate the expression levels of those genes of interest by controlling the timing in a specific tissue , as shown by others ( e . g . [53] , [54] ) . The efficacy of the ey-GAL4 driver in specifically knocking down HP1a expression in eye lineage cells is demonstrated by a cytological assay using the eye disc . HP1a staining is very faint before the morphogenetic furrow ( where ey-GAL4 is active ) , but becomes much stronger after the morphogenetic furrow ( where ey-GAL4 is shut off ) ( Figure 4B ) . The foci of HP1a dense regions , seen immediately after HP1a expression is restored , overlap with the DAPI-dense regions of the nuclei . The dramatic contrast in HP1a staining on either side of the morphogenetic furrow ( Figure 4B , middle panel , left ) demonstrates that HP1a KD by the ey-GAL4 driver during the eye development is very efficient . The BL2 reporter sequence also carries the white gene down stream of its minimal endogenous promoter sequences [41]; the adjacent hsp70-lacZ and white reporters show concordant expression in the fly eyes [38] , [41] . The eye pigment levels from the BL2 reporter were assayed to investigate the effect of HP1a or Piwi knock down in the developing eye lineage ( using ey-GAL4 ) on silencing . Indeed , HP1a depletion leads to increased expression of the white reporter gene in eyes , while Piwi depletion has no impact ( Figure 4C ) . In addition , pigment assays indicate that KD of HP1a results in a dramatic suppression of variegation for both wm4 and the hsp70-w reporter line 118E10 , while KD of Piwi and AGO2 in these somatic cells does not have any impact on the reporters ( Figure 4D ) . The KD of EGG , which codes for an H3K9 histone methyl transferase , results in different impacts depending on the location of the reporters , as anticipated . For the reporter on chromosome 4 ( 118E10 ) , EGG KD in eye-lineage cells results in significantly increased hsp70-w expression ( suppression of PEV ) , while for the reporter juxtaposed to the pericentric heterochromatin of the X chromosome ( wm4 ) there is less effect . In the latter case , while the quantitative pigment assay does not show a significant difference , the pattern of the pigment in the eye does change , exhibiting a non-uniform distribution instead of a uniform “pepper and salt” pattern ( Figure 4D ) . A specific role for EGG in maintaining the heterochromatic nature of chromosome four has been previously demonstrated [54]–[57] . No impact was seen on KD of G9a , as anticipated . The results above argue that HP1a and EGG , but not RNAi components Piwi or AGO2 , are actively involved in the maintenance of heterochromatin at reporter sites in dividing somatic cells during development after the late embryo stage . The PEV assays shown above argue that for somatic tissues in flies , Piwi's impact on heterochromatin formation stems from its function in the early embryo , a stage at which Piwi is enriched in both pole cells and the bulk of the embryo [46] . After the onset of zygotic transcription , Piwi expression is found primarily in the gonads . The X-gal staining and quantitative β-galactosidase assays ( Figures 2–3 ) suggest that Piwi depletion in early embryos contributes to an altered chromatin state seen in later stage animals . As the majority of the larval tissues are somatic , we used the larval stage to investigate the impact of embryonic depletion of Piwi on the HP1a enrichment profile . Data were obtained from piwi2/piwi2 null larvae ( offspring of piwi2/+ heterozygous parents ) , which have ∼50% maternally loaded Piwi protein compared to wild type in early embryos ( Figure 2A ) . An HP1a ChIP-array assay was performed to study the genome-wide impact of early zygotic Piwi depletion on chromatin structure . In piwi2/piwi2 null larvae ( 50% depletion of Piwi in the early embryo ) , HP1a enrichment ( measured by M value , see Methods for details ) exhibits a small decrease in all heterochromatic regions investigated: pericentric heterochromatin ( 6 . 7% decrease ) , chromosome 4 ( 5 . 7% ) , piRNA clusters ( defined by [58] ) ( 8 . 7% ) . The largest decrease was observed for TEs ( 13 . 4% ) ( Figure 5A ) . We further calculated the HP1a enrichment levels for individual TE classes to look for any differences . In wild type larvae , HP1a enrichment varies among different TE classes , ranging from very little HP1a enrichment ( e . g . roo , DMRP1 and XDMR; blue arrows , Figure 5B ) , to M-values larger than 2 ( gypsy5 , invader3 , DIVER2; brown arrows , Figure 5B ) ( see Methods for detailed definition of the M-value ) . In piwi2/piwi2 null larvae , the HP1a levels for most TE classes were decreased ( 69 out of 83 classes investigated , Figure 5B ) ; this was found for Bari1 , Invader1 , mdg1 and telomere associated Het-A ( red arrows , Figure 5B ) , TEs that are sensitive to Piwi depletion in the germ line [20] , [23] , [51] . A second telomere associated non-LTR element TART also falls into the group with the largest HP1a reduction . However , HP1a levels do not change much for some TE classes , including roo , where HP1a enrichment is very low and does not change on Piwi depletion in the female germ line ( see also [20] ) . Consistent with this observation , we find that depletion of HP1a in parents and developing animals ( Su ( var ) 20504/Su ( var ) 20505 larvae from heterozygous parents , data from [57] ) results in loss of the heterochromatin mark H3K9me2 at TE classes such as HET-A and TART , but not at roo ( Figure S6 ) , suggesting that different TEs have different sensitivity to the Piwi-HP1a silencing system . These observations indicate that HP1a binding at TEs , and the impact of the piRNA system on that binding , varies; the pattern of association seen here mimics that reported in earlier studies of Piwi function in TE silencing in the ovary [20] , [23] Similarly , the HP1a enrichment in piRNA clusters shows an overall decrease in piwi2/piwi2 mutant larvae , but the changes differ in different clusters ( Figure 5B ) . Some piRNA clusters , including the longest piRNA cluster in the 42AB region ( cluster #1 defined by [58]; red arrow in Figure 5B right ) , actually show an increase in HP1a enrichment , consistent with the previous report for this region [59] . But for the majority of the piRNA clusters ( 80 out of 96 investigated , Figure 5B , right ) , HP1a enrichment decreases ( generally to a small extent ) with Piwi depletion . Piwi is reported to be essential for the recruitment of active histone marks in the sub-telomeric 3R-TAS region , with loss of Piwi leading to the increase of HP1a and silencing of white reporters inserted in 3R-TAS in flies [60] . There are few probes in the array that could be uniquely mapped back to the highly repetitious TAS region , so we plot the HP1a enrichment at the tips of the chromosome arms 2L , 2R , 3L and 3R to study the role of Piwi in the recruitment of HP1a in sub-telomeric regions . Indeed , increased HP1a enrichment is observed in the most distal regions of the assembled chromosome sequences ( Figure S7 ) in piwi2/piwi2 null larvae , suggesting an active role of Piwi in those sub-telomeric regions . In piwi2/piwi2 null animals , the gonads are tiny and rudimentary . One might wonder whether the small changes in HP1a association with TEs could result from the absence of gonadal tissues in piwi2/piwi2 null larvae . To consider this question we analyzed the HP1a levels in ovary . The HP1a enrichment for the individual TE classes is highly correlated between ovary and larvae ( Figure S8A , R2 = 0 . 874 ) . If the reduced HP1a level seen for some TEs in piwi2/piwi2 null larvae were a consequence of the depletion of gonadal tissue , one would expect to see that those TEs would have higher HP1a enrichment in ovary compared to larvae ( should fall below the dashed line ) ; this is not observed when the 10 TEs showing the largest HP1a reduction in piwi2/piwi2 null larvae are plotted ( red dots in Figure S8A ) . Furthermore , there is no correlation between the HP1a reduction in piwi2/piwi2 null larvae and the HP1a levels in ovary ( Figure S8B ) . Thus , the observed HP1a change is unlikely to be due to the depletion of gonadal cells in the mutant animals . Reduced HP1a levels in pericentric heterochromatin and TEs have also been observed in newly eclosed piwi null adult flies , which have little gonadal tissue [61] . Overall , the HP1a profiles demonstrate that depletion of Piwi results in a small overall reduction in HP1a at heterochromatic sequences in general , with variation among different heterochromatin classes and elements . Among these , HP1a enrichment drops most significantly at TEs on Piwi depletion . This result argues for a multiplicity of mechanisms for heterochromatin formation , with Piwi playing a significant role at a subset of TEs .
These results , coupled with earlier findings , support a model for heterochromatin targeting that utilizes Piwi in the early zygote ( Figure 6 ) : we suggest that Piwi and the associated piRNA system are required ( directly or indirectly ) to guide HP1a to a subset of TEs , and that the deposition of HP1a further recruits other components to establish H3K9me2-enriched heterochromatin status in those TE regions . Specificity could be achieved via a base-pairing mechanism utilizing piRNAs [19] . Subsequent mitotic transmission of this HP1a/H3K9me2 enriched heterochromatic state during development does not appear to depend on the piRNA system . This targeting mechanism may be of primary importance for TEs in border regions between heterochromatin masses and adjacent euchromatin , the situation for PEV reporters utilized here . When Piwi is depleted , the HP1a level is significantly decreased at these sites . Some loss of HP1a is seen in general in heterochromatic regions , presumably because heterochromatin is enriched in TEs and other repetitious elements . Thus the silencing of PEV reporters , which are dependent on the spreading of the local heterochromatin , can be released . The silent chromatin state is apparently transmitted by the heterochromatin system during development , when the piRNA system is largely absent in non-gonadal somatic cells . Using mutant alleles , we assayed the effect of maternal depletion ( which results in depletion in early embryos ) and zygotic depletion of HP1a or Piwi on the expression of PEV reporters . Functional HP1a depletion in either the early zygote or developing animals leads to suppression of variegation of the PEV reporters , coupled with decreased levels of HP1a itself as well as the silencing mark H3K9me2 in the reporter regions ( Figure 1 ) . This suggests a critical role for HP1a in both early establishment and subsequent maintenance of heterochromatin , and demonstrates that the impact of early depletion can be seen using an adult phenotype , even when wild type alleles of HP1a are present in the developing zygote . In the case of Piwi , only maternal or early zygotic depletion has a significant effect on the reporters in non-gonadal somatic cells ( Figures 2–4 ) . Surprisingly , zygotic Piwi depletion in embryos from wild type mothers does cause a small decrease in PEV silencing of the BL1 reporter in carcasses ( Figure 2B ) , and of 118E10 and wm4 reporters in eyes ( Figure 2E ) . However , this effect is not as significant as that caused by maternal depletion . A small zygotic effect of Piwi depletion is consistent with prior observations [62] . At the same time , the results of Piwi knock down in the eye lineage argue that Piwi is dispensable for the maintenance of heterochromatin silencing after embryogenesis ( Figures 4C , 4D ) . Note that the ey-GAL4 driver becomes active in late embryogenesis , much later than the onset of zygotic expression in the 2-hour embryo . Overall , the data demonstrate that Piwi's role in recruiting HP1a and other components to some TE regions happens early in development , while HP1a is essential for heterochromatin formation during every cell cycle . This is in congruence with their expression patterns . Piwi mRNA is present in gonadal cells and early embryos , with little detectable expression in non-gonadal somatic cells , with some exceptions ( e . g . , larval fat body , possibly nerve cells ) , while HP1a is expressed in all cells/tissues during development [40] , [63] . Thus any HP1a-Piwi interaction likely occurs in gonadal cells and in early embryos , where they are both highly enriched and observed to be nuclear proteins . These stages are also enriched in small RNAs and piRNA pathway components [64] , supporting a model of piRNA-mediated heterochromatin assembly . As it is a structural protein of heterochromatin , one would anticipate that HP1a would be essential for heterochromatin formation in any dividing cell ( such as those in the eye imaginal disc ) when heterochromatin is re-established after DNA replication as is observed ( Figure 4 ) . A second protein found to be important for silencing state maintenance for some reporters is the histone methyltransferase EGG ( Figure 4D; also [54] ) . EGG has been suggested to be essential for heterochromatin formation in specific regions , including chromosome 4 ( in somatic cells; [54]–[57] , [65] ) and piRNA clusters ( gonads; [65] ) . It is of interest that cells in the mature organism “remember” the loss of HP1a in the early zygote , exhibiting HP1a and H3K9me2 reduction in the reporter promoter region in the adult ( Figure 1 ) . The depletion of HP1a at the critical stage of heterochromatin establishment during early development , even when the overall HP1a level is presumably recovered soon after the onset of zygotic transcription , results in diminished heterochromatic regions that apparently cannot be fully re-established , and only partially recover . This implies that both genetic and environmental insults sustained at the critical embryonic stage can have a long-lasting impact on the individual . The reporters exhibiting PEV used here either lie near the break point between heterochromatin and euchromatin caused by inversion or translocation ( e . g . BL1 , BL2 and wm4 ) , or have been inserted into heterochromatic domains by P element transposition ( e . g . 118E10 ) . Their silencing is dependent on the spreading of the adjacent heterochromatin structure , making them sensitive to even small changes in the heterochromatin environment and chromatin assembly systems [66] . For example , when Piwi is knocked down in the early embryo , we observe suppression of variegation of the BL2 reporter coupled with significant HP1a loss at the promoter of the reporter ( Figure 3 ) . However , no dramatic change of HP1a enrichment in is observed in most other heterochromatic regions ( Figure 3 ) . The sensitivity of the BL2 reporter to Piwi depletion might be explained by its position at the edge of a heterochromatic mass , and the requirement for spreading of the heterochromatic assembly . The HP1a ChIP-array data in piwi mutant larvae further confirms that depletion of Piwi will lead to a small decrease in the HP1a level at some TE classes , coupled with an overall small decrease of HP1a levels in heterochromatic sequences ( Figure 5 ) . However , the data obtained from the ChIP-array includes only the unique probes in the assembled genome sequence , so only a small portion of the TEs have been analyzed . It is possible that the actual overall decrease of HP1a enrichment is greater , as most of the TE sequences are not included in this analysis . Nonetheless , the PEV reporters may be particularly sensitive to Piwi manipulation , either because of their dependency on spreading of heterochromatin , or because the Piwi-dependent response itself is triggered by transcription , more likely to occur in these flanking regions . While our studies have focused on the role of Piwi , the resulting model is consistent with earlier work examining several components of the piRNA pathway ( Piwi , Aubergine , Armitage , Spn-E ) . Mutations in these components are reported to have an impact on the repression of transcription and maintenance of a closed chromatin structure for several TE classes when assayed in the female germ line [20] , [23] , [51] . We demonstrate here two additional features: first , that maternal depletion of Piwi has an impact on silencing PEV reporters that can be seen in somatic cells of larvae and adults , and second , that depleting Piwi in early zygotic cells ( but not maternally ) also impacts PEV assayed in later stages . Our results further suggest that the piRNA system observed here most likely acts in the context of multiple mechanisms for heterochromatin formation . In the yeast S . pombe the RNAi system is redundant with other heterochromatin protein interaction systems in heterochromatin establishment [12]; such DNA-protein interaction systems have also been inferred in Drosophila [67] , [68] . The interplay among these systems remains to be investigated . The system of selective depletion developed here should allow further investigation of the role of various components in targeting and maintaining heterochromatin at different heterochromatin domains .
Stocks were maintained and crosses carried out using cornmeal sucrose-based medium [69]; fly culture conditions were set at 25°C with 70% humidity . Stocks used in this study are listed in Table S1 . Ethanol-based pigment extraction and quantification was performed as described in Sun et al 2004 [70] with some minor modifications . Flies were homogenized in 250 µl pigment assay buffer , followed by incubation at 65°C for 10 minutes for pigment extraction . A final volume of 150 µl of pigment extract was used to read OD at 480 nm . For each assay , data from 4–8 samples ( each sample made up of five 3–6 day old flies , randomly picked from the population ) were collected . Detection of β-galactosidase in adult testes and ovaries was performed by a modification of the protocol described in Gonczy et al 1992 [71] . Flies were heat shocked at 37°C for 1 hour and allowed to recover at 25°C for 30 minutes before dissection or freezing for quantitative assays . Ovaries from 2–4 day old flies ( heat shocked as described ) were dissected in PBS-T ( phosphate-buffered saline , 0 . 1% Triton ) and fixed in Glutaraldehyde Fixative ( 2 . 5% glutaraldehyde , 50 mM PIPIEs ) for 10 minutes . Imaginal discs from 3rd instar larvae were dissected in PBS and fixed in PBS-4% formaldehyde for 15 minutes . Tissues were incubated in 0 . 2% X-gal staining solution at 37°C for an appropriate time to visualize staining . The mutant and wild type control samples were handled in parallel to insure equivalent staining conditions and times . For quantitative galactosidase assays , flies or tissues were homogenized in 300 ul of assay buffer ( 50 mM potassium phosphate , 1 mM MgCl2 , pH 7 . 5 ) , followed by spinning to pellet the debris . An aliquot of the extract was transferred to CPRG solution ( 1 mM Chlorophenol Red β-D-galactopyranoside in assay buffer ) and the OD at 574 nm measured at intervals over a 2-hour period . The β-galactosidase activity was calculated as a function of the change in OD . For the whole animal assay , data from 4–8 samples ( five 3rd instar larvae or 2–4 day old flies each , randomly picked from the population ) were generated . The male flies used for the assays in Figure 3 were offspring of 5-day post-eclosion ( or older ) mothers . For the dissected ovaries and carcasses , data from 5–8 samples ( 15 flies each , randomly picked from the population ) were generated . The ovaries were dissected in PBS from 2–4 day old females fed with yeast . RNA was isolated in TRIzol ( Invitrogen ) following the vendor's instructions . For each sample , tissues were homogenized in 0 . 5 ml TRIzol by using an electric grinder for 1–2 minutes . After DNase I treatment , RNAs were used for cDNA synthesis by using oligoA primer and the SuperScrit III kit ( Invitrogen ) . The cDNAs were used as template for quantitative PCR to measure the abundance of a certain transcript . Quantitative PCR was performed using iQ SYBR Green Supermix ( Bio-Rad ) on a Cepheid Smart Cycler . Primers used are listed in Table S2 . Results were analyzed by using the ΔΔCT method [72] using RPL32 as the control locus . Fifteen pairs of ovaries were collected for each RNA sample; for embryos , 200–300 embryos were used for each RNA preparation . Two biological and two technical replicates were performed and analyzed . Imaginal discs were dissected in PBS from 3rd instar larvae , and fixed in 4% formaldehyde in PBS for 15 minutes . The primary antibodies used were C1A9 anti-HP1a ( 1∶10 ) [73] . Alexa Fluor-conjugated antibodies ( Invitrogen ) were applied as secondary antibodies . Images were collected on a Nikon A1 confocal microscope . Western blot analysis was performed by standard methods using whole cell lystae from staged embryos or nuclei from ovaries . Primary antibodies used were: P4D2 anti-PIWI ( 1∶100 ) [74] , 3C7 anti-myosin VI ( 1∶20 ) [75] , and JLA20 anti-actin ( 1∶100; from Developmental Studies Hybridoma Bank ) . Chemiluminescent detection of HRP ( horseradish peroxidase ) conjugated goat secondary antibodies ( KPL ) was performed according to vendors' instructions . Chromatin preparation and chromatin immuno-precipitation were carried out following the modENCODE protocols ( http://www . modencode . org ) using antibodies wa191 anti-HP1a ( 1∶50 ) [76] or anti-H3K9me2 ( 1∶100 , Abcam 1220 ) . Homozygous piwi mutant third instar larvae were recovered from the stocks carrying the piwi2 [77] allele over a GFP balancer by selecting for lack of GFP . For each chromatin preparation , 1 gram of third instar larvae or 2–4 day old flies ( enough for 6 ChIP samples ) are collected and homogenized in liquid nitrogen . Following formaldehyde fixation , nuclei are prepared and lysed , followed by sonication for 5 times 5 minutes using a Bioruptor ( Diagenode ) . The size of the DNA fragments after sonication is about 200–500 bp . The relative enrichment of each mark at the designated region was determined by quantitative PCR ( iQ SYBR Green Supermix , Bio-Rad ) . Primers used are listed in Table S2 . The pull-down efficacy of each ChIP at each locus was determined by using input sample dilutions . Relative enrichment at a given locus was then determined by normalizing the pull-down efficacy of the target locus over α-actinin pull-down efficacy . Two biological and 2–4 technical replicates were performed and analyzed . Array hybridization conditions are as previously described [5] . Two biological replicates are performed for arrays . The M value ( log2 ratio of signal intensities between ChIP and input ) , data normalization and identification of regions ( or peaks ) with significant enrichment were performed as described [5] , [57] , [78] . For heatmap analyses , 500 bp bins were used to average the enrichment levels . Heterochromatin/euchromatin border positions previously identified for larvae by H3K9me2 enrichment were used to define pericentric heterochromatin [78] . The accession numbers for the array data sets are: GSE44884 ( HP1 wa191 . D . mel 3rd Instar Larvae Nuclei piwi2 mutant ) and GSE45523 ( HP1 wa191 . ovary ) .
|
Most eukaryotes harbor a high proportion of transposable elements ( TEs ) in their genomes . Heterochromatin , a condensed chromatin state found at domains enriched for TEs and other repetitious elements , is important for silencing TEs and maintaining the integrity of the genome . The RNAi system has been shown to be important for the establishment and maintenance of heterochromatin in both fungi and plants . To investigate whether this mechanism is also utilized in animals , we selectively depleted the proteins of interest in different developmental stages of Drosophila melanogaster , the fruit fly . Observing reporters subject to Position Effect Variegation ( silencing due to heterochromatin formation ) , and using chromatin immunoprecipitation to establish heterochromatin protein levels at these sites , we find that the piRNA component Piwi is important for heterochromatin establishment during embryogenesis , but not for maintenance of the chromatin state in somatic cells . In contrast , the heterochromatin structural protein HP1a is essential for both establishment and maintenance . This suggests that RNAi components ( specifically Piwi ) play a guiding role in heterochromatin structure formation during embryogenesis , but have a minor role in maintenance . The findings emphasize the importance of maternal materials for the development of the Drosophila offspring , and demonstrate that failure of heterochromatin formation in the embryo impacts the adult phenotype .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
Maternal Depletion of Piwi, a Component of the RNAi System, Impacts Heterochromatin Formation in Drosophila
|
Praziquantel-based mass treatment is the main approach to controlling schistosomiasis mansoni in endemic areas . Interventions such as provision and use of safe water , minimising contact with infested water , disposal of stool in latrines and snail control provide key avenues to break the transmission cycle and can sustain the benefits of mass treatment in the long term . Efforts are also being made to develop a schistosomiasis vaccine which , if effective , might reduce the incidence of re-infection after treatment . However , any interventions deployed need to be acceptable to , and sustainable by , the target communities . In this qualitative study , we investigated the perceptions of six Lake Victoria island communities of Koome , Uganda , about interventions to control Schistosoma mansoni infection and their willingness to participate in Schistosoma vaccine trials . Thirty-two in-depth interviews , 12 key informant interviews and 10 focus group discussions were conducted . Data were analysed using a thematic content approach . Intestinal schistosomiasis was not regarded as a serious health problem because a mass treatment programme is in place . However , the communities lack safe water sources and latrines . Mass treatment with praziquantel , safe water supplies and use of toilets were deemed the most acceptable interventions by the participants . The communities are willing to participate in Schistosoma vaccine trials . Knowledge of a community’s perception about interventions to control schistosomiasis can be valuable to policy makers and programme implementers intending to set up interventions co-managed by the community members . In this study , the views of the Lake Victoria island communities of Koome are presented . This study also provides data to guide further work on alternative interventions such as Schistosoma vaccine trials in these communities .
Schistosomiasis affects an estimated 240 million people worldwide and over 90% of all Schistosoma infections are found in sub-Saharan Africa [1] . In Uganda , an estimated four million people are infected with Schistosoma and about 55% of the population is at risk of infection [2] . The Lake Victoria island communities of Koome sub-county Mukono district carry a large burden of intestinal schistosomiasis: in a recent study 52% of the inhabitants had S . mansoni infections detected on a single stool sample and over 70% using a rapid urine antigen test [3] . The morbidity caused by schistosomiasis is chronic and results in a huge socio-economic burden that is often underestimated [4] . The transmission cycle of intestinal schistosomiasis requires contamination of surface water by egg-laden human excreta , specific freshwater snails as intermediate hosts , and human water contact [5] . To break this cycle , existing intervention strategies include treatment with praziquantel , snail control , proper sanitation and provision of safe water supplies . For the interventions to be effective and sustainable , communities need to be provided with adequate health education [6–8] . Periodic mass treatment of communities with praziquantel is the most widely used approach to control schistosomiasis [9] with numerous gains reported [10 , 11] . Although effective , it does have its drawbacks . Praziquantel does not kill immature schistosomes [12] and therefore does not clear all infection with single treatment , especially in individuals with high intensity infection [13 , 14] . Since praziquantel does not prevent re-infection , the treatment must be provided repeatedly on a regular basis and the side effects can be unpleasant [15] . Due to these drawbacks and other factors such as poor drug coverage and poor drug compliance , control of schistosomiasis by mass treatment with praziquantel has not been consistent in the long-term [11 , 16] . Therefore , effective long-term control of schistosomiasis by praziquantel mass treatment will rely on modifying the other components that facilitate infection transmission [17] . Without those additional interventions , re-infection rates will likely remain very high [18 , 19] . However , a vaccine against Schistosoma could potentially overcome the challenges posed by mass treatment [17]; but currently none is commercially available [20] . Various antigens and vaccine candidates have been proposed [21] and some are investigated in clinical trials [22 , 23] but none is yet readily approved and used for the public . A long-lasting intervention against schistosomiasis needs to be cost-effective , acceptable to and sustainable by the recipient community . Therefore , gauging the readiness of communities in schistosomiasis endemic areas for intervention trials is of paramount importance[24] . In this paper , we investigated the perceptions of island communities with a high burden of the disease about their perceptions of schistosomiasis transmission and control . Specifically , we assessed their knowledge of schistosomiasis , their views on the various control strategies and the control interventions most acceptable to them . We also sought their opinion on their willingness to participate in Schistosoma vaccine trials . This knowledge can be valuable to policy makers and programme implementers intending to transition from project-provided interventions to interventions managed by the community members and local health facilities . This work also provides data to guide further work on alternative interventions such as Schistosoma vaccine trials in these communities .
The study was carried out in the Lake Victoria island villages of Koome sub-county , Mukono district , Uganda . Mass drug administration of praziquantel is being provided to these communities as an intervention in a cluster randomised trial investigating the effects of anthelminthic intervention on health outcomes–the Lake Victoria Island Intervention Study on Worms and Allergy-related diseases ( LaVIISWA ) [3]–in collaboration with the Vector Control Division , Ministry of Health , Uganda . In this trial , 13 villages were randomised to receive the standard intervention against helminths ( single dose praziquantel once a year , single dose albendazole twice a year ) and 13 were randomised to intensive intervention ( single dose praziquantel four times a year , triple dose albendazole four times a year ) . The interventions were rolled out in 2012 and are , in 2017 , on-going . The study presented here was cross-sectional and employed qualitative methods through in-depth Interviews ( IDI ) , key informant interviews ( KII ) and focus group discussions ( FGDs ) . Six of the 27 fishing villages in the sub-county were randomly selected by the trial statistician to participate , taking into consideration that big and small villages were equally represented . Using STATA software ( Stata Corp . , College Station , TX , USA ) , a random selection of participating households in each village was generated . Participants for the in-depth interviews were selected from six households in each village . Adult members of the household who had lived in Koome sub-county for at least 6 months were eligible to participate . On day 1 , the selected households were contacted by the research team , a community leader and a member of the village health team . Eligible household members were invited to participate and appointments made to conduct the interviews during the week ( Monday to Friday ) . One adult member was interviewed from each household , alternating between male and female and choosing the most “senior” adult available in each household ( preferably the household head if this person was of the required gender ) . Six participants were interviewed per village ( the first three male and three female participants to consent ) . Two key informants were purposively selected per village . These were community leaders such as Local Council ( LC ) 1 chairpersons , Beach Management Unit ( BMU ) chairpersons , religious leaders and health workers . Beach Management Units are community fisheries management institutions set up in each fishing village . Two focus group discussions were planned for each village , one for each gender ( male and female ) . Five members were purposively selected for each group by the community leaders . These were community members aged 18 years and above and had lived in the sub-county for at least 6 months . For each village , all the interviews ( in-depth and key informant ) and focus group discussions were conducted in a space of one week ( Monday–Friday ) . Prior to commencement of data collection , the study was presented to the district health team and consultations held with them . Thereafter , meetings were held in each of the six villages to present and explain the work to community members and answer questions about the study . The data were collected by an experienced Social Science interviewer from the research team using both an audio recorder and field notes . Each interview lasted for about an hour . All the interviews and discussions were conducted in Luganda , the local language , using a translated topic guide ( S4 Text , S5 Text , S6 Text ) . After the tools ( information sheets , consent forms , interview topic guides and standard operating procedures ) were developed , the procedures were piloted in one of the study villages . The key informant interviews were conducted in each village to obtain the views of the opinion leaders . The key informant interviews were conducted before focus group discussions . For each FGD , a moderator ( from the research team ) led the discussions , and a note-taker was present to document all verbal and nonverbal responses . To assess awareness of the existence , causes , transmission , health problems and control of schistosomiasis , data were collected , using open-ended questions , on the following: The attitudes of the community members towards the following intervention strategies were assessed: mass administration of praziquantel , disposal of faeces in toilets or latrines , provision and use of safe water supplies , minimising contact with infested water and snail control . To determine the schistosomiasis control interventions most acceptable for the community , views were solicited from the participants on the interventions they thought would work and were willing to adopt . Their opinions were also obtained on who should provide the interventions , how people can be motivated to use them and what the obstacles are . The community members’ willingness to participate in future Schistosoma mansoni vaccine trials was also assessed . Characteristics of a mock vaccine trial were utilized and readiness assessed . The vaccine trial attributes considered were: All the notes and audio recordings were transcribed and the data were analysed manually using a thematic content approach . Responses were categorized into themes and ideas formulated by looking at the pattern of responses . After transcription , data were analysed thematically by closely reading and re-reading the interview scripts looking out for commonalities or recurring opinions and any other thoughts or ideas emerging from the data which formed our themes . The themes were then classified into subthemes and organised in relation to the study objectives . Narrative text was applied around the themes and participant direct quotes were added to illustrate the text . The themes included assessment of community health problems , knowledge and awareness of schistosomiasis , perceptions about interventions to control schistosomiasis , knowledge about vaccines , and willingness to participate in Schistosoma vaccine trials . Ethical approval to conduct this study was granted by the Research Ethics Committee of Uganda Virus Research Institute ( reference number GC/127/15/05/510 ) ( S1 Text ) , the London School of Hygiene and Tropical Medicine ( reference number 10109 ) ( S3 Text ) and the Uganda National Council for Science and Technology ( reference number SS 3831 ) ( S2 Text ) . All participants provided written informed consent prior to the interviews and discussions .
A majority ( 63 out of 94 ) of the participants did not consider schistosomiasis to be a major health problem . To them , schistosomiasis had been a big problem in the past , which has been averted by mass drug administration of praziquantel . Now , schistosomiasis was not considered dangerous during its early stages and was regarded as not affecting activities of daily living . Respondents said that it only becomes more severe if left untreated and results in abdominal distension , body weakness , loss of appetite and eventually death . The scarcity of toilets , safe water sources and health facilities were frequently reported as major health issues . The most frequently mentioned diseases affecting the community were malaria ( mentioned by 32 participants ) , diarrhoea ( 27 participants ) , respiratory infections ( 28 participants ) and HIV ( 25 participants ) . Most of the participants had previously heard about schistosomiasis from community health workers and community leaders . A few said they studied about schistosomiasis at school . However , despite having heard about it , all participants from the in-depth interviews , 7 out of 12 key informants and 33 out of 50 FGD participants reported , incorrectly , that the main source of infection with schistosomiasis was drinking infested water . Fifty-seven participants also correctly stated that contact with infested lake water ( while fishing , fetching and washing from the lake , swimming and playing in the lake ) and open defecation were sources of infection . Three participants said that eating half-cooked food or food that has been contaminated by flies which have come into contact with faeces causes schistosomiasis . Although everyone who lives or works in the study villages was perceived to be at risk of schistosomiasis infection , fishermen were identified as the most at risk . The other groups of people identified to be at risk are women who do laundry from the lake , children who swim in the lake for recreation and the youth who load and off load passenger boats . Most ( 63 out of 94 ) participants stated that it is very hard to control schistosomiasis in their communities . They attributed this to the nature of their activities which revolve around the lake . Six key informants blamed this on the movement of people from one village to another , and inability of the community members to utilise the control measures in place . The control measures suggested by a few participants include improving sanitation , having access to safe and adequate water facilities , increasing the coverage of mass treatment and health education . Having a vaccine was also suggested as one of the interventions . One female participant reported placing water for domestic use in the sun for seven hours as a method she uses to prevent infection . Two female participants stated that their husbands draw water for domestic use from the middle of the lake as a measure to control schistosomiasis . They believed that water in the middle of the lake is unsuitable for the organisms to survive because it is warm . They also said they draw water for domestic use from routes used by ferries and other boats with big engines because they believe the organisms are repelled by the engines . The treatment seeking behaviour of the community members was said to still be poor . Despite the free mass drug administration of praziquantel , some community members are said to be unwilling to accept treatment . One reason respondents cited for not taking treatment was being “too busy with their work” . It was also revealed that residents are reluctant to seek treatment for schistosomiasis during its early stages . This is because , they said , at these stages , they are asymptomatic and their daily activities are unaffected . Lack of health facilities in the sub-county for testing and treating schistosomiasis was also highlighted as a challenge . Mass administration of praziquantel was perceived to be beneficial by 90 of the 94 participants . The fear of side effects notably dizziness , vomiting , fatigue , diarrhoea and loss of appetite was reported by a few participants as the reason for people’s refusal to take praziquantel . Men , especially fishermen , were reported to be the most notorious for dodging mass treatment for this reason . The side effects were described by one participant as being more severe than the disease itself . Most ( 56 out of 94 ) of the participants said the presence of side effects was because of infection with schistosomiasis . They said that the side effects show that the treatment is effective and these effects are short lived , while the benefits last longer . One participant complained about the bitter taste of the treatment tablets . Most participants ( 79 out of 94 ) viewed disposal of faeces in toilets as a good intervention to control schistosomiasis . They reported that disposal of faeces in latrines/toilets also helps to control the spread of other diseases like cholera , diarrhoea , dysentery and other worm infections . They noted that they lack latrines/toilets in their communities and this has resulted in open defecation . The major reason given for not owning a latrine was high cost of materials required for construction . In addition , it was said that the close proximity to the lake renders the soils too weak to keep a latrine firm . Eight participants blamed their landlords for the failure to have toilets in place . Both the community members and the opinion leaders said some traditional beliefs deter people from owning and using toilets . However , one participant said negligence is the only reason why they do not have toilets . She said people in the islands believe that they are always on the move ( temporarily settled ) and so labouring to have toilets in place would be a waste of time and money . Nearly all participants ( 90 out of 94 ) perceived the provision of safe water supplies to be an effective intervention to control schistosomiasis . They reported that this intervention would mostly favour the women , children and people whose work does not involve regular contact with the lake such as bar and shop owners . For this intervention to work more effectively , participants suggested that many safe water facilities must be erected . This will minimise long queues when accessing safe water and make fetching water directly from the lake less appealing . Most of the participants had reservations about the use of biological agents to control snails . Thirty-six participants felt that chemicals may affect the fish negatively because they ( the chemicals ) could be non-selective and kill the fish as well as snails . Two participants were keen to know the dose of the chemical required to kill all the snails . They also felt that this chemical may remain on the shores and not reach farther into the lake rendering it less effective . On the use of biological agents such as fresh water prawns [25 , 26] or ducks [27] to feed on the host snails , the participants were concerned about the numbers required . They expressed the fear that the agents ( especially ducks ) may be poached . Participants said that in the past , there were many wild ducks on Lake Victoria but now , they are almost extinct . On the use of competitor snails [28 , 29] to control schistosomiasis , most ( 70 out of 94 ) participants wondered how snails can be predators of other snails . Some of the community leaders questioned the effectiveness of this intervention . Six participants who showed interest in the biological agents called for community engagement so that people can appreciate the potential of the intervention . There were mixed reactions about this intervention . The majority of female participants perceived it to be a good intervention saying that it is practical , provided there are other water sources in place . The male participants felt that this intervention is ‘unrealistic’ and ‘unwanted’ because it will directly affect their income . Another participant from the female FGD said: Most participants ( 51 out of 94 ) mentioned mass drug administration of praziquantel as an acceptable intervention which they are willing to adopt . This was closely followed by the provision of safe water sources ( 35 participants ) and improved sanitation through disposal of faeces in toilets and latrines ( 30 participants ) . They said that the residents need to be educated on the proper disposal of faeces in order to curb open defecation . Most participants felt that putting up such interventions requires extensive infrastructure which they are unable to provide because of poverty . Forty-four participants ( 47% ) said the government should provide these interventions . Some ( 25 participants ) suggested LaVIISWA should provide , and the rest suggested a joint venture between government and LaVIISWA or between government and non-governmental organisations ( NGOs ) or well-wishers . Twelve participants ( including three key informants ) were unhappy with the way government operates . They said that government has on numerous occasions pledged to provide them with safe water and toilets . These pledges are yet to be fulfilled . To sustain these interventions , participants felt that , once stringent penalties were in place , fines should be levied on those who do not comply . A few participants called for public engagement and health education . They said that community members should be told the advantages and disadvantages of the interventions before any penalties are enforced . Some participants said providing free access to the interventions will motivate people to use them . Participants demonstrated that they had a basic knowledge about vaccines and their role in disease prevention . Many participants acknowledged that they were unaware of the availability of any Schistosoma vaccine . One participant thought that mass treatment with praziquantel was a form of vaccination . All participants said they would welcome a vaccine becoming available . The community leaders also reported that a vaccine would be the best intervention for controlling schistosomiasis . The majority ( 74 out 94 ) of the participants expressed willingness to participate in a Schistosoma vaccine trial . They stated various reasons for the interest in enrolment: service to humanity , benefit from preventing schistosomiasis and trust in the researchers conducting the trial . The reasons for not participating were religious beliefs , conspiracy theories and fear of side effects . The participants were willing to participate in a vaccine trial for a duration of three years although three participants felt that duration was too long . The majority of the participants were willing to accept vaccine administration by injection . Seven participants ( six of them were female ) said they feared the pain caused by injections and would not participate in the trial for that reason . Most participants had no problems with being randomised to any study arm ( including placebo ) . Eight participants preferred the arm with the candidate vaccine and gave it as a pre-condition for participating in the trial . Most participants said that they were willing to provide the required volumes of blood . Eight participants complained about the volumes and said they would not participate . Eleven participants said they should be given food supplements to replace the blood drawn and be provided with medical treatment during the trial . All but three female participants were willing to delay or defer pregnancy during the trial using birth control methods such as oral and injectable contraceptives . Five participants ( all male ) said they should be acknowledged as heroes and given some monetary compensation at the end of the vaccine trial .
In this paper , we have shown that the inhabitants of this schistosomiasis-endemic area prefer mass treatment with praziquantel , safe water supplies and use of toilets to minimising contact with infested water and snail control as the interventions they are willing to embrace . Despite awareness about the existence of schistosomiasis in their communities , they do not consider it as a major health priority because of a mass treatment programme in place . Gaps exist in their basic knowledge about schistosomiasis transmission and prevention such as regarding drinking of infested water as the main source of infection . Provision of mass treatment with praziquantel has faced obstacles such as inadequate supplies of praziquantel , the costs associated with delivery to the target communities and lack of compliance with treatment [30] . Factors such as population migration , change in food supply , conspiracy theories about the intentions of MDA , fear of drug side effects and relations between drug distributors and the target community have been identified to determine MDA success in communities in Uganda [31] . MDA is also not taken up because of inappropriate and inadequate health education and differing biomedical and local understanding of schistosomiasis , absence during drug distribution , pregnancy , breast feeding and feeling healthy [32 , 33] . Indeed the uptake of mass treatment with praziquantel has been sub-optimal in the Lake Victoria island communities [34] and other nearby communities [35] . With support from development partners such as the Schistosomiasis Control Initiative , praziquantel has become more affordable and the supplies more consistent . Logistical support and a motivated drug distribution network in these communities would ensure that the communities access the medication . Health education would address compliance . Providing safe water supplies to the Lake Victoria island communities is still a challenge . Once achieved , the communities need to be educated on the benefits in order to maximise its use . Water for domestic use could be obtained from these safe sources and in the process , vulnerable groups such as women working at home and children would have less contact with the infested lake water . It will still be challenging to stop recreational contact and harder to convince the fishermen to minimise contact with the lake but attempts have to be made . In such a resource limited setting , a concerted effort involving the local communities , government and development partners is required to establish and sustain this intervention . Despite the willingness to use toilets , coverage is very low ( <10% ) in these communities [3] . The communities feel that it is costly to construct and maintain latrines due to the terrain and some landlords are unwilling to provide land . Health education is also key in addressing the misconceptions about the source of infection and wrong beliefs about toilet use . Despite mass treatment with praziquantel , safe water supplies and use of toilets being the most acceptable interventions , the communities felt that they are unable set up and sustain the interventions on their own . Reasons such as prohibitive costs in setting up and maintaining the interventions , mobility of the population , lack of unity in the communities owing to the diverse cultural backgrounds and uncooperative landlords were stated . The communities are willing to participate in sustaining the interventions and they provided suggestions such as setting up stringent byelaws , the need for health education and community engagement . Indeed , health education is important because knowledge about the transmission , severity and consequence of schistosomiasis may be poor [36] . As demonstrated elsewhere , the communities need to be involved in designing the interventions in order to promote ownership of the intervention [24] . To our knowledge , this is the first qualitative study to assess the willingness of a highly endemic community to take part in a potential Schistosoma vaccine trial . The community members were interested and willing to engage in discussion about a trial . However , for the success of such a trial , the concerns raised in this study need to be adequately addressed: the goals of the trial and requirements such as blood sample volume and trial duration need to be clearly explained , and adequate recognition must be given to participants’ contribution to the exercise . As the world targets the elimination of neglected tropical diseases such as schistosomiasis , the perspectives of the target communities about the control strategies do provide very useful insights , especially to policy makers . Community-specific solutions can be designed to address potential barriers to the acceptability and sustainability of an otherwise scientifically proven intervention .
|
Schistosomiasis , a neglected tropical disease caused by the blood fluke Schistosoma , is still a huge burden in sub-Saharan Africa . The modalities for its control are mass treatment of the population with praziquantel , minimising contact with infested water , provision and use of safe water , intermediate host snail control and disposal of stool in toilets/latrines . For sustainable control of the parasite , the recipient communities need to embrace the interventions . In this study , we investigated the perceptions of fishing communities on the Lake Victoria Islands about interventions to control schistosomiasis and their willingness to participate in Schistosoma vaccine trials . We assessed their knowledge of schistosomiasis , their views on the interventions and the interventions most acceptable to them . We show that the community members of this schistosomiasis-endemic area prefer mass treatment with praziquantel , safe water supplies and use of toilets to minimise contact with infested water and snail control . The communities are also willing to participate in Schistosoma vaccine trials . This information is valuable to policy makers and programme implementers intending to set up interventions co-managed by the recipient communities . In addition , the study provides support for future Schistosoma vaccine trials in these communities .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"geomorphology",
"schistosoma",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"surface",
"water",
"water",
"resources",
"landforms",
"helminths",
"topography",
"tropical",
"diseases",
"parasitic",
"diseases",
"animals",
"health",
"care",
"vaccines",
"aquatic",
"environments",
"bodies",
"of",
"water",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"islands",
"hydrology",
"infectious",
"diseases",
"natural",
"resources",
"lakes",
"marine",
"and",
"aquatic",
"sciences",
"helminth",
"infections",
"schistosomiasis",
"freshwater",
"environments",
"eukaryota",
"earth",
"sciences",
"biology",
"and",
"life",
"sciences",
"health",
"education",
"and",
"awareness",
"organisms"
] |
2017
|
Perceptions about interventions to control schistosomiasis among the Lake Victoria island communities of Koome, Uganda
|
Protein signaling networks are static views of dynamic processes where proteins go through many biochemical modifications such as ubiquitination and phosphorylation to propagate signals that regulate cells and can act as feed-back systems . Understanding the precise mechanisms underlying protein interactions can elucidate how signaling and cell cycle progression occur within cells in different diseases such as cancer . Large-scale protein signaling networks contain an important number of experimentally verified protein relations but lack the capability to predict the outcomes of the system , and therefore to be trained with respect to experimental measurements . Boolean Networks ( BNs ) are a simple yet powerful framework to study and model the dynamics of the protein signaling networks . While many BN approaches exist to model biological systems , they focus mainly on system properties , and few exist to integrate experimental data in them . In this work , we show an application of a method conceived to integrate time series phosphoproteomic data into protein signaling networks . We use a large-scale real case study from the HPN-DREAM Breast Cancer challenge . Our efficient and parameter-free method combines logic programming and model-checking to infer a family of BNs from multiple perturbation time series data of four breast cancer cell lines given a prior protein signaling network . Because each predicted BN family is cell line specific , our method highlights commonalities and discrepancies between the four cell lines . Our models have a Root Mean Square Error ( RMSE ) of 0 . 31 with respect to the testing data , while the best performant method of this HPN-DREAM challenge had a RMSE of 0 . 47 . To further validate our results , BNs are compared with the canonical mTOR pathway showing a comparable AUROC score ( 0 . 77 ) to the top performing HPN-DREAM teams . In addition , our approach can also be used as a complementary method to identify erroneous experiments . These results prove our methodology as an efficient dynamic model discovery method in multiple perturbation time course experimental data of large-scale signaling networks . The software and data are publicly available at https://github . com/misbahch6/caspo-ts .
Regarding the training of BNs with respect to multiple perturbation datasets , CellNOpT ( CNO ) [16] assembles BNs from a Prior Knowledge Network ( PKN ) and phosphoproteomic datasets . Their tool has been implemented using stochastic search algorithms ( more precisely , a genetic algorithm ) , to suggest multiple BNs explaining the data [17] . However , stochastic search methods cannot generate a complete set of solutions , hence they cannot guarantee a global optimal solution . In [11 , 12] , the authors overcome this problem by proposing caspo , an approach based on ASP to infer BNs explaining the underlying protein signaling network . This approach can generate all possible optimal Boolean models as compared to the CellNOpt approach . The authors in [14] , presented a framework based on integer linear programming ( ILP ) to learn the subset of interactions best fitted to the experimental data . Recently , another approach based on ILP has been proposed to reconstruct BNs from experimental data . Their learning approach do not require the information about the activation/repression properties of the network’s edges [13] . The methods mentioned above are very useful but restrain themselves to learn from only two time points , assuming the system has reached an early steady-state when the measurements are performed . This assumption prevents us from capturing interesting characteristics like loops [3] . To overcome this issue , the caspo time series ( caspo-ts ) method was proposed in [8] . This method learns BNs from multiple perturbation phosphoproteomic time series data given a PKN . The proposed method is based on ASP and a model-checking step is needed to detect true positive BNs . They tested their approach on synthetic data for a small PKN ( ≈17 nodes and ≈50 edges ) [8] . More recently , an approach based on genetic algorithms was proposed to learn context specific networks given a PKN and experimental information about stable states and their transitions but it does not scale well with large networks and finding a global optimum is not guaranteed [18] .
The structure of the protein signaling network was generated by mapping the experimentally measured phosphorylated proteins ( HPN-DREAM dataset ) to their equivalents from literature-curated databases and connecting them together within one network ( see Materials and methods ) . The reference network ( Fig 2 ) was built using the ReactomeFIViz app ( also called the ReactomeFIPlugIn or Reactome FI Cytoscape app ) [25] , which accesses the interactions existing in the Reactome and other databases [25 , 26] . The PKN shown in Fig 2 consists of 64 nodes ( 7 stimuli , 3 inhibitors , and 23 readouts ) and 178 edges . The learning and testing datasets used in this study were extracted from the HPN-DREAM challenge and correspond to time series protein measurements upon different perturbations of four breast cancer cell lines—UACC812 , BT20 , BT549 , and MCF7 [20 , 21] ( see Materials and methods ) . Since readout signals are measured on variable ranges depending on the protein , a normalization step was necessary . The learning dataset had a few noisy , inconsistent and incomplete time series data points . The caspo-ts system identified these inconsistencies existing in the time series data . The recurrent experimental inconsistency observed was an oscillation in the protein signal upon experimental inhibition of the same protein . To resolve the above mentioned issues , we performed the following data processing steps on the learning dataset: The experimental errors pointed in steps 2-5 were raised as warning or exceptions by caspo-ts . Steps 1 to 5 were applied on the learning dataset . Only step 1 was applied on the testing dataset . In this section , we show the generated BNs for each cell line . For this , we used caspo-ts to learn the BNs from the PKN ( Fig 2 ) and the phosphoproteomic data of four breast cancer cell lines—BT20 , BT549 , MCF7 , and UACC812 . We inferred a family of cell line specific BNs for each cancer cell line and they are shown in the Supplementary Figures ( S1 , S2 , S3 and S4 Figs ) . As explained in the caspo-ts modeling framework section , the caspo-ts method produces BNs fulfilling two criteria , ( i ) satisfaction of the over-approximation criteria ( see Materials and methods ) and ( ii ) optimality with respect to the RMSE objective function . ASP-optimal solutions were fast to collect , their computation time ranged from 36 seconds to 3 minutes depending on the cell line ( S1 Table ) . Afterwards , these ASP-optimal BNs were given to the model-checker for further verification . This second step is more complex and we put a restriction for the computation time of 7 days for each cell line . The number of verified BNs varies from one cell line to another , depending on a number of factors such as the number of perturbations , the order of answer sets in the solutions space , and the perturbation order . The total number of verified ASP-optimal BNs within the 7 days time-frame were 231 , 52 , 188 and 150 for the BT549 , MCF7 , BT20 and UACC812 cell lines respectively . We obtained 191 , 21 , and 72 true positive BNs for BT549 , MCF7 , and BT20 cell lines respectively with an optimal fit to the data . For the UACC812 cell line , we were unable to obtain true positive BNs within the 7 day time limit for verification . Hence , we kept the first 20 BNs from the 150 ASP-optimal BNs for the UACC812 cell line . The caspo-ts method uses the ASP solver ( clingo ) , which is able to exhaustively enumerate all solutions . The clingo solver by default uses an enumeration scheme , in which , once a solution is found , it backtracks to the first point from where the next solution can be found . This typically leads to the situation where successive solutions only change in a small part . As a result , caspo-ts may enter a solution space where BNs are clustered together . We have observed that given the size of the PKN and the small number of perturbations in the experimental data , the solution space of the caspo-ts can be rather very large containing billions of BNs making it difficult to enumerate true positive BNs ( because of the model checking overhead ) in a reasonable time if it gets stuck in a cluster of false positive BNs . An aggregated network was built ( Fig 3 ) by combining the BN families ( with 191 , 21 , 72 , and 20 BNs for BT549 , MCF7 , BT20 , and UACC812 cell lines respectively ) obtained for the four cell lines by keeping the hyper-edges ( Boolean functions ) having a frequency higher than 0 . 3 within each BN family . The frequency is calculated by counting the number of common Boolean functions and dividing it by the total number of Boolean functions within the BN family of each cell line . This aggregated network contains 34 nodes and 74 Boolean functions involving 36 AND gates . As compared to the PKN ( Fig 2 ) , the inferred networks are highly specific to each cell line . In Fig 3 , all cell lines share only 4% of Boolean functions which are shown in thick black colored edges . This shows that the inferred BNs of these four breast cancer cell lines are very diverse and different from each other . To measure cell lines similarity , we calculated the similarity score by applying the Graph Similarity Measure ( see Materials and methods ) on the family of BNs ( with 191 , 21 , 72 , and 20 BNs for BT549 , MCF7 , BT20 , and UACC812 cell lines , respectively ) . This algorithm receives two parameters as input: ( 1 ) one gold standard BN and ( 2 ) a family of BNs . It outputs a score in [0; 1] , measuring the average of the similarity scores between each BN in the family and the gold standard BN . In our case , the gold standard BN is the aggregation of one family of BNs . The similarity scores between all pairs of breast cancer cell lines are shown in Table 1 . Fig 3 agrees with the results presented in Table 1 as we can see the clear discrepancies among the four cell lines . It can be seen that 23% of the Boolean functions are shared among BT549 and MCF7 , and also between BT20 and UACC812 . BT20 shares the least number of Boolean functions ( 15% ) with BT549 . This table revealed pronounced differences among different cell lines of breast cancer . We also analyzed the diversity of Boolean functions among the family of BNs within the same cell line . The similarity among Boolean functions from BT20 ( 0 . 73 ) and MCF7 ( 0 . 63 ) is higher than the ones from BT549 ( 0 . 43 ) and UACC812 ( 0 . 46 ) cell lines . There are a total of 69 distinct Boolean functions shown in Fig 4 along with their respective frequencies . It is interesting to note that the B549 and UACC812 cell lines have more distinct models among their family of BNs with a variable frequency range . This shows that these cell lines have different mechanisms agreeing with the results obtained through graph similarity measure given in Table 1 . Fig 5 shows the common Boolean functions along with their frequency in all BNs . Interestingly , only 4% of the Boolean functions are shared in all cell lines and 88% of these shared functions have the same frequency . In this figure , there is only one Boolean function which is frequent in 3 cell lines and has a lower frequency in BT20 . The union of the BNs learned for each cell line is displayed in the Supplementary Figures ( S1 , S2 , S3 and S4 Figs ) . The caspo-ts method revealed that cell line specific reactions are clustered around the AKT , MAPK3 , and PIK3R1 proteins . PI3K is an important factor for cancer development in HER2 amplified cancers ( UACC812 ) as compared to non-HER2 amplified ( BT20 , BT549 and MCF7 ) cancer cell lines . We can see from the Supplementary Figures ( S1 , S2 , S3 and S4 Figs ) that PIK3R1 exists in all cell lines but is rather more connected in the UACC812 cell line with 10 incoming edges while in others with only 1 incoming edge . The PIK3R1 node in UACC812 ( S4 Fig ) has a centrality measure of 0 . 37 while in the other three cell lines the centrality measure is less than 0 . 11 . The centrality measure is used to quantify the most important node within a network i . e . , the number of times a node has been used as a bridge ( along the shortest path ) to connect to other nodes in the network [27] . It has been established that P1K3R1 ( the regulatory unit of PI3K ) plays an important role in suppressing tumors [28 , 29] . Recently , it has been found that PIK3R1 is mutated in 3% of breast cancer cell lines[30] . Nonetheless , it is worth studying the impact of the PIK3R1 regulatory unit in breast cancer . The performance of the caspo-ts method is evaluated using three criteria: 1 ) RMSE calculation using a typical learning and testing data approach , 2 ) random data comparison , 3 ) AUROC ( Area Under the Operating Curve ) score . The BNs are learned using the learning dataset ( see Materials and methods ) only . The prediction accuracy is evaluated by comparing the RMSE of trajectories in the testing dataset with those recovered by the learned networks ( see Eq 1 ) . There are two types of RMSE—discrete and model . The discrete RMSE is imposed by the discretization of the method . Since we use a discrete learning approach , our recovered traces will be in {0 , 1} and this introduces an error with respect to continuous measurements in [0;1] . The model RMSE refers to the learned BN error with respect to the normalized time series data; that is , the model RMSE is at least as large as the discrete RMSE . When the difference between these two is zero then the inferred BNs are able to recover the discrete trajectories without any error . If the model RMSE is greater than the discrete RMSE then the inferred BNs have some errors in the recoverability of the discrete time series data . To check how our method performs in case of random time series , we have calculated the RMSE score for random data and compared it with learning and testing data . Next , the validity of these networks is verified by comparing them with the canonical MTOR signaling pathway using two parameters , i . e . , true positive rate ( TPR ) and false positive rate ( FPR ) .
As a future direction , we are planning to investigate several aspects of the caspo-ts method , such as ( i ) the order of the solution space of over-approximated Boolean models; ( ii ) the computational time for checking reachability; ( iii ) designing an efficient experimental design strategy and applying it prior to selecting the most informative experiments . Because caspo-ts uses an ASP solver to enumerate BNs , in the resulting sequence of solutions similar BNs are typically clustered together . This can be problematic for large scale problems where we cannot explore the whole solution space in reasonable time . We are currently working on sampling to randomly select BNs from the solution space . Further , we are also studying another technique , which allows for shuffling the order in which solutions are enumerated [32] . We are planning to implement this by dynamically modifying the heuristic of the ASP solver at execution time . Finally , to reduce the false positive rate , we are planning to use multi-shot ASP solving [33] allowing us to customize the search and modify the underlying ASP program at runtime . In our case , we can call the model checker during solving to learn and add constraints to prune wrong BNs early on .
The DREAM portal provides unrestricted access to complex , pre-tested data to encourage the development of computational methods . In this study , we are focused on the HPN-DREAM challenge , which was motivated by the fact that the same perturbation may lead to different signaling behaviors in different backgrounds , making it necessary to build a model which can perform unseen predictions ( absent from the learning data ) . The main goal of the HPN-DREAM challenge is to learn signaling networks efficiently and effectively to predict the dynamics of breast cancer [19] . PKNs are available in different databases such as Reactome , PID , and KEGG among others [26 , 34–45] . We can construct a PKN through a tool such as ReactomeFIViz [25] which is available as a Cytoscape [46] plugin . A PKN alone cannot be used to build reliable dynamical models or to explain underlying biological behaviors [16 , 47] , especially in the case of multiple perturbations data because of the need of specificity . In order to overcome this issue , methods have been proposed which take into account both literature based knowledge and experimental data to build logic models [3 , 7 , 11 , 12 , 16] . In the caspo-ts modeling framework section , we have given the formal definitions of the inputs and the output ( BN ) of the caspo-ts . Here , we formally describe the over-approximation criteria . Finally . we give pseudo encodings of the input of the ASP part of the caspo-ts . This work introduces the study of a graph similarity measure in order to check the variability among the families of BNs generated by caspo-ts . We compare the reactions existing in the gold standard network ( A ) with the family of BNs ( B ) and is based on the Jaccard similarity coefficient which measures the similarity of these models .
|
Traditional canonical signaling pathways help to understand overall signaling processes inside the cell . Large scale phosphoproteomic data provide insight into alterations among different proteins under different experimental settings . Our goal is to combine the traditional signaling networks with complex phosphoproteomic time-series data in order to unravel cell specific signaling networks . In this study , we have applied the caspo time series ( caspo-ts ) approach which is a combination of logic programming and model checking , over the time series phosphoproteomic dataset of the HPN-DREAM challenge to learn cell specific BNs . The learned BNs can be used to identify the cell specific topology . Our analysis suggests that caspo-ts scales to real datasets , outputting networks that are not random with a lower fitness error than the models used by the 178 methods which participated in the HPN-DREAM challenge . On the biological side , we identified the cell specific and common mechanisms ( logical gates ) of the cell lines .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"phosphorylation",
"learning",
"medicine",
"and",
"health",
"sciences",
"breast",
"tumors",
"ontologies",
"protein",
"interaction",
"networks",
"signaling",
"networks",
"social",
"sciences",
"cancers",
"and",
"neoplasms",
"neuroscience",
"learning",
"and",
"memory",
"oncology",
"data",
"management",
"cognitive",
"psychology",
"network",
"analysis",
"information",
"technology",
"data",
"processing",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"mathematical",
"functions",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"breast",
"cancer",
"proteomics",
"biochemistry",
"psychology",
"post-translational",
"modification",
"biology",
"and",
"life",
"sciences",
"cognitive",
"science"
] |
2018
|
Computational discovery of dynamic cell line specific Boolean networks from multiplex time-course data
|
Monocytes are innate immune cells that play a pivotal role in antifungal immunity , but little is known regarding the cellular metabolic events that regulate their function during infection . Using complementary transcriptomic and immunological studies in human primary monocytes , we show that activation of monocytes by Candida albicans yeast and hyphae was accompanied by metabolic rewiring induced through C-type lectin-signaling pathways . We describe that the innate immune responses against Candida yeast are energy-demanding processes that lead to the mobilization of intracellular metabolite pools and require induction of glucose metabolism , oxidative phosphorylation and glutaminolysis , while responses to hyphae primarily rely on glycolysis . Experimental models of systemic candidiasis models validated a central role for glucose metabolism in anti-Candida immunity , as the impairment of glycolysis led to increased susceptibility in mice . Collectively , these data highlight the importance of understanding the complex network of metabolic responses triggered during infections , and unveil new potential targets for therapeutic approaches against fungal diseases .
The immune system is constantly challenged by pathogens , and this requires immune cells to optimize the management of metabolic resources in order to exert their crucial role in host defense . A number of studies have shown how different stimuli induce metabolic reprogramming in immune cells , required for the response against microbial infections [1–3] . The recognition of pathogen-associated molecular patterns ( PAMPs ) triggers substantial changes in cellular metabolism of immune cells , leading to modulation of the effector functions of these cells . Recent studies led to the understanding that the differential use of carbon and nitrogen sources can subsequently affect the immune response . In this sense , proinflammatory macrophages and neutrophils favor aerobic glycolysis over oxidative phosphorylation [4] , anti-inflammatory macrophages rely more on fatty acid oxidation and TCA cycle [5] , whereas full T cell activation also requires the induction of mitochondrial ROS [6] . Candida albicans is a dimorphic fungus that normally colonizes skin and mucosal surfaces in the majority of the healthy population [7] , but in immunocompromised hosts can cause severe life threatening infections [8] . Although the cell wall of both Candida yeast and hyphae contains a variety of glucans , mannans and glycoproteins that can be recognized by a wide range of PRRs , the expression of these molecules greatly varies between yeast and hyphal forms , leading to substantial differences in cytokine induction [9] . Monocytes undergo metabolic and functional reprogramming after exposure to β-glucans and C . albicans yeast , leading to a ‘trained immunity’ functional status characterized by an enhanced cytokine production after secondary stimulation with related or non-related stimuli [10] . In addition , monocytes have been shown to play a crucial role against C . albicans infection , as the deficiency in this immune cell subset has been related with higher susceptibility to fungal infections both in mice and humans [11 , 12] . Few data are available regarding the role of cellular metabolism for the immune function of monocytes , especially the impact on antifungal host defense . This prompted us to study the metabolic pathways triggered by C . albicans in human monocytes after yeast or hyphal stimulation , analyzing the different degree of engagement of the main PRRs involved in C . albicans recognition with the cellular metabolic changes induced , and to study the influence of these alterations on the cytokine response profiles . We report an association between C . albicans-specific recognition by C-type lectin receptors ( CLRs ) and the enhancement of glucose metabolism and aerobic glycolysis in monocytes . These metabolic changes are connected with an enhanced proinflammatory cytokine production in a differential way after yeast or hyphal stimulation . The relevance of these immunometabolic changes was validated in vivo , showing that inhibition of glucose metabolism led to impaired cytokine production , lower fungicidal activity , and a higher susceptibility to systemic C . albicans infection .
Since the molecules expressed in the cell wall of C . albicans can be recognized by a great variety of receptors that could be involved in the upregulation of different metabolic pathways , we measured genome-wide transcriptional profiles in peripheral blood mononuclear cells ( PBMCs ) from healthy volunteers upon stimulation with C . albicans for 4 h and 24 h . Transcriptomic analysis of the genes involved in the main cellular metabolic pathways revealed that the only pathway whose gene expression was consistently upregulated after stimulation was glycolysis , and that this enhancement occurred at 24h , but not at 4h after stimulation ( Figs 1A and S1 ) . We validated the upregulation observed in those genes by qPCR in monocytes isolated from healthy volunteers stimulated with heat-killed yeast or hyphae for 24 h finding a significant upregulation of some of the main enzymes involved in glycolysis such as hexokinase ( HK ) and phosphofructokinase ( PFKP ) . We also found an upregulation of the expression of glutaminase ( GLS ) , an enzyme that allows the entrance of glutamine into tricarboxylic acid ( TCA ) cycle by converting it into glutamate that is subsequently transformed into α-ketoglutarate ( Fig 1B ) . The upregulation of the expression of genes involved in glycolysis has been related to a boost of glucose consumption and lactate production [13] . In agreement with these data , we observed a significant increase in the lactate concentrations released in the supernatants of C . albicans-stimulated monocytes , which was accompanied by an increase in glucose consumption both after yeast and hyphal stimulation ( Fig 2A ) , reflecting the induction of the glycolytic pathway [14] . In line with these data , the increased basal and maximal extracellular acidification rate ( ECAR ) values measured in these cultures reflected an enhancement of the glycolytic activity of monocytes after C . albicans yeast stimulation ( Fig 2B ) . Importantly , the oxygen consumption rate ( OCR ) , which is accepted to be an indicator of the oxidative phosphorylation activity [15] , was also higher in monocytes that had been stimulated with C . albicans both for 4 and 24 h ( Fig 2C ) , also reflecting an enhancement of the oxidative mitochondrial activity in C . albicans-stimulated monocytes . Of note , the OCR/ECAR ratio did not change within the time points measured reflecting a proportional increase of glycolysis and OXPHOS ( S2 Fig ) . In addition to this , C . albicans-stimulated monocytes showed a slightly increased mitochondrial spare respiratory capacity ( SRC ) , especially 4 h after stimulation ( Fig 2D ) . Since the stimulation of monocytes with C . albicans led to an increase in glycolysis and oxidative phosphorylation , we quantified the levels of different metabolites of the TCA cycle . Interestingly , after 4 h stimulation with yeast a slight decrease was observed in the levels of glutamate , fumarate and malate , metabolites than can be synthesized from glutamine [16] . On the other hand , overnight stimulation induced a general increase in the intracellular metabolite levels ( Fig 2E ) , suggesting that within 24 h after stimulation , cells had time to induce an extensive activation of the cellular metabolic machinery and fulfill the energy requirements needed for the functional changes induced by cell activation . Of note , these data correspond to the increases in ECR and OCAR observed 4 and 24 h after stimulation ( Fig 2C and 2D ) , reflecting an enhancement of the cellular metabolic activity after C . albicans recognition by monocytes . We identified no overall changes in cell numbers after stimulation of cells with C . albicans , we can thus conclude that cell growth or an enhanced survival of monocytes stimulated with C . albicans is not the reason of the differences observed in ECAR or OCR . Heat-killing alters the cell wall structure and exposes antigens and PAMPs on the surface of C . albicans yeasts and hyphae [17] , and we validated the results obtained with heat-killed forms of the fungus by using live yeast and hyphae . In order to distinguish between the metabolic changes induced either by yeast or hyphae , we cultured monocytes with a yeast-locked strain of C . albicans ( Δhgc1 ) or with the hyphae-forming wild-type corresponding strain ( hgc1 ) , as the in vivo culture conditions used stimulate hyphal development from live yeast . As described for heat-killed forms , stimulation with live C . albicans increased lactate production by human monocytes ( Fig 2F ) . Of note , the increase in lactate measured after stimulation with live fungal forms was lower than with heat-killed forms , most likely due to the effective masking of β–glucan in the cell wall of live yeasts [18] . In line with this hypothesis , the levels of intracellular metabolites 24 h after infection were significantly higher for hyphae-forming live Candida than for yeast-locked Candida ( Fig 2G ) . Apart from phagocytosis and killing , one of the main effector functions of monocytes and macrophages during C . albicans infection is the production of proinflammatory cytokines , required for the development of a protective immune response [8] . Therefore , we tested how the specific inhibition of various metabolic pathways affected the proinflammatory cytokine production after stimulation with C . albicans-yeast and hyphae ( Fig 3A ) . The inhibition of glycolysis with 2-deoxyglucose ( 2-DG ) , a competitive inhibitor of hexokinase ( HK ) , or with dichloroacetate ( DCA ) , a compound that skews the glycolytic flux through TCA cycle by reducing the transformation of pyruvate into lactate by enhancing the activity of pyruvate dehydrogenase ( PDH ) , strongly downregulated C . albicans-induced IL-1β , TNFα and IL-6 production in human monocytes ( Fig 3B ) . Monocyte training with β-glucan , a ligand from C . albicans cell wall , has been reported to cause a switch from oxidative phosphorylation to aerobic glycolysis via activation of the PI3K-Akt-mTOR axis in human monocytes [19] . We found that inhibition of the mTOR pathway with Torin1 ( a direct mTOR inhibitor ) or with AICAR ( an indirect mTOR inhibitor via AMPK activation ) caused a decrease in the cytokine production by human monocytes after stimulation with yeast , but not with hyphae ( Fig 3B ) . In addition , inhibition of glutaminolysis by BPTES , a selective inhibitor of glutaminase ( GLS ) , also impaired the production of IL-1β and IL-6 in yeast-stimulated monocytes , although to a lower extent ( Fig 3B ) . The inhibition of β-oxidation with etomoxir or the interference of the pentose phosphate pathway with 6-aminonicotinamide ( 6-AN ) did not produce any significant differences in the production of the cytokines measured ( Fig 3B ) . Importantly , impairment of oxidative phosphorylation with oligomycin , an ATP synthase inhibitor , caused a significant decrease in the proinflammatory cytokine production after stimulation with C . albicans yeast ( Fig 3B ) , and this effect is consistent with the increased OCR reported after C . albicans stimulation ( Fig 2B ) . We confirmed the importance of glycolysis in these processes as 2-DG treatment of monocytes abolished the differences observed in ECAR and OCR after C . albicans recognition ( Fig 3C and 3D ) . As a whole , these data suggest that C . albicans yeast-induced cytokine production in monocytes relied on an mTOR-dependent enhanced glycolysis as well as on an increased oxidative phosphorylation activity of the cells and , to a lesser extent , on glutamine metabolism . In the case of hyphae-induced cytokine production , these data suggest that it mostly relies on glycolysis . Since Th1 and Th17 responses have been reported to play a protective role in C . albicans infection [20 , 21] we also measured how the inhibition of the different metabolic pathways affected the production of Th1/Th17-derived cytokines after C . albicans stimulation of human PBMCs . We found that the inhibition of glucose metabolism notably impaired IL-17 , IL-22 , IFNγ and IL-10 production , while glutamine metabolism also played a role in the production of Th17-derived cytokines after yeast but not after hyphal stimulation ( S3 Fig ) . In the case of the anti-inflammatory cytokine IL-10 , its production after hyphal stimulation was only significantly affected by glycolysis inhibition , suggesting that the metabolic pathways leading to the production of pro- or anti-inflammatory cytokines might be regulated by different metabolic routes , as already described for macrophages [22] . C . albicans recognition by human cells is known to rely on a wide series of Pattern Recognition Receptors ( PRRs ) such as Toll-Like Receptors ( TLRs ) and C-type lectin receptors ( CLRs ) [8] . Since C . albicans-induced cytokine production in human monocytes seemed to be under the control of glycolysis , we wondered which receptors were responsible for triggering the metabolic changes reported . To this aim , we blocked different PRRs involved in C . albicans recognition and assessed lactate production in cell supernatants after overnight culture . Interestingly , neither the treatment of monocytes with a specific mAb against TLR2 , nor the blockade of TLR4 with Bartonella quintana LPS , a natural antagonist of this receptor [23] produced any changes in the lactate production triggered by heat-killed Candida stimulation , indicating that TLR-derived signaling did not play a role in the induction of glycolysis after Candida recognition ( Fig 4A ) . Nevertheless , blockade of C-type lectin receptors with laminarin ( a dectin-1 specific antagonist ) or with a specific mAb against mannose receptor ( MR ) , caused a significant decrease in the lactate production measured upon stimulation with heat-killed Candida yeasts , but not hyphal stimulation . Of note , blockade of CR3 , a receptor that has a lectin domain involved in C . albicans [24] and β-glucan [25] recognition , also led to a decrease in the extracellular lactate levels determined after C . albicans yeast stimulation ( Fig 4A ) . These results reflect that metabolic changes induced by recognition of C . albicans by monocytes were mainly driven by CLR-mediated rather than TLR-mediated signaling , in contrast to the metabolic rewiring induced by bacteria [26] . These data also confirmed the differences in the intracellular metabolic requirements triggered after yeast or hyphal recognition . PRR blockade after monocyte stimulation with live yeast-locked C . albicans did not produce any significant changes in lactate production except for the case of monocytes with an impaired dectin-1 signaling , which showed a discrete reduction in this readout ( Fig 4B ) . This can be again attributed to the low degree of β-glucan exposure in the cell wall of the live wild-type C . albicans [27] . The differences seen after live dectin-1 blockade in wild-type C . albicans-stimulated cells can be explained by the fact that β-glucan , the dectin-1 ligand , is highly exposed in the cell wall of newly-formed hyphae as described by Cheng et al . [18] , and further confirmed that yeast and hyphae triggered metabolic changes in human monocytes in a differential fashion . A number of studies have related the production of Reactive Oxygen Species ( ROS ) with the resolution of C . albicans infection [8 , 28] . We hypothesized that ROS induction could be affected by monocyte metabolism after C . albicans stimulation . Glycolysis inhibition with 2-DG prior to stimulation with yeast or hyphae almost completely abolished ROS production by monocytes ( Fig 5B ) . In line with this , the impairment of the glycolytic routes with DCA treatment also impaired ROS production strongly ( Fig 5C ) . We did not find any other metabolic pathways involved in ROS production after C . albicans stimulation ( S5 Fig ) , except for the case of the pentose phosphate pathway , for which the specific inhibition of its oxidative branch with 6-AN led to a strong decrease in ROS production ( Fig 5D ) . This can be related to the drop in availability of NADPH , a key factor for the induction of ROS , as already described for LPS-activated macrophages [29] . On the other hand , phagocytosis of C . albicans yeast was not significantly affected by treatment of monocytes with 2-DG ( S6 Fig ) . Because the data presented above argue for a crucial role of monocyte glycolysis for antifungal host defense , we wanted to validate these results in an in vivo model of systemic C . albicans infection . In this model , C57BL/6 mice were intravenously injected with a single dose of 105 colony-forming units ( CFU ) of C . albicans yeast , causing a disseminated infection [30] . In order to validate the role of glucose and glutamine metabolism in vivo , we treated these mice with 2-DG or BPTES prior to systemic C . albicans intravenous challenge and evaluated the systemic response to infection . Treatment with 2-DG led to a significant increase in the fungal burden measured in the kidneys of these mice 5 days after C . albicans infection , while BPTES-treated animals had a fungal burden comparable to control individuals ( Fig 6A ) . Of note , one of the mice treated with 2-DG had to be euthanized during the experiment due to the infectious process . We also assessed the candidacidal activity of blood neutrophils , which have been described to be the main effector cells in this model of infection [8] , finding a strong impairment of their fungicidal potential in the case of 2-DG treated mice ( Fig 6B ) . Mouse neutrophils treated in vitro with 2-DG also had a significantly lower candidacidal activity than control cells ( S7 Fig ) . We also measured cytokine production after C . albicans restimulation of splenocytes obtained from 2-DG or BPTES-treated mice after the infection , finding a significant reduction in the production capacity of IL-1β , IL-6 , IL-10 , TNFα and IFNγ in mice treated with 2-DG . Thus , the impairment of glucose metabolism alters the capacity of splenocytes to respond to a secondary C . albicans stimulation ( Fig 6C ) . In mice treated with BPTES , we found reduced levels of IL-6 , which is in agreement with the data obtained from human monocytes . Therefore , while the inhibition of glutamine metabolism seems to have a relatively small effect in systemic antifungal response in vivo , these data confirm that glycolysis plays a central role in the induction of an effective anti- C . albicans host response both in vitro and in vivo ( Fig 7 ) .
C . albicans is the most important fungal pathogen , and immunotherapeutic approaches to boost antifungal host defense are urgently needed to decrease mortality in systemic candidiasis which currently reaches up to 30–40% [31] . Here we aimed to investigate for the first time the role of cellular metabolism of immune cells for the induction of an effective immune defense against C . albicans . We observed that C . albicans yeast and hyphae induce differential rewiring of cellular metabolism: while yeast-stimulated monocytes rely on glycolysis , oxidative phosphorylation and glutaminolysis to mount cytokine responses , monocytes stimulated with hyphae rely mainly on glycolysis . These processes are mediated by fungal recognition by CLRs , but not by TLRs , and glycolysis is crucial for an effective host defense in vivo against disseminated candidiasis . The metabolic circuits triggered in immune cells following pathogen recognition are very complex . Microbial stimuli such as LPS promote glucose conversion into lactate and decrease oxidative phosphorylation in monocytes and macrophages , in a process known as the Warburg effect [4] . However , a recent study demonstrated that induction of the Warburg effect in monocytes is a specific feature of LPS engagement of TLR4 , as the engagement of other TLRs by their specific ligands or by complete microorganisms led to a much more complex response , notably a strong increase in the oxidative phosphorylation activity of the cells after stimulation [26] . While several studies investigated the immunometabolic circuits involved in anti-bacterial responses , very few have addressed the role of these pathways in anti-fungal immunity . In this report , we describe differential roles for glucose metabolism , glutamine metabolism , oxidative phosphorylation , and the pentose phosphate pathway in the immune response against C . albicans . The increase of glycolysis following microbial stimulation of myeloid cells has been reported in several studies , being usually linked to an enhanced function of mTOR-related pathways [19 , 32] . Our data confirmed that after C . albicans recognition , human monocytes underwent an increase not only in their glycolytic activity , as demonstrated by the higher glucose consumption , lactate production and ECAR reported after stimulation , but also in their oxidative phosphorylation capacity and OCR . This enhancement of both aerobic glycolysis and oxidative phosphorylation is reminiscent of the metabolic stimulation induced by other whole microorganisms [16 , 26] , in contrast to purified ligands such as LPS or β-glucan [19 , 33] , reflecting the complex nature of the cellular metabolic networks induced by the engagement of different PRRs . An important discovery is the difference in induction of cellular metabolism between C . albicans yeasts and hyphae . The differences in the abundance and the degree of exposure of cell wall components such as β-glucans , mannans or glycoproteins between yeasts and hyphae have been described to induce distinct profiles of cytokine responses [9] . Along this line , we showed that stimulation with yeast or hyphae led to different metabolic responses in monocytes . Heat-killed C . albicans yeast induced production of proinflammatory cytokines , through a process highly demanding for the cells , which makes use of glycolysis , oxidative phosphorylation and glutaminolysis in order to fulfill the expensive energy requirements . In the case of heat-killed C . albicans hyphae , cytokine production was solely dependent on glycolysis . The transition to hyphal growth creates the opportunity for improved recognition of inner cell wall components such as β–glucan , which become more exposed , allowing their interaction with PRRs and triggering the proinflammatory cytokine production [18] . The data presented here , correlates with previously published data showing a strong induction of glycolysis by β–glucan [19] and argues for a model in which the poor β–glucan recognition in yeasts [18] requires different recognition systems for activation of multiple metabolic pathways , while the broad exposure of β–glucan by hyphae induces a much stronger induction of glycolysis in the immune cells which is sufficient to fulfill their energy requirements . Glucose that enters the cell is not solely processed via glycolysis , but can be also transformed into fatty acids , aldoses , glycogen , or enter the pentose phosphate pathway to generate NADPH and pentoses [34] . ROS production has been historically linked to NADPH oxidase and mitochondrial activity [35] , and several studies have assigned an important role for ROS production in host defense against microbial infections [29 , 36] . In our study , we found that glucose metabolism is crucial for the generation of ROS in human monocytes after C . albicans stimulation . Nonetheless , this could not be only attributed to glycolysis , as the impairment of the pentose phosphate pathway also induced a severe impairment of ROS production , as similarly described for LPS-activated macrophages , where ROS were generated through pentose phosphate pathway-dependent NADPH-oxidase activity [29] . These data suggest a differential role for the different metabolic pathways triggered after C . albicans stimulation in monocytes . While glycolysis appeared to play a central role in cytokine and ROS production , other routes such as oxidative phosphorylation or glutaminolysis seemed to play a preferential role in fueling cytokine induction . In contrast , the pentose phosphate pathway , which did not play a role in cytokine production , seems fundamental for ROS generation . These results further emphasize the variety and the specificity of the metabolic changes that cells undergo after making contact with a pathogen , highlighting the necessity of studying the distinctive features of the various stimuli . The presence of suitable carbon sources in the environment is fundamental not only for the host but also for Candida cells . Regarding this , some studies have shown that the exposure of Candida to high concentrations of lactate is able to modulate its cell wall architecture , drug resistance and virulence [37 , 38] . In our study , the concentrations of lactate reached after stimulation are much lower than those demonstrated to alter cell wall structure of C . albicans , being glucose the major carbon source in all the experimental conditions tested . Therefore , in this case we find improbable that any of the experimental conditions might have been affected by the minor presence of lactate in the medium , compared to glucose . TLR-mediated immunometabolic reprogramming have been reviewed by a number of authors [39–41] . C . albicans expresses a large variety of structures in its cell wall , which represent PAMPs that bind to different families of PRRs on the immune cells , among which CLRs and TLRs are the most important [8] . Especially deficiencies in CLRs lead to an impaired cytokine production and higher susceptibility to C . albicans infections [42 , 43] . In line with this , blockade of recognition of fungal β-glucan and mannans by C-type lectins led to a decrease in the glycolytic activity of monocytes triggered after stimulation with heat-killed yeast or with live wild-type C . albicans . Of note , the different background of the strains used can contribute to the complexity of the interpretation of the results obtained in this work . Besides this , our results suggest that the impact of metabolism in the outcome of Candida infection is not strain-specific , as experiments with heat-killed stimuli and with live fungi were carried out with strains with different backgrounds ( UC820 in case of heat-killed fungi and SC5314 in case of live fungi ) . We validated in vivo results on the role of cellular metabolism for antifungal host defense in a mouse model of systemic C . albicans infection , in which glycolysis-mediated mechanisms were demonstrated to be crucial for the defense against the pathogen . Inhibition of glycolysis led to hosts that were significantly more susceptible to the infection , presenting lower fungicidal activity and defective cytokine production capacity . Our data also suggest that glutamine metabolism could play a role in the production of IL-1β and its known downstream target IL-6 [44] after C . albicans recognition both in human and mouse . The relationship between IL-1β and IL-6 could be also playing a role in the effects observed on Th1 and Th17-derived cytokines in human PBMC cultures , whose decrease could be due to direct effects of the pharmacological inhibitors employed or to indirect effects of the dampening of IL-1β and IL-6 released from monocytes . Our results reflect the concept that the inhibition of glucose metabolism during C . albicans infection has an impact on the immune system at different levels , as the impairment of glycolysis decreased the ability to fight candidiasis both by direct and indirect mechanisms , as previously described for NK cells in an in vitro study [45] . In this sense , treatment with 2-DG induced a decrease in the production of monocyte-derived cytokines proved crucial to boost Candida clearance by neutrophils such as IL-1β , IL-6 and TNFα[8 , 46 , 47] and also reduced the fungicidal potential of neutrophils directly treated with this metabolic inhibitor . These results suggest that the establishment of a functional response against systemic C . albicans infections in vivo largely relies on the proper functioning of glucose metabolism in different immune subsets , demonstrating the importance of glycolysis in the development of an efficient antifungal response and highlighting the central role of metabolism as a cornerstone of the immune function . The different pathways of the cellular metabolism are connected through a very complex network of enzymes and mediators . Our findings suggest that the immune functions of monocytes in C . albicans infection rely on the activation of glucose metabolism , but also required the participation of other additional metabolic pathways such as glutaminolysis or the pentose phosphate pathway . These results also delve into the distinctive features of CLR-mediated antifungal responses and highlight the need of studying the particular characteristics of the metabolic mechanisms underlying the immune responses against the wide variety of human pathogens . As combining effective anti-fungal treatment with adjuvant immunotherapy is proposed to improve the poor outcome in disseminated C . albicans infections , these data suggest that cellular metabolism of immune cells may represent a novel potential therapeutic target .
Buffy coats from healthy donors were obtained after written informed consent ( Sanquin Blood Bank , Nijmegen , the Netherlands ) . Samples were anonymized to safeguard donor privacy . The use of the samples received IRB approval . Peripheral blood mononuclear cell ( PBMC ) isolation was performed by dilution of blood in pyrogen-free PBS and differential density centrifugation over Ficoll-Paque ( GE Healthcare ) . Cells were washed three times in PBS . Percoll isolation of monocytes was performed as previously described ( Repnik et al . , 2003 ) . Briefly , 150–200 x 106 PBMCs were layered on top of a hyper-osmotic Percoll solution ( 48 . 5% Percoll [Sigma-Aldrich] , 41 . 5% sterile H2O , and 0 . 16 M filter-sterilized NaCl ) and centrifuged for 15 min at 580g . The interphase layer was isolated and cells were washed with cold PBS . Cells were re-suspended in RPMI culture medium ( RPMI medium Dutch modified , Invitrogen ) supplemented with 50 μg/mL gentamicin , 2 mM Glutamax , and 1 mM pyruvate , and counted . An extra purification step was added by adhering Percoll-isolated monocytes to polystyrene flat bottom plates ( Corning ) for 1 h at 37°C; a washing step with warm PBS was then performed to yield maximal purity . Candida albicans UC820 ( ATCC MYA-3573 ) [48] was grown overnight to generate yeast cells in Sabouraud dextrose broth at 29°C , with shaking . Cells were harvested by centrifugation , washed twice with PBS , and re-suspended in culture medium ( RPMI 1640 Dutch modification ) . To generate hyphae , yeast cells were inoculated and grown overnight at 37°C in culture medium adjusted to pH 6 . 4 with hydrochloric acid . C . albicans yeast or hyphae were heat-killed for 30 min at 95°C . The hyphae-specific G1 cyclin-null hgc1Δ mutant and hgc wild-type strains , kindly provided by Dr . Yue Wang ( Institute of Molecular and Cell Biology , Singapore [49] ) were grown under similar conditions . PBMCs at 5 x 106 cells/mL were stimulated for 4 or 24 h with RPMI or 105 heat-killed C . albicans yeast . Global gene expression was profiled using Illumina Human HT-12 Expression BeadChip according to manufacturer’s instructions . Image analysis , bead-level processing and quantile normalization of array data were performed using the Illumina LIMS platform , BeadStudio . Only samples that had cells for which paired fold-changes could be calculated were used ( i . e . for which cells from the same individual were used for both RPMI and Candida stimulation ) . At 4 h there were 19 matched samples , and at 24 h there were 29 matched samples . Log2 fold changes ( gene expression after Candida divided by expression after RPMI stimulation ) were calculated for each individual separately . The average of these values was used in the plot . Gene identifiers were mapped to entrez ids , and formatted to match the transcript gene identifiers as defined in “RECON1” , the model that contains the enzymes associated with each reaction . Gene expression values were always mapped to alternative transcript 1 , i . e . all entrez identifiers were appended with “_AT1” . The tool “Escher” [50] was used to generate a pathway to visualize the expression data containing the most interesting parts of it . 100 μL monocytes at 1 x 106 cells/mL or PBMCs at 5 x 106 cells/mL were added to flat-bottom or round-bottom 96-well plates ( Greiner ) , respectively . Cells were incubated with culture medium with 10% serum only as a negative control or incubated with 11 mM 2-deoxyglucose ( Sigma ) , 100 nM Torin1 ( Tocris ) , 1 mM AICAR ( 5-aminoimidazole-4-carboxamide-1-β-D-ribofuranoside , Brunschwig Chemie , Amsterdam , The Netherlands ) , 50 nM BPTES ( Sigma ) , 3 μM potassium dichloroacetate ( DCA , Sigma ) , 500 nM 6-aminonicotinamide ( 6-AN , Sigma ) , 10 μM etomoxir ( Sigma ) , 1 μM oligomycin ( Sigma ) . After 1 h , cells were stimulated with 105 heat-killed Candida yeast , 105 heat-killed Candida hyphae , 104 live hgc1 Candida yeast or 104 live Δhgc1 Candida yeast . The different concentrations of Candida used , were based on optimization experiments and employed as seen in works from different groups [51–53] . Similar concentrations of yeasts and hyphae were employed at similar concentrations in order to get a better approach to in vivo situations . Supernatants from monocytes were collected 24 h after stimulation . Supernatants from PBMCs were collected 7 days after stimulation , except for IL-10 production , which was measured after 48 h . Concentrations of inhibitors were selected as being the highest non-cytotoxic concentrations ( see S8 Fig ) . All supernatants were stored at -20°C until analyzed . For receptor blockade experiments , before stimulation with C . albicans , monocytes were preincubated for 1 h with 100 μg/mL laminarin ( Sigma ) , 10 μg/mL anti-CR3 antibody and control IgG ( R&D ) , 100 ng/mL B . quintana LPS , 10 μg/mL TLR2-blocking antibody ( anti-TLR2 ) and its control IgA1 ( InvivoGen , San Diego , CA ) , 5 μg/mL MR-blocking antibody ( anti-MR , R&D ) and IgG1 isotype control ( BD Biosciences ) . Cytokine production from human cells was determined in supernatants using commercial ELISA kits for IL-1β , TNFα , IL-17A , IL-22 ( R&D Systems , Minneapolis , MN ) IL-6 , IFNγ , and IL-10 ( Sanquin , Amsterdam , The Netherlands ) , following the instructions of the manufacturer . Lactate was measured from cell culture supernatants using a coupled enzymatic assay in which lactate was oxidized and the resulting H2O2 was coupled to the conversion of Amplex Red reagent to fluorescent resorufin by HRP ( horseradish peroxidase ) [54] . Glucose consumption was measured according to the manufacturer instructions using the Amplex Red Glucose/Glucose Oxidase Assay Kit ( Life Technologies ) . Glutamate , fumarate , malate , α-ketoglutarate and succinate concentrations were determined by commercial assay kits ( all Sigma ) following the instructions of the manufacturer from at least one million monocytes lysed in 1 mL 0 . 5% Triton-X in PBS at 4 and 24 h after stimulation . Real-time analysis of ECAR , OCR and SRC on monocytes was performed using an XF-96 Extracellular Flux Analyzer ( Seahorse Bioscience ) as described in Lachmandas et al . , 2016 . CD14+ monocytes were purified from freshly isolated PBMCs using MACS microbeads for positive selection , according to the manufacturer’s instructions ( Miltenyi Biotec ) . These monocytes were seeded in quintuplicate in XF-96 cell culture plates ( 2 × 105 monocytes/well ) in the presence of RPMI or C . albicans for 4h or 24 h in 10% human pooled serum . For the measurements of oxygen consumption and acidification rates it is therefore important to have a homogenous cell population , in this case monocytes . The CD14+ isolation is performed on PBMCs , which contain usually less than 5% of neutrophils . With subsequent CD14+ selection we obtain 95% of monocytes , so the amount of neutrophils and lymphocytes in the CD14+ selected cells is negligible . The metabolic rates of monocytes were analyzed in four consecutive measurements in XF Base Medium ( unbuffered DMEM with 5 . 5 mM glucose and 2 mM L-glutamine , pH adjusted to 7 . 4 ) . After three basal measurements , three consecutive measurements were taken following the addition of 1 . 5 μM oligomycin , 1 μM carbonyl cyanide-4- ( trifluoromethoxy ) phenylhydrazone ( FCCP ) , 2 μM antimycin together with 1 μM rotenone , glucose ( 20 mM ) , pyruvate ( 1 mM ) and/or 50 mM 2-DG in order to determine basal and maximum OCR and ECAR . SRC was determined as the absolute increase in OCR after FCCP injection compared with basal OCR . All compounds used during the Seahorse runs were acquired from Sigma-Aldrich . Cells were cultured as described above . After 4 h and 24 h mRNA was extracted by TRIzol ( Life Technologies ) , according to the manufacturer’s instructions , and cDNA was synthesized using iScript reverse transcriptase ( Invitrogen ) . Relative mRNA levels were determined using the Applied Biosciences StepOne PLUS and the SYBR Green method ( Invitrogen ) . Values are expressed as fold increases in mRNA levels , relative to those in non-stimulated cells , with HPRT as housekeeping gene . Primers are listed in S1 Table . Oxygen radical production levels of isolated monocytes were evaluated using luminol-enhanced chemiluminescence and determined in an automated LB96V Microlumat plus luminometer ( EG & G Berthold , Bald Wilberg , Germany ) as previously described [55] . Briefly , monocytes ( 1 × 105 per well ) were seeded into 96-well plates and incubated in medium containing either RPMI , phorbol 12-myristate 13-acetate ( PMA; 5 μg/ml ) , heat-killed C . albicans yeast or heat-killed C . albicans hyphae ( 107 CFU/ml ) . Luminol was added to each well in order to start the chemiluminescence reaction . Each measurement was carried out in at least duplicate repetitions . Chemiluminescence was determined every 145 s at 37°C for 1 h . Luminescence was expressed as relative light units ( RLU ) per second . Monocytes and PBMCs were isolated from blood donated by healthy volunteers after written informed consent . Ethical approval was obtained from the CMO Arnhem-Nijmegen ( NL32357 . 091 . 10 ) . Buffy coats from healthy donors were obtained after written informed consent ( Sanquin Blood Bank , Nijmegen , the Netherlands ) . Samples were anonymized to safeguard donor privacy . The use of the samples received IRB approval . All animal work was approved by the Animal Care and Use Committee of the Centro Nacional de Biotecnología—CSIC ( protocol number 312–2014 ) in accordance with Spanish RD 1201/2005 and international EU guidelines 2010/63/UE about protection of animals used for experimentation and other scientific purposes and Spanish national law 32/2007 about animal welfare in their exploitation , transport and sacrifice . 8–12 week-C57BL/6J mice were randomized and treated with a daily intraperitoneal dose of 100 mg/kg 2-deoxyglucose ( n = 6 ) or 100 μg BPTES ( n = 6 ) every morning for 5 days starting at the same day with the C . albicans intravenous infection . PBS was injected as a control ( n = 6 ) . C . albicans SC5314 yeast were grown on YPD plates ( Sigma-Aldrich , St Louis , MO ) at 30°C for 48 h . Then , C . albicans cells were centrifuged , washed in PBS and counted using a hematocytometer . Mice were infected by intravenous injection of 1 x 105 C . albicans yeast via the lateral tail vein and daily monitored for health and survival following the institutional guidance . After 5 days , mice were euthanized in a CO2 rodent euthanasia chamber and kidneys were aseptically removed , weighed and homogenized in PBS using a T10 basic Ultra-Turrax homogenizer ( Ika , Staufen , Germany ) . Fungal burden was determined by plating organ homogenates in serial dilutions on YPD plates . Colony forming units ( CFUs ) were counted after growth for 48 h at 30°C . For ex vivo stimulation experiments splenocytes were obtained from mice at day 5 of i . v . infection with C . albicans and stimulated ex vivo with LPS ( 10 ng/mL ) heat-killed Candida yeast ( 1 × 107/mL ) or heat-killed C . albicans hyphae ( 1 × 106/mL ) . Splenocytes were obtained by gently squeezing spleens in a sterile 100 mm filter . After centrifugation and washing , splenocytes were resuspended in complete RPMI 1640 medium supplemented with 10% FCS , 2 mM L-glutamine , 100 U/mL penicillin , 100 μg/mL streptomycin and 50 μM 2-mercaptoethanol , and counted using a hematocytometer . Splenocyte concentration was adjusted to 5 × 106/mL . 200 μL of the cell suspension were cultured in round-bottom 96-well plates ( Corning , Durham , NC ) and stimulated with RPMI or 1 × 106 heat-killed C . albicans yeast or hyphae/mL . The measurement of cytokine concentrations was performed in supernatants collected after 48 h of incubation at 37°C in 5% CO2 . Cytokine production from mouse cells was determined in supernatants using commercial ELISA kits for IL-1β , TNFα , IL-6 , IFNγ and IL-10 , all from BD Pharmingen ( San Diego , CA ) . Circulating neutrophils were isolated from blood drawn by cardiac puncture , diluted in PBS containing 5 mM EDTA and 3% FCS , overlaid over a density gradient of Histopaque 1119 and Histopaque 1077 ( Sigma ) and centrifuged for 30 minutes at 400 g . Neutrophil preparations had a purity >80% . To test neutrophil killing activity , 5 x 104 C . albicans yeast were exposed to 104 neutrophils for 2 h; neutrophils were then lysed with water and the number of surviving yeast cells was assessed on YPD agar . Killing activity was expressed as the percentage of C . albicans cells surviving in the presence of neutrophils compared to C . albicans cells surviving in the absence of neutrophils . Percoll monocytes were plated in 96 flat bottom plates at 1x105 cells / well . Cells were allowed to phagocytose 1 x 106 ( MOI 1:10 ) heat inactivated FITC-labeled C . albicans for 2h in the presence or absence of α-antitrypsin 10 mg/mL or 100 mg/mL . Subsequently , the fluorescence signal of extracellular non-phagocytosed Candida was quenched using trypan blue . The monocytes that phagocytosed one or more C . albicans yeast were enumerated by their positivity for the FITC signal , and could be divided into two populations: FITC- monocytes ( those that did not engulf C . albicans ) and FITC+ monocytes ( those that did ) . Cell viability was assessed using Annexin-V ( Biovision , San Francisco , CA ) and propidium iodide ( Sigma ) staining . Cells were stained for 15 minutes using Annexin-V-FITC using the protocol supplied by the manufacturer to detect early apoptotic cells . Subsequently cells were stained with for 5 minutes in 10 ug/mL propidium iodide . Cells were assessed for annexin-V and PI positivity using a FC500 flow cytometer ( Beckman Coulter ) . Annexin-V+ cells were considered as early apoptotic cells and Annexin-V+ PI+ cells were considered as late apoptotic cells .
|
Fungal infections are a major health concern for immunocompromised individuals due to the lack of success of the currently available antifungal therapies . Unveiling the metabolic processes involved in the immune function offers a promising opportunity for the development of new therapeutic approaches against these infections . In this report , we describe how changes in monocyte glucose metabolism are crucial for host defense against infections caused by the opportunistic pathogenic yeast Candida albicans . We report how the participation of various metabolic routes , such as glycolysis , oxidative phosphorylation and the pentose phosphate pathway , were differentially required after yeast or hyphal exposure , depending on the cellular energy requirements for each response . The proper control of metabolic reprogramming of immune cells was crucial to afford protection against fungal infections in vivo .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"cell",
"physiology",
"carbohydrate",
"metabolism",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"cytokines",
"yeast",
"infections",
"pathogens",
"metabolic",
"processes",
"immunology",
"microbiology",
"glucose",
"metabolism",
"cell",
"metabolism",
"glycolysis",
"developmental",
"biology",
"fungi",
"experimental",
"organism",
"systems",
"molecular",
"development",
"fungal",
"diseases",
"fungal",
"pathogens",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"mycology",
"white",
"blood",
"cells",
"animal",
"cells",
"medical",
"microbiology",
"microbial",
"pathogens",
"yeast",
"immune",
"system",
"biochemistry",
"candida",
"eukaryota",
"cell",
"biology",
"monocytes",
"physiology",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"cellular",
"types",
"metabolism",
"organisms",
"candida",
"albicans"
] |
2017
|
Rewiring monocyte glucose metabolism via C-type lectin signaling protects against disseminated candidiasis
|
The yellow fever vaccines ( YF-17D-204 and 17DD ) are considered to be among the safest vaccines and the presence of neutralizing antibodies is correlated with protection , although other immune effector mechanisms are known to be involved . T-cell responses are known to play an important role modulating antibody production and the killing of infected cells . However , little is known about the repertoire of T-cell responses elicited by the YF-17DD vaccine in humans . In this report , a library of 653 partially overlapping 15-mer peptides covering the envelope ( Env ) and nonstructural ( NS ) proteins 1 to 5 of the vaccine was utilized to perform a comprehensive analysis of the virus-specific CD4+ and CD8+ T-cell responses . The T-cell responses were screened ex-vivo by IFN-γ ELISPOT assays using blood samples from 220 YF-17DD vaccinees collected two months to four years after immunization . Each peptide was tested in 75 to 208 separate individuals of the cohort . The screening identified sixteen immunodominant antigens that elicited activation of circulating memory T-cells in 10% to 33% of the individuals . Biochemical in-vitro binding assays and immunogenetic and immunogenicity studies indicated that each of the sixteen immunogenic 15-mer peptides contained two or more partially overlapping epitopes that could bind with high affinity to molecules of different HLAs . The prevalence of the immunogenicity of a peptide in the cohort was correlated with the diversity of HLA-II alleles that they could bind . These findings suggest that overlapping of HLA binding motifs within a peptide enhances its T-cell immunogenicity and the prevalence of the response in the population . In summary , the results suggests that in addition to factors of the innate immunity , “promiscuous” T-cell antigens might contribute to the high efficacy of the yellow fever vaccines .
The yellow fever ( YF ) vaccines ( YF-17D-204 and 17DD ) are considered to be among the most effective vaccines [1] , [2] . Antibody and T-cell responses are believed to mediate protection [3] , [4] , [5] , and recent studies have also implicated the innate immune responses as one of the critical elements for developing the immune responses [6] . However , the immune adaptive mechanisms that make this vaccine so highly effective are unclear . T-cell immune responses against YF wild type virus and other flaviviruses , such as dengue and West Nile virus [7] , [8] , are considered to be important for development of neutralizing antibodies , and activation of CD4+ helper T-cells and CD8+ cytotoxic T lymphocytes ( CTLs ) against YF wild type virus has been reported [6] , [9] , [10] . The CTL responses appear 14 days after vaccination and these cells differentiate into long-lived memory T-cells after a few months [11]; however , only a few YF wild type virus T-cell epitopes have been described in humans [12] , [13] . In order to expand the repertoire of human leukocyte antigens ( HLA ) restricted YF wild type virus epitopes , and as part of the Immune Epitope Database - IEDB program ( http://www . immuneepitope . org/ ) , we studied the repertoire of T-cell responses present in a cohort of YF-17DD vaccinees established by Melo et al . [14] and identified 16 peptides that are immunogenic in more than 10% of the individuals tested . Analysis of the most prevalent immunogenic peptides indicated that they contain overlapping HLA binding motifs and suggested that the prevalence T-cell immunogenicity in response to YF vaccine ( 17DD ) is correlated with the ability of the peptide to bind multiple HLA types .
This study was reviewed and approved by the ethics committee of the Brazilian Ministry of Health ( CONEP: 12138; Process n° 25000 . 103608/2005-39 ) . In addition , the Johns Hopkins University and University of Pittsburgh Institutional Review Boards also reviewed and approved this study as protocol JHM-IRB-3: 03-08-27-01 and PRO09090146 respectively . Written informed consent was obtained from all volunteers and all clinical investigation was conducted according to the principles expressed in the Declaration of Helsinki . A cohort of YF-17DD vaccinees was established and described elsewhere [14] . Briefly , healthy subjects , IgG-negative for YFV , were immunized with the YF-17DD vaccine ( Biomanguinhos , Oswaldo Cruz Foundation , Brazil ) at The Brazilian National Health Surveillance Agency ( ANVISA ) office , International Airport of Recife . After a full explanation of the study , written consent was obtained from each volunteer and blood samples were collected before ( Day 0 , used as negative controls ) and after vaccination ( 2 months to 4 years ) . Blood samples were collected from 220 subjects into 10 mL Vacutainer® tubes ( Becton Dickinson , Franklin Lakes , NJ ) . Serum was obtained by centrifuging the whole blood at 1 , 600×g for 10 min and aliquots ( 1 mL per tube ) were stored at −20°C for further use on serological tests . For harvesting PBMC , blood samples collected in heparin Vacutainer® tubes ( Becton Dickinson , Franklin Lakes , NJ ) were used . Isolation of PBMC was carried out through gradient density using Ficoll-Paque™ PLUS ( Amersham Biosciences Ab , Uppsala , Sweden ) according to the protocol suggested by the manufacturer . Residual red blood cells were lyzed with Ammonium-Chloride-Potassium lysing buffer ( Gibco BRL , Gaithersburg , MD ) for 3 min at room temperature . Then , cells were resuspended in Hybridoma Serum Free Media ( SFM; Gibco BRL , Gaithersburg , MD ) for ELISPOT assaying . PBMC samples from at least 3 subjects , known to respond against the peptides Env57–71 , Env345–359 , Env361–375 , NS3137–151 , NS4b77–91 , NS5341–355 , NS5345–359 , NS5465–479 and NS5481–495 were depleted of CD4 T-cells using anti-CD4 mAb coated microbeads ( Miltenyi Biotec , Auburn , CA ) and LD columns ( Miltenyi Biotec , Auburn , CA ) according to manufacturer's directions . This approach consistently yielded cell samples in which depletion was greater than 95% for CD4+ T-cells , according to flow cytometry analysis ( data not shown ) . CD8+ T-cells were isolated from PBMC samples by the negative selection method using the isolation kits and LS columns provided by Miltenyi Biotec ( Auburn , CA ) , according to the manufacturer's manual . This approach consistently yielded cell samples in which purity was greater than 90% for CD8+ T-cells according to flow cytometer analysis ( data not shown ) . To measure the breadth and magnitude of T-cell responses , a library of 653 peptides ( 15-mers ) , overlapping by 11 amino acids , spanning the length of the 17DD YFV proteins Env ( n = 120 ) , NS1 ( n = 47 ) , NS2a ( n = 32 ) , NS2b ( n = 28 ) , NS3 ( n = 148 ) , NS4a ( n = 45 ) , NS4b ( n = 17 ) and NS5 ( n = 216 ) ( sequence NCBI entry U17066 ) was synthesized by Schafer-N ( Copenhagen , Denmark ) . These peptides were HPLC-purified to 80% purity or greater , with the exception of four peptides that could not be purified and were used as crude products . The Env peptides were pooled in groups of 6 peptides , totalizing 20 pools , Box S1 . NS peptides were arranged into matrix pools [NS1 ( 7×7 ) ; NS2a ( 8×4 ) ; NS2b ( 7×4 ) ; NS3 ( 15×10 ) ; NS4a ( 8×5 ) ; NS4b ( 4×4 ) and NS5 ( 17×13 ) ] ensuring that each peptide was present in two different pools , Box S2 [15] . Ten NS peptides were tested individually ( no pooling ) . The peptides NS4b77–85 , NS4b76–84 , NS4b75–83 , NS4b75–84 were synthesized and purified by Genescript ( New Jersey , USA ) . The purity of these peptides was greater than 95% . Stock solutions of all peptides were sterile-prepared at 2 mg/mL in 10% ( v/v ) dimethil sulphoxide ( DMSO; Sigma-Aldrich ) and stored at −20°C until use . Transporter associated with antigen processing-deficient T2 cells expressing HLA-A*0201 were maintained in RPMI 1640 medium ( Invitrogen , Brown Dee , WI ) supplemented with 10% ( v/v ) , fetal bovine serum ( FBS; Hyclone , Logan , UT ) , 1% ( v/v ) glutamine ( Gibco , Gaithersburg , MD ) and 1% ( v/v ) penicillin-streptomycin ( Gibco , Gaithersburg , MD ) . For pulsing the peptides , one million target cells were washed , resuspended in 1 mL of media and incubated at 37°C for 2 h with peptides at 10 µg/mL . Cells were then washed 3 times and resuspended in SFM at 1×106 cells/mL for T-cell function assays . Interferon-γ ELISPOT assays were performed by using the ELISPOT set from BD-Biosciences Pharmingen ( San Diego , CA ) , according to the manufacturer′s protocol . Briefly , 96-well nitrocellulose-bottom plates were coated with 100 µL/well primary antibody anti-human IFN-γ at 5 µg/mL and incubated at 4°C overnight . The following day , plates were blocked with 200 µL/well of RPMI 1640 ( Invitrogen , Brown Dee , WI ) containing 10% ( v/v ) heat-inactivated FBS ( Hyclone , Logan , UT ) , 1% ( v/v ) glutamine ( Gibco , Gaithersburg , MD ) and 1% ( v/v ) penicillin-streptomycin ( Gibco , Gaithersburg , MD ) for 2 h at room temperature . Then , the cells were plated , in duplicate , at a range of 1–3×105 cells/well along with peptides at 10 µg/mL ( concentration of each peptide when tested as a pool and also the concentration of individual peptides when tested individually ) both in SFM . During peptide pool analysis and deconvolution , total PBMC were used as source of T-cells . For the experiments aimed to confirm whether CD8+ T-cells were driving the cellular responses , in addition to total PBMC , CD4-depleted PBMC samples were included in the analysis . Media ( SFM ) containing 1% ( v/v ) glutamine , 1% ( v/v ) penicillin-streptomycin alone ( background ) and media supplemented with phorbol 12-myristate 13-acetate ( PMA; Sigma ) at 250 ng/mL and ionomycin ( Sigma ) at 250 ng/mL were used as negative and positive controls , respectively . After 16 h incubation at 37°C , 5% CO2 , the plates were washed twice with distilled water and 4 times with PBS containing 0 . 05% Tween-20 ( PBS-T; Sigma ) followed by incubation for 2 h , at room temperature , with 4 µg/mL of biotinylated anti-human IFN-γ . Plates were again washed four times with PBS-T and incubated for 1 h , at room temperature , with avidin-horseradish peroxidase at 10 µg/mL . After another washing cycle ( 5× ) using PBS-T followed by three washes with PBS , plates were developed with substrate 3-amino-9-ethyl carbazole ( AEC; BD-Biosciences Pharmingen , San Diego , CA ) for 40 min . The reaction was stopped with distilled water , the plates were air-dried and spots were counted using the ImmunoSPOT reader using ImmunoSPOT software version 3 . 2 ( Cellular Technology Ltd . ) . Totals for duplicated wells were averaged and normalized to numbers of IFN-γ spot forming cells ( SFC ) per 1×106 PBMCs . In experiments aimed at confirm the HLA-A*0201 restriction of T-cell epitopes identified , ELISPOT was performed using T2 cell lines as target to activate CD8+ T-cells . T2 cells were pulsed with each of the NS4b peptides ( NS4b77–85 , NS4b76–84 , NS4b75–83 and NS4b75–84 ) individually . Peptide-pulsed target cells and effector cells ( CD8 T-cells ) were co-cultured at 1×105 cells/well each and incubated for 16 h at 37°C , 5% CO2 on ELISPOT plate . After incubation , the cells were removed and IFN-γ was detected as described above . The negative controls were effector cells only and effector cells plated with unpulsed target cells , whereas the positive control was PMA/ionomycin as mentioned previously . The criteria to identify positive peptide pools/individual peptides relied on the combination of the following equations: ( i ) mean number of spots ( peptides ) −2 standard deviations ( SD ) >mean number of spots ( background ) ; ( ii ) mean number of spots ( peptides ) >mean number of spots ( background ) +2 SD; ( iii ) mean number of spots ( peptides ) - mean number of spots ( background ) >10 . The peptide pools that met all criteria listed above were directly selected . For the NS proteins , T-cell responses among volunteers analyzed during pool analysis was low and , thus , difficult to comply with the analysis criteria shown above; therefore , an additional test was applied in order to increase the number of pools to deconvolute . This additional test consisted of the calculation of a cut-off based on average ( Avg ) SFC of the top 25% peptide pool and standard error through the following equation: Avg SFC Top 25% peptide pool +2 SE . The peptide pool in which SFC value was greater than the cut-off was also considered positive . The combinations in which the frequency was greater than 7% ( for NS1 and NS3 peptides ) ; 10% ( for NS2a , NS2b , NS4a and NS4b ) ; and 16% to NS5 were taken into account , and their respective common peptide was selected to be tested individually . Genomic DNA was extracted from PBMCs collected from 142 volunteers by using the PureLink Genomic DNA MiniKit ( Invitrogen , Carlsbad , CA ) following the manufacturer protocol . HLA alleles genotyping was performed using the polymerase chain reaction ( PCR ) test sequence-specific primer ( SSP ) UniTray Kit ( Invitrogen , Brown Dee , WI ) , which provides low to intermediate HLA typing results according to the manufacturer protocol . After amplification , the PCR products were separated in 2% ( w/v ) agarose gel pre-stained with ethidium bromide ( 0 . 25 µg/mL gel ) . Gels were electrophoresed for 30 min at 150 volts in 0 . 5× Tris-Boric acid-EDTA ( TBE ) buffer , then examined under Ultraviolet illumination and documented by photography . The types of HLA-A , B , C were then determined by specific electrophoresis bands using the UniMatch Software ( Invitrogen , Brown Dee , WI ) . Further analysis was performed to determine the HLA allele in high-resolution for the individuals HLA-A*02 positives using the same methodology above mentioned . The peptides NS4b77–85 , NS4b76–84 , NS4b75–83 and NS4b75–84 were incubated individually with HLA-A*0201/anti-hCD28 aAPCs [16] , [17] , [18] at concentration of 50×106 beads/mL for 5 days at 4°C . Then , one million CD8+ T-cells isolated from HLA-A*0201 positive individuals were co-cultured with 1×106 of each peptide-loaded aAPCs , individually , in RPMI 1640 media supplemented with 8% ( v/v ) T-cell growth factor ( TCGF ) and 5% ( v/v ) human AB serum . On day 4 , the plates were replenished with media containing TCGF and AB human serum and cultured for 3 more days . On day 7 , the aAPCs were removed and the cells were washed , counted and their concentration adjusted to 1×106 cells/mL in RPMI 1640 media ( Invitrogen , Brown Dee , WI ) supplemented with 10% ( v/v ) FBS ( Hyclone , Logan , UT ) . CD8+ T-cells ( 1×105 ) that underwent one round of expansion with different aAPCs were “challenged” with T2 cells ( 1×105 ) pulsed , individually , with each NS4b peptides for 2 h at 37C , 5% CO2 . Then , Golgi stop ( BD Biosciences , San Diego , CA ) was added to the culture to stop protein trafficking and secretion . The cells were cultured for 4 h at 37C , 5% CO2 and then washed twice with staining buffer [PBS containing 0 . 05% sodium azide and 2% ( v/v ) FBS ( Hyclone , Logan , UT ) ] . The cells were stained with anti-CD8 FITC ( Sigma Aldrich , St . Louis , MO ) for 20 min at 4°C and washed 3 times with staining buffer . After that , the cells were permeabilized with cytoperm/cytofix buffer ( BD Biosciences , San Diego , CA ) for 20 min at 4°C followed by 3 washes with perm/wash buffer ( BD Biosciences , San Diego , CA ) . Then staining with either anti-hCD107a-PE ( BD Biosciences , San Diego , CA ) or anti-hIFN-γ-PE ( BD Biosciences , San Diego , CA ) was performed for 30 min at 4°C , after which the cells were washed twice with perm/wash buffer and once with staining buffer . Finally , the cells were acquired on a FACS Calibur flow cytometer ( BD Bioscience , San Diego , CA ) and the data was analyzed using Flowjo for Macintosh version 8 . 8 . 6 ( Tree star , Ashland , OR ) . Purification of HLA class II MHC molecules by affinity chromatography , and the performance of assays based on the inhibition of binding of a high affinity radiolabeled peptide to quantitatively measure peptide binding , were performed essentially as detailed elsewhere [19] , [20] , [21] , [22] , [23] . Briefly , EBV transformed homozygous cell lines were used as sources of MHC molecules . A high affinity radiolabeled peptide ( 0 . 1–1 nM ) was co-incubated at room temperature or 37C with purified MHC in the presence of a cocktail of protease inhibitors . Following a two-day incubation , MHC bound radioactivity was determined by capturing MHC/peptide complexes on Ab coated Lumitrac 600 plates ( Greiner Bio-one , Frickenhausen , Germany ) , and measuring bound cpm using the TopCount ( Packard Instrument Co . , Meriden , CT ) microscintillation counter . The concentration of peptide yielding 50% inhibition of the binding of the radiolabeled peptide was calculated . Under the conditions utilized , where [label]<[MHC] and IC50≥[MHC] , the measured IC50 values are reasonable approximations of the true Kd values . Each competitor peptide was tested at six different concentrations covering a 100 , 000-fold range , and in three or more independent experiments . As a positive control , the unlabeled version of the radiolabeled probe was also tested in each experiment . Fisher's two-tailed test was used to associate the response against a given peptide and HLA genotype . When significant p-values were achieved ( p<0 . 05 ) , the odds ratio and 95% confidence intervals ( CI ) were calculated ( R Statistical Package 2 . 9 . 0 ) . HLA-A , B and C genotype distributions were checked with Hardy-Weinberg equilibrium . In addition , we compared the allelic distributions of the five genotyped HLA genes ( HLA-A , HLA-B , HLA-C , HLA-DR , HLA-DQ ) to HLA genotype data obtained from the MHC Database ( dbMHC ) of the NCBI , consisting of two studies with Brazilian populations: one with volunteers from the state of Minas Gerais ( which we refer to as “dbMHC1” ) , and another with volunteers from the state of São Paulo ( referred to as “dbMHC2” ) . Pearson's correlation was used to assess the co-variation between HLA promiscuity and frequency of recognition of the yellow fever T-cell epitopes identified .
The immunogenicity of the YF-17DD peptides was determined by IFN-γ ELISPOT performed ex vivo using PBMCs from immunized volunteers . A flowchart summarizing the assays and strategies for characterization of the T-cell responses are show in Figure 1 . Blood samples were negative for anti-YF-17DD antibodies prior to vaccine inoculation and all vaccinees seroconverted by one month after immunization [14] . The peptides were organized in groups corresponding to the Env , NS1 , NS2 , NS3 , NS4 and NS5 proteins and screened in two rounds . In the first round the peptides were organized in pools or matrices and tested in samples from a series of donors as indicated in Table 1 . In the second round , immunogenic peptides were selected from the pools screened individually as indicated in Table 2 . Screening of immunogenic T-cell peptides within the structural proteins: The initial screening was performed with 120 peptides from the Env protein . The peptides were organized in 20 pools with six peptides in each and tested in samples from 146 volunteers . Five pools ( 3 , 9 , 14 , 15 and 20 ) were immunogenic in 13 to 34 of the 146 individuals tested . Peptide pool number 15 was immunogenic in 34 of the 146 individuals tested and was most prevalent immunogenic pool in this population ( data not shown ) . Thirty individual peptides from these five pools were selected for a second round of analysis and tested individually ( Table 2 ) . Thirteen peptides were positive in the second round ( Table S1 ) . Six ( Env57–71 , Env65–79 , Env73–87 , Env337–351 , Env345–359 and Env361–375 ) out of the 30 Env peptides tested individually were positive in at least 10% the volunteers . Their protein position , amino acid sequence , frequency with in the cohort and magnitude of the ex vivo T-cell response are shown on Table 3 . Screening of immunogenic T-cell peptides within Non-structural ( NS ) proteins . A similar strategy was applied to screen the T-cell responses against NS peptides . In the first round , the pools were organized in matrices with each peptide being present in two pools to allow identification of the immunogenic peptide within each pool . The matrices were assembled as described in the methods section and tested in a series of samples from different volunteers ( Table 1 ) . The peptide pool matrix analysis , in many cases , allowed the precise identification of the immunogenic peptide within each pool . The peptides that most frequently activated T-cell responses in the individuals tested are shown in Table S1 . These peptides were selected for the second round of screening as individual peptides in a second set of samples . The summary of the results of the second round is shown in Table 2 . In total , 7 ( NS1 ) , 3 ( NS2a ) , 6 ( NS2b ) , 18 ( NS3 ) , 11 ( NS4a ) , 2 ( NS4b ) and 13 ( NS5 ) peptides were shown to be immunogenic on 1 to 11 vaccinees tested in the second round ( Table S1 ) . Ten peptides among NS proteins ( NS2b97–111 , NS2b113–127 , NS3137–151 , NS4a197–211 , NS4b77–91 , NS5341–355 , NS5345–359 , NS5465–479 , NS5469–483 , NS5481–495 ) were immunogenic in at least 10% of the YF-17DD vaccinees . Their position , sequence , frequency and magnitude of T-cell response are shown on Table 3 . In order to investigate the HLA restriction related to the T-cell responses to each of the peptides shown on Table 3 , HLA genotyping was performed for the loci HLA-A , HLA-B and HLA-C on 142 YF-17DD vaccinees ( Table S2 ) . The frequency of the HLA types present in our cohort was compared to the frequencies reported by two Brazilian blood banks ( Minas Gerais - “dbMHC1” and São Paulo - “dbMHC2” ) and deposited at the cohort dbMHC database . Statistical analysis indicated a large degree of correlation between the HLA diversity present in the study volunteers and the ones on the dbMHC database , for all available genes ( p<0 . 01 ) , suggesting that the HLA diversity of the cohort is representative of the general Brazilian population . The HLA types of the samples used on the second round of screening and the respective peptide with which they reacted are shown in Table S3 . Subsequently , we investigated the possibility of some HLA types to be overrepresented among the individuals responding to a peptide . Indeed , some HLA types were very prevalent among the responders of some peptides . For example , HLA-A*23 was very frequent among the individuals responding to the NS5481–495 peptide while it was seldom observed among individuals that did not react to this peptide suggesting the possibility that this HLA type might be involved in the presentation of an epitope present within the immunogenic 15-mer . In order to confirm these associations , Fisher's test was carried out comparing the frequency of the HLA types present in the vaccinees responding to a given peptides versus the HLA frequency in the general population or among the ones that did not respond . The HLA alleles with significant statistical associations are shown in Table 4 . Nine out of the 16 immunogenic peptides could be statistically associated with one or more HLA type . Seven peptides presented significant associations with one HLA type , four HLA-A ( A*02 , A*11 , A*23 , A*26 ) , three HLA-B ( B*15 , B*18 , B*39 ) and one HLA-C ( C*12 ) . Two peptides , Env361–375 and NS5341–355 showed significant association with two HLA-types ( A*26 and B*18 ) and ( B*39 and C*12 ) respectively . Three immunogenic peptides Env57–71 , Env345–359 and NS4b77–91 , were associated with HLA-A*02 and other two NS5341–355 and NS5345–359 were associated with B*39 . However these two peptides associated with B*39 overlap by 11 amino acids suggesting that these associations are likely directed to an immunogenic determinants shared by these two adjacent peptides . The peptides Env57–71 , Env345–359 and NS4b77–91 induced immune responses predominantly in HLA-A*02 individuals ( Tables 4 and S3 ) . In addition , the statistical analysis indicated that the peptides Env361–375 , NS3137–151 , NS5341–355 , NS5345–359 , NS5465–479 and NS5481–495 were associated with the presence of the other HLA class I alleles . In order to validate these associations , we analyzed whether CD8+ T-cells derived from the immunized individuals could be activated by these peptides . For these analysis , PBMCs from 3 volunteers were collected , the CD4+ T-cell depleted , and the T-cell responses to each peptide measured by ELISPOT . It is important to note that some antigen presenting cells ( APC ) were unintentionally removed , because CD4 is also expressed on monocytes and dendritic cells , albeit at lower levels than on CD4+ T helper cells . Seven of the nine peptides selected could activate CD4+ depleted PBMCs from YF-17DD vaccinees ex vivo , whereas PBMCs collected pre-vaccination were negative , corroborating the hypothesis that these immune responses are likely being mediated by CD8+ T-cells ( Table 5 ) . In order to further determine the HLA-restriction involved on the T cell response against the peptide NS4b77–91 , previously shown to be associated with HLA-A*02 , all the subjects bearing this HLA were genotyped at high resolution for this locus . . According to the analysis , 78% of the subjects responding against the peptide NS4b77–91 were HLA-A*02:01 ( Table S3 ) . Subsequently , all the 9-mers covering the NS4b77–91 sequence were synthesized and tested for T-cell activation on PBMC from HLA-A*0201 subjects ( n = 3 per peptide ) . For this experiment , the initial 15-mer was used as positive control . Among the 9-mers derived from the NS4b77–91 , the NS4b77–85 ( LWNGPMAVS ) was the one that induced the most T-cell activation ( NS4b77–91 , used as positive control in Figure 2C ) . Interestingly , Akondy et al . [24] identified an immunodominant epitope HLA-A*02-restricted on NS4b protein of YF-17D vaccine , the position of which was at amino acid 76–84 ( LLWNGPMAV ) , one amino acid to N-terminal side ( Figure 2A ) . This observation highlights one important caveat of screening peptide libraries , and points at the presence of epitopes not detected due to the manner in which the 15-mer peptides of the library were spliced . Therefore , we tested the adjacent peptide at the position 75–91 for the presence of potential HLA-A*02:01-restricted epitopes within its sequence . Since these peptides have a common core ( LWNGPMA ) , they would represent possible products of antigen processing that could activate the same T cell clone . Thus , to investigate that these peptides could be presented by HLA-A*02:01 , CD8+ T-cells were isolated from HLA-A*02:01-positive individuals and the ELISPOT assay for IFN-γ was performed using T2 cells , which expresses HLA-A*02:01 exclusively , as target for T-cell activation . Figure 2B depicts representative ELISPOT data set . All peptides analyzed could activate CD8+ T-cells , but to different degrees . The peptides with the core LLWNGPMAV ( NS4b76–84 and NS4b75–84 ) induced the highest number of spots , whereas the peptides in which the core was incomplete ( NS4b77–85 and NS4b75–83 ) the T-cell response tended to be lower . Finally , we tested the ability of these peptides to expand the population of CD8+ T-cells of the patients for one week using aAPC loaded with each of the NS4b peptides . One million purified CD8 T-cells were co-cultured with each aAPC ( ratio 1∶1 ) and after one week the yield of cellular proliferation of the CD8+ T-cell population expanded by the aAPCs loaded with NS4b77–85 , NS4b76–84 , NS4b75–83 or NS4b75–84 peptides were 2×106 , 4×106 , 2 . 5×106 and 3 . 5×106 respectively , indicating a proliferation of 2 to 4 fold . Flow cytometry analysis showed that after one week of expansion , epitope-specific CD8+ T cells were activated , as degranulating cells ( surface CD107a; data not shown ) producing IFN-γ ( Figure 2C ) were identified upon challenging with the same epitopes used during cell expansion . Notably , the expansion of NS4b76–84-specific CD8+ T-cells was the highest among the peptides tested and reached 22% . Interestingly , the peptide NS4b75–84 , which on unexpanded CD8+ T-cells tested ex vivo elicited comparable levels of T-cell response as NS4b76–84 , had , after one week expansion with the aAPC , a frequency of NS4b75–84 specific-CD8+ T-cells that was 4-fold less as compared to NS4b76–84 . These results suggest that there might be different products of antigen processing that could potentially be activating the same T cell clone , however at different degrees . The physiological role of these products of antigen processing for an effective T cell response needs to be further investigated . CD4+ and CD8+ responses cooperate with each other , enhancing and also regulating T-cell responses . CD4+ T-cell helper cytokine responses are required for a proper activation of naïve CD8+ T-cells [25] and CD4+ helper epitopes linked to each other can also cooperate to allow more efficient T-cell priming of weaker epitopes [26] , [27] . Thus , it is reasonable to expect that the most immunogenic peptides could contain both HLA-I and HLA-II binding motifs . Therefore to address this , the binding affinity of the immunogenic peptides was determined for 14 different HLA class II molecules ( DRB1*0101 , 0301 , 0401 , 0404 , 0405 , 0701 , 0802 , 0901 , 1101 , 1302 , 1501 , DRB3*0101 , DRB4*0101 , DRB5*0101 ) and the results are shown in Table 6 . All the peptides tested could bind with high affinity ( IC50≤1000 nM , see [28] ) to at least one HLA-DR molecule . Levels of peptide promiscuity varied from low ( affinity to 2 to 5 HLA class II molecules bound ) , e . g . NS5341–355 and NS5345–359; to intermediate ( affinity to 6 to 9 HLA class II molecules bound ) , e . g . Env337–351 , NS2b97–111 and NS5465–479; and highly promiscuous ( binding to ≥10 HLA molecules ) , e . g . Env57–71 , Env345–359 , NS4b77–91 , NS5469–483 , and NS5481–495 . Nine peptides ( Env57–71 , Env345–359 , Env361–375 , NS3137–151 , NS4b77–91 , NS5341–355 , NS5345–359 , NS5465–479 and NS5481–495 ) contained overlapping HLA-I an HLA-II binding motifs ( Tables 3 and 6 ) . This suggested that our screening process targeted the identification of promiscuous T-cell antigens . We then tested whether the frequency of recognition of the peptides on the cohort was dependent on the level of promiscuity of those to different HLA-II molecules . There was a significant correlation ( R2 = 0 . 45 , p = 0 . 01 ) between the number of HLA class II molecules bound to the peptides ( Figure 3 ) and the frequency of positivity in the vaccinees , suggesting that the prevalence of the immunogenicity of an antigen in the population is associated with the HLA promiscuity of the T-cell antigen .
This study presents the screening of 653 peptides of the YF-17DD Env , NS1 , NS2 , NS3 , NS4 and NS5 proteins in the context of a cohort of healthy adults immunized with the YF-17DD vaccine . Considering the two rounds of peptide screening , each peptide was tested in at least 75 individuals , and in the case of the Env protein each peptide was tested in 208 volunteers . The screening allowed the identification of 16 T-cell peptides that were immunogenic in 10% or more of the individuals tested . Only a few YF-17D epitopes have been characterized previously [12] , [13] , [24] . Overall , among the 16 immunogenic peptides identified herein , 14 contain new human T-cell antigens ( Env57–71 , Env65–79 , Env73–87 , Env337–351 , Env345–359 , Env361–375 , NS2b97–111 , NS3137–151 , NS4a197–211 , NS5341–355 , NS5345–359 , NS5465–479 , NS5469–483 , NS5481–495 ) . Two peptides , NS2b113–127 [12] and the NS4b77–91 [24] , contained epitopes previously described in humans , while the Env57–71 and NS2b113–127 peptides contain epitopes that have been described in the murine H2d background . The murine epitope ( Env57–71 ) was reported to be able to stimulate both CD8+ and CD4+ T-cells to secrete IFN-γ in YF-17DD vaccine immunized H2 d mice and also HLA A02 , B07 and A24 transgenic mice [29] , [30] . These highly prevalent immunogenic peptides were shown to contain multiple HLA binding motifs and that the degree of prevalence of its immunogenicity was correlated with the HLA promiscuity . Previous studies have shown some degree of correlation between predicted binding affinity and immunogenicity [30] . However , additional studies are required to determine the precise breath and differences in functionalities of these immunogenic peptides in different HLA contexts . Biochemical binding assays indicated that the most prevalent immunogenic peptides could bind multiple HLA molecules and the prevalence of their immunogenicity was correlated with the presence of multiple HLA binding motifs . Interestingly , the level of promiscuity of the class II binding was correlated with the prevalence of the immunogenicity of the peptide in the cohort ( Figure 2 ) . No promiscuous T-cell immunogens ( peptides that bind more than one HLA allelic variant ) have been described for YF wild type virus or other flaviviruses until now . Computational strategies for determination of promiscuous HLA-I and II binding peptides [28] , [31] have been used for cancer [32] and infectious disease [33] . Besides the theoretical advantage of being broadly reactive in the population , the biological characteristics of promiscuous T-cell epitopes are not clear and only few promiscuous epitopes have been identified . CD4+ T-cells play a major role in the generation of CD8+ cytotoxic T-cell responses and maturation of neutralizing antibodies [34] , [35] , [36] . In addition , virus-specific CD4+ T-cells may be able to tolerate more sequence diversity in their target epitopes than CD8+ T-cells , thus being more resistant to mutational escape [37] . Most of the immunogenic peptides , subsequently assessed using competitive binding assays bound to at least six HLA-DR alleles , indicating that an individual bearing at least one such HLA-DR molecule could develop broad CD4+ T-cell responses against the YF-17DD . Previous work has shown that many peptides are capable of binding with good affinity to multiple DR alleles [19] , [28] , [38] , [39] , [40] . However , it is still not clear how a given promiscuous peptide binds different HLA class II molecules . For example , Kilgus et al . [41] showed that distinct sites on the malaria T-cell epitope interact in different ways with the three DR molecules analyzed [41] . On the other hand , Panina-Bordignon et al . [42] showed that promiscuous peptides interact in a similar way with different DR molecules , possibly by binding to the conserved DR residues [42] . Chicz et al . [43] suggested that the ability of peptides to bind multiple HLA alleles must be dependent on the composition and location of several key amino acids within the primary structure , which led to the hypothesis that rigid allele-specific motifs for the class II molecule do not exist , thus permitting a broad binding specificity [43] . To our knowledge , the present study is the first report showing that the frequency of recognition of the peptides is dependent on the level of promiscuity of those to different HLA molecules . The exact role of the promiscuous peptides in protective immune response still unknown , however one would expect that a vaccine built with multiple immunodominant promiscuous epitopes , capable to bind several HLA molecules , could lead to an increased coverage of the human population . Although the cellular and humoral responses play a central role in effectiveness of YF vaccines , the innate immunity , which is known to shape the development of adaptive immune responses , also contribute to vaccine-mediated protection through other mechanisms , including its live replicative nature and its ability to trigger several Toll-like receptors [6] . In conclusion , the identification of this set of highly prevalent class I and II T-cell epitopes will enable detailed studies of the role of T-cell responses on the development of yellow fever immunity and the identification of the structural requirements of immunogenic promiscuous T-cell epitopes .
|
T-cell responses are considered to be very important; however , the role of T-cell responses in vaccine mediated immunity is still controversial . One reason may be that most studies of human T-cell responses are focused on a few epitopes . We still lack a systematic view of the repertoire of peptides presented by the different HLA class I and II molecules and how the peptides presented by the different HLAs interact within the host to develop T-cell responses . Here we present a study of the T-cell responses against the YF-17DD vaccine in the context of a cohort of 220 volunteers and observed that the most prevalent T-cell responses are targeted at peptides that bind to multiple types of HLA molecules . Based on these results we postulate that promiscuous T-cell epitopes might have a critical role in the development of adaptive immunity . These results may have broader implications for other pathogens , since the yellow fever vaccine is currently being developed as a vaccine vector for other diseases . Therefore , these epitopes might have a functionally cooperative role in boosting specific neutralizing antibody responses . In addition , we propose that promiscuous T-cell antigens may be better immunogens for vaccine development; however more studies are necessary .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"immune",
"cells",
"clinical",
"research",
"design",
"jungle",
"yellow",
"fever",
"flavivirus",
"immunity",
"to",
"infections",
"dengue",
"immunology",
"vaccines",
"adaptive",
"immunity",
"cohort",
"studies",
"neglected",
"tropical",
"diseases",
"immunizations",
"vaccination",
"viral",
"hemorrhagic",
"fevers",
"infectious",
"diseases",
"major",
"histocompatibility",
"complex",
"dengue",
"fever",
"t",
"cells",
"arboviral",
"infections",
"immune",
"response",
"yellow",
"fever",
"west",
"nile",
"fever",
"antigen",
"processing",
"and",
"recognition",
"clinical",
"immunology",
"immunity",
"tropical",
"diseases",
"(non-neglected)",
"vaccine",
"development",
"prospective",
"studies",
"viral",
"diseases"
] |
2013
|
T-Cell Memory Responses Elicited by Yellow Fever Vaccine are Targeted to Overlapping Epitopes Containing Multiple HLA-I and -II Binding Motifs
|
Anatomic connections between brain areas affect information flow between neuronal circuits and the synchronization of neuronal activity . However , such structural connectivity does not coincide with effective connectivity ( or , more precisely , causal connectivity ) , related to the elusive question “Which areas cause the present activity of which others ? ” . Effective connectivity is directed and depends flexibly on contexts and tasks . Here we show that dynamic effective connectivity can emerge from transitions in the collective organization of coherent neural activity . Integrating simulation and semi-analytic approaches , we study mesoscale network motifs of interacting cortical areas , modeled as large random networks of spiking neurons or as simple rate units . Through a causal analysis of time-series of model neural activity , we show that different dynamical states generated by a same structural connectivity motif correspond to distinct effective connectivity motifs . Such effective motifs can display a dominant directionality , due to spontaneous symmetry breaking and effective entrainment between local brain rhythms , although all connections in the considered structural motifs are reciprocal . We show then that transitions between effective connectivity configurations ( like , for instance , reversal in the direction of inter-areal interactions ) can be triggered reliably by brief perturbation inputs , properly timed with respect to an ongoing local oscillation , without the need for plastic synaptic changes . Finally , we analyze how the information encoded in spiking patterns of a local neuronal population is propagated across a fixed structural connectivity motif , demonstrating that changes in the active effective connectivity regulate both the efficiency and the directionality of information transfer . Previous studies stressed the role played by coherent oscillations in establishing efficient communication between distant areas . Going beyond these early proposals , we advance here that dynamic interactions between brain rhythms provide as well the basis for the self-organized control of this “communication-through-coherence” , making thus possible a fast “on-demand” reconfiguration of global information routing modalities .
In Arcimboldo's ( 1527–1593 ) paintings , whimsical portraits emerge out of arrangements of flowers and vegetables . Only directing attention to details , the illusion of seeing a face is suppressed ( Figure 1A–B ) . Our brain is indeed hardwired to detect facial features and a complex network of brain areas is devoted to face perception [1] . The capacity to detect faces in an Arcimboldo canvas may be lost when lesions impair the connectivity between these areas [2] . It is not conceivable , however , that , in a healthy subject , shifts between alternate perceptions are obtained by actual “plugging and unplugging” of synapses , as in a manual telephone switchboard . Brain functions –from vision [3] or motor preparation [4] up to memory [5] , attention [6]–[8] or awareness [9]– as well as their complex coordination [10] require the control of inter-areal interactions on time-scales faster than synaptic changes [11] , [12] . In particular , strength and direction of causal influences between areas , described by the so-called effective connectivity [13]–[15] , must be reconfigurable even when the underlying structural ( i . e . anatomic ) connectivity is fixed . The ability to quickly reshape effective connectivity –interpreted , in the context of the present study , as “causal connectivity” [16] or “directed functional connectivity” ( see Discussion ) – is a chief requirement for performance in a changing environment . Yet it is an open problem to understand which circuit mechanisms allow for achieving this ability . How can manifold effective connectivities –corresponding to different patterns of inter-areal interactions , or brain states [17]– result from a fixed structural connectivity ? And how can effective connectivity be controlled without resorting to structural plasticity , leading to a flexible “on demand” selection of function ? Several experimental and theoretical studies have suggested that multi-stability of neural circuits might underlie the switching between different perceptions or behaviors [18]–[22] . In this view , transitions between many possible attractors of the neural dynamics would occur under the combined influence of structured “brain noise” [23] and of the bias exerted by sensory or cognitive driving [24]–[26] . Recent reports have more specifically highlighted how dynamic multi-stability can give rise to transitions between different oscillatory states of brain dynamics [27] , [28] . This is particularly relevant in this context , because long-range oscillatory coherence [12] , [29] –in particular in the gamma band of frequency ( 30–100 Hz ) [29]–[32]– is believed to play a central role in inter-areal communication . Ongoing local oscillatory activity modulates rhythmically neuronal excitability [33] . As a consequence , according to the influential communication-through-coherence hypothesis [31] , neuronal groups oscillating in a suitable phase coherence relation –such to align their respective “communication windows”– are likely to interact more efficiently than neuronal groups which are not synchronized . However , despite accumulating experimental evidence of communication-through-coherence mechanisms [34]–[38] and of their involvement in selective attention and top-down modulation [30] , [39] , [40] , a complete understanding of how inter-areal phase coherence can be flexibly regulated at the circuit level is still missing . In this study we go beyond earlier contributions , by showing that the self-organization properties of interacting brain rhythms lead spontaneously to the emergence of mechanisms for the robust and reliable control of inter-areal phase-relations and information routing . Through large-scale simulations of networks of spiking neurons and rigorous analysis of mean-field rate models , we model the oscillatory dynamics of generic brain circuits involving a small number of interacting areas ( structural connectivity motifs at the mesoscopic scale ) . Following [41] , we extract then the effective connectivity associated to this simulated neural activity . In the framework of this study , we use a data driven rather than a model driven approach to effective connectivity [16] ( see also Discussion section ) , and we quantify causal influences in an operational sense , based on a statistical analysis of multivariate time-series of synthetic “LFP” signals . Our causality measure of choice is Transfer Entropy ( TE ) [42] , [43] . TE is based on information theory [44] ( and therefore more general than causality measures based on regression [45] , [46] ) , is “model-agnostic” and in principle capable of capturing arbitrary linear and nonlinear inter-areal interactions . Through our analyses , we first confirm the intuition that “causality follows dynamics” . Indeed we show that our causal analysis based on TE is able to capture the complex multi-stable dynamics of the simulated neural activity . As a result , different effective connectivity motifs stem out of different dynamical states of the underlying structural connectivity motif ( more specifically , different phase-locking patterns of coherent gamma oscillations ) . Transitions between these effective connectivity motifs correspond to switchings between alternative dynamic attractors . We show then that transitions can be reliably induced through brief transient perturbations properly timed with respect to the ongoing rhythms , due to the non-linear phase-response properties [47] of oscillating neuronal populations . Based on dynamics , this neurally-plausible mechanism for brain-state switching is metabolically more efficient than coordinated plastic changes of a large number of synapses , and is faster than neuromodulation [48] . Finally , we find that “information follows causality” ( and , thus , again , dynamics ) . As a matter of fact , effective connectivity is measured in terms of time-series of “LFP-like” signals reflecting collective activity of population of neurons , while the information encoded in neuronal representations is carried by spiking activity . Therefore an effective connectivity analysis –even when based on TE– does not provide an actual description of information transmission in the sense of neural information processing and complementary analyses are required to investigate this aspect . Based on a general information theoretical perspective , which does not require specifying details of the used encoding [44] , we consider information encoded in spiking patterns [49]–[53] , rather than in modulations of the population firing rate . As a matter of fact , the spiking of individual neurons can be very irregular even when the collective rate oscillations are regular [54]–[57] . Therefore , even local rhythms in which the firing rate is modulated in a very stereotyped way , might correspond to irregular ( highly entropic ) sequences of codewords encoding information in a digital-like fashion ( e . g . by the firing –“1”– or missed firing –“0”– of specific spikes at a given cycle [58] ) . In such a framework , oscillations would not directly represent information , but would rather act as a carrier of “data-packets” associated to spike patterns of synchronously active cell assemblies . By quantifying through a Mutual Information ( MI ) analysis the maximum amount of information encoded potentially in the spiking activity of a local area and by evaluating how much of this information is actually transferred to distant interconnected areas , we demonstrate that different effective connectivity configurations correspond to different modalities of information routing . Therefore , the pathways along which information propagates can be reconfigured within the time of a few reference oscillation cycles , by switching to a different effective connectivity motif . Our results provide thus novel theoretical support to the hypothesis that dynamic effective connectivity stems from the self-organization of brain rhythmic activity . Going beyond previous proposals , which stressed the importance of oscillations for feature binding [59] or for efficient inter-areal “communication-through-coherence” , we advance that the complex dynamics of interacting brain rhythms allow to implement reconfigurable routing of information in a self-organized manner and in a way reminiscent of a clocked device ( in which digital-like spike pattern codewords are exchanged at each cycle of an analog rate oscillation ) .
In order to model the neuronal activity of interacting areas , we use two different approaches , previously introduced in [60] . First , each area is modeled as a large network of thousands of excitatory and inhibitory spiking neurons , driven by uncorrelated noise representing background cortical input ( network model ) . Recurrent synaptic connections are random and sparse . In these networks , local interactions are excitatory and inhibitory . A scheme of the network model for a local area is depicted in Figure 2A ( left ) . In agreement with experimental evidence that the recruitment of local interneuronal networks is necessary for obtaining coherent gamma cortical activity in vitro and in vivo [61] , [62] , the model develops synchronous oscillations ( ) when inhibition is strong , i . e . for a sufficiently large probability of inhibitory connection [54]–[57] , [63] . These fast oscillations are clearly visible in the average membrane potential ( denoted in the following as “LFP” ) , an example trace of which is represented in Figure 2A ( bottom right ) . Despite the regularity of these collective rhythms , the ongoing neural activity is only sparsely synchronized . The spiking of individual neurons is indeed very irregular [54] , [56] and neurons do not fire an action potential at every oscillation cycle , as visible from the example spike trains represented in Figure 2A ( top right ) . Structural network motifs involving areas are constructed by allowing excitatory neurons to establish in addition long-range connections toward excitatory or inhibitory neurons in a distant target area ( see a schematic representation of an structural connectivity motif in Figure 2C ) . The strength of inter-areal coupling is regulated by varying the probability of establishing an excitatory connection . In a second analytically more tractable approach , each area is described by a mean-field firing rate variable ( rate model ) . The firing rate of a local population of neurons obeys the non-linear dynamical equation ( 4 ) ( see Methods ) . All incorporated interactions are delayed , accounting for axonal propagation and synaptic integration . Local interactions are dominantly inhibitory ( with coupling strength and delay ) . Driving is provided by a constant external current . A cartoon of the rate model for a local area is depicted in Figure 2B ( left ) . As in the network model , the firing rates undergo fast oscillations for strong inhibition ( , [60] ) . An example firing rate trace is shown in Figure 2B ( right ) . In order to build structural networks involving areas , different mean-field units are coupled together reciprocally by excitatory long range interactions with strength and delay ( see a schematic representation of an structural motif in Figure 2D ) . Remarkably , the rate model and the network model display matching dynamical states [60] ( see also later , Figures 3 , 4 and 5 ) . More details on the network and the rate models are given in the Methods section and in the Supporting Text S1 . For simplicity , we study fully connected structural motifs involving a few areas ( ) . Note however that our approach might be extended to other structural motifs [64] or even to larger-scale networks with more specific topologies [41] , [65] . In the simple structural motifs we consider , delays and strengths of local excitation and inhibition are homogeneous across different areas . Long-range inter-areal connections are as well isotropic , i . e . strengths and delays of inter-areal interactions are the same in all directions . Delay and strength of local and long-range connections can be changed parametrically , but only in a matching way for homologous connections , in such a way that the overall topology of the structural motif is left unchanged . As previously shown in [60] , different dynamical states –characterized by oscillations with different phase-locking relations and degrees of periodicity– can arise from these simple structural motif topologies . Changes in the strength of local inhibition , of long-range excitation or of delays of local and long-range connections can lead to phase transitions between qualitatively distinct dynamical states . Interestingly , however , within broad ranges of parameters , multi-stabilities between dynamical states with different phase-locking patterns take place even for completely fixed interaction strengths and delays . We generate multivariate time-series of simulated “LFPs” in different dynamical states of our models and we calculate TEs for all the possible directed pairwise interactions . We show then that effective connectivities associated to different dynamical states are also different . The resulting effective connectivities can be depicted in diagrammatic form by drawing an arrow for each statistically significant causal interaction . The thickness of each arrow encodes the strength of the corresponding interaction . This graphical representation makes apparent , then , that effective connectivity motifs or , more briefly , effective motifs , with many different topologies emerge from structural motifs with a same fixed topology . Such effective motifs are organized into families . All the motifs within a same family correspond to dynamical states which are multi-stable for a given choice of parameters , while different families of motifs are obtained for different ranges of parameters leading to different ensembles of dynamical states . We analyze in detail , in Figures 3 , 4 and 5 , three families of motifs arising for strong intra-areal inhibition and similarly small values of delays for local and long-range connections . We consider ( panels A and B ) and ( panels C and D ) structural motifs . Panels A and C show TEs for different directions of interaction , together with “LFPs” and example spike trains ( from the network model ) , and rate traces ( from matching dynamical states of the rate model ) . Panels B and D display motifs belonging to the corresponding effective motif families . A first family of effective motifs occurs for weak inter-areal coupling . In this case , neuronal activity oscillates in a roughly periodic fashion ( Figure 3A and C , left sub-panel ) . When local inhibition is strong , the local oscillations generated within different areas lock in an out-of-phase fashion . It is therefore possible to identify a leader area whose oscillations lead in phase over the oscillation of laggard areas [60] . In this family , causal interactions are statistically significant only for pairwise interactions proceeding from a phase-leading area to a phase-lagging area , as shown by the the box-plots of Figure 3A and C ( right sub-panel , see Discussion and Methods for a discussion of the threshold used for statistical significancy ) . As commented more in detail in the Discussion section , the anisotropy of causal influences in leader-to-laggard and laggard-to-leader directions can be understood in terms of the communication-through-coherence theory . Indeed the longer latency from the oscillations of the laggard area to the oscillations of the leader area reduces the likelihood that rate fluctuations originated locally within a laggard area trigger correlated rate fluctuations within a leading area [35] ( see also Discussion ) . Thus , out-of-phase lockings for weak inter-areal coupling give rise to a family of unidirectional driving effective motifs . In the case of , causality is significant only in one of two possible directions ( Figure 3B ) , depending on which of the two areas assumes the role of leader . In the case of , it is possible to identify a “causal source” area and a “causal sink” area ( see [66] for an analogous terminology ) , such that no direct or indirect causal interactions in a backward sense from the sink area to the source area are statistically significant . Therefore , the unidirectional driving effective motif family for contains six motifs ( Figure 3D ) , corresponding to all the possible combinations of source and sink areas . A second family of effective motifs occurs for intermediate inter-areal coupling . In this case , the periodicity of the “LFP” oscillations is disrupted by the emergence of large correlated fluctuations in oscillation cycle amplitudes and durations . As a result , the phase-locking between “LFPs” becomes only approximate , even if it continues to be out-of-phase on average . The rhythm of the laggard area is now more irregular than the rhythm in the leader area . Laggard oscillation amplitudes and durations in fact fluctuate chaotically ( Figure 4A and C , left sub-panel ) . Fluctuations in cycle length do occasionally shorten the laggard-to-leader latencies , enhancing non-linearly and transiently the influence of laggard areas on the leader activity . Correspondingly , TEs in leader-to-laggard directions continue to be larger , but TEs in laggard-to-leader directions are now also statistically significant ( Figure 4A and C , right sub-panel ) . The associated effective motifs are no more unidirectional , but continue to display a dominant direction or sense of rotation ( Figure 4B and D ) . We refer to this family of effective motifs as to a family of leaky driving effective motifs ( containing two motifs for and six motifs for ) . Finally , a third family of effective motifs occurs for stronger inter-areal coupling . In this case the rhythms of all the areas become equally irregular , characterized by an analogous level of fluctuations in cycle and duration amplitudes . During brief transients , leader areas can still be identified , but these transients do not lead to a stable dynamic behavior and different areas in the structural motif continually exchange their leadership role ( Figure 5A and C , left sub-panel ) . As a result of the instability of phase-leadership relations , only average TEs can be evaluated , yielding to equally large TE values for all pairwise directed interactions ( Figure 5A and C , right sub-panel ) . This results in a family containing a single mutual driving effective motif ( Figure 5B and D ) . Further increases of the inter-areal coupling strength do not restore stable phase-locking relations and , consequently , do not lead to additional families of effective motifs . Note however that the effective motif families explored in Figures 3 , 4 and 5 are not the only one that can be generated by the considered fully symmetric structural motifs . Indeed other dynamical configurations exist . In particular , anti-phase locking ( i . e . locking with phase-shifts of for and of for ) would become stable when assuming the same interaction delays and inter-areal coupling strengths of Figures 3 , 4 and 5 , but a weaker local inhibition . Assuming different interaction delays for local and long-range interactions , out-of-phase lockings continue to be very common , but in-phase and anti-phase locking can become stable even for strong local inhibition , within specific ranges of the ratio between local and long-range delays [60] . For , in the case of general delays , more complex combinations can arise as well , like , for instance , states in which two areas oscillate in-phase , while a third is out-of-phase . In-phase locking between areas gives rise to identical TEs for all possible directed interactions , resulting in effective motifs without a dominant directionality . Anti-phase lockings for give rise to relatively large inter-areal phase-shifts and , correspondingly , to weak inter-areal influences ( at least in the case of weak inter-areal coupling ) , resulting in small TE levels which are not statistically significant ( not shown ) . However , in the framework of this study , we focus exclusively on out-of-phase-locked dynamical states , because they are particularly relevant when trying to achieve a reconfigurable inter-areal routing of information ( see later results and Discussion section ) . To conclude , we remark that absolute values of TE depend on specific parameter choices ( notably , on time-lag and signal quantization , see Methods ) . However , the relative strengths of TE in different directions –and , therefore , the resulting topology of the associated effective motifs– are rather robust against changes of these parameters . Robustness of causality estimation is analyzed more in detail in the Discussion section . How can asymmetric causal influences emerge from a symmetric structural connectivity ? A fundamental dynamical mechanism involved in this phenomenon is known as spontaneous symmetry breaking . As shown in [60] , for the case of the structural motif , a phase transition occurs at a critical value of the strength of inter-areal inhibition . When local inhibition is stronger than this critical threshold , a phase-locked configuration in which the two areas oscillate in anti-phase loses its stability in favor of a pair of out-of-phase-locking configurations , which become concomitantly stable . The considered structural motif is symmetric , since it is left unchanged after a permutation of the two areas . However , while the anti-phase-locking configuration , stable for weak local inhibition , share this permutation symmetry with the full system , this is no more true for the out-of-phase-locking configurations , stable for strong local inhibition . Note , nevertheless , that the configuration in which leader and laggard area are inverted is also a stable equilibrium , i . e . the complete set of stable equilibria continue to be symmetric , even if individual stable equilibria are not ( leading thus to multi-stability ) . In general , one speaks about spontaneous symmetry breaking whenever a system with specific symmetry properties assumes dynamic configurations whose degree of symmetry is reduced with respect to the full symmetry of the system . The occurrence of symmetry breaking is the signature of a phase transition ( of the second order [67] ) , which leads to the stabilization of states with reduced symmetry . The existence of a symmetry-breaking phase transition in the simple structural motifs we analyze here ( for simplicity , we consider the case ) can be proven analytically for the rate model , by deriving the function , which describes the temporal evolution of the phase-shift between two areas when they are weakly interacting [47]: ( 1 ) The function for the rate model is shown in the left panel of Figure 6B . Stable phase lockings are given by zeroes of with negative slope crossing and are surrounded by basins of attraction ( i . e . sets of configurations leading to a same equilibrium ) , whose boundaries are unstable in- and anti-phase lockings ( Figure 6A ) . For the network model , a function with an analogous interpretation and a similar shape , shown in the right panel of Figure 6B , can be extracted from simulations , based on a phase description of “LFP” time-series ( see Methods and Supporting Figure S1A ) . The analogous distribution of the zero-crossings of and results in equivalent phase-locking behaviors for the rate and network models . Thus spontaneous symmetry breaking leads to multi-stability between alternative out-of-phase-lockings and to the emergence of unidirectional effective driving within a symmetric structural motif . Because of multi-stability , transitions between effective motifs within a family can be triggered by transient perturbations , without need for structural changes . We theoretically determine conditions for such transitions to occur . The application of a pulse of current of small intensity advances or delays the phase of the ongoing local oscillation ( see Supporting Figure S1B ) . This is true for rate oscillations of the mean-field rate model , but also for “LFP” oscillations reflecting rhythmic synchronization in the network model . In the latter case , the collective dynamics is perturbed by synchronously injecting pulse currents into all of the neurons within an area . The induced phase shift depends on the perturbation strength but also on the phase at which the perturbation is applied . For the network model , this can be measured directly from numeric simulations of a perturbed dynamics ( see Methods and right panel of Figure 6D ) . For the rate model , the phase shift induced by an instantaneous phased perturbation can be described analytically in terms of the Phase Response Curve ( PRC ) [47] ( see Figure 6D , left , and Supporting Text S1 ) . After a pulse , the phase-shift between two areas is “kicked out” of the current equilibrium locking and assumes a new transient value ( solid paths in Figure 6C ) , which , for weak perturbations and inter-areal coupling , reads: ( 2 ) where the approximate equality between square brackets holds for the mean-field rate model . If falls into the basin of attraction of a different phase-locking configuration than , the system will settle within few oscillation cycles into an effective connectivity motif with a different directionality ( dashed green path in Figure 6C ) . Even relatively small perturbations can induce an actual transition , if applied in selected narrow phase intervals in which the induced grows to large values . For most application phases , however , even relatively large perturbations fail to modify the effective driving direction ( dashed red path in Figure 6C ) , because the induced perturbation is vanishingly small over large phase intervals ( Figure 6D ) . This is a robust property , shared by the two ( radically different ) models we consider here and –we hypothesize– by any local circuit generating fast oscillations through a mechanism based on delayed mutual inhibition . As a consequence , for a given perturbation intensity , a successful switching to a different effective motif occurs only if the perturbation is applied within a specific phase interval , that can be determined analytically from the knowledge of and of for the rate model , or semi-analytically from the knowledge of and ( see Methods ) . Figure 6E–F reports the fraction of simulated phased pulses that induced a change of effective directionality as a function of the phase of application of the perturbation . The phase intervals for successful switching predicted by the theory are highlighted in green . We performed simulations of the rate ( Figs . 6E–F , left column ) and of the network ( Figs . 6E–F , middle column ) models , for unidirectional ( Figs . 6E ) and leaky driving ( Figs . 6F ) effective motifs . Although our theory assumes small inter-areal coupling and is rigorous only for the rate model , the match between simulations and predictions is very good for both models and families of motifs . In Figs . 6E–F , we perturb the dynamics of the laggard area , but changes in directionality can also be achieved by perturbing the leader area ( Supporting Figure S2 ) . Note also that , in the network model , direction switchings can take place spontaneously , due to noisy background inputs . Such noise-induced transitions , however , occur typically on time-scales of the order of seconds , i . e . slow in terms of biologic function , because the phase range for successful switching induction is narrow . A second non-linear dynamic mechanism underlying the sequence of effective motifs of Figures 3 and 4 is effective entrainment . In this phenomenon , the complex dynamics of neural activity seems intriguingly to be dictated by effective rather than by structural connectivity . We consider as before a rate model of reciprocally connected areas ( Figure 2D ) . In order to properly characterize effective entrainment , we review the concept of bifurcation diagram [68] . As shown in [60] , when the inter-areal coupling is increased , rate oscillations become gradually more complex ( cfr . Figure 7A ) , due to the onset of deterministic chaos ( see also [69] for a similar mechanism in a more complex network model ) . For small , oscillations are simply periodic ( e . g . ) . Then , for intermediate ( e . g . ) , the peak amplitudes of the laggard area oscillation assume in alternation a small number of possible values ( period doubling ) . Finally , for larger ( e . g . ) , the laggard peak amplitudes fluctuate in a random-like manner within a continuous range . This sequence of transitions can be visualized by plotting a dot for every observed value of the peak amplitudes of oscillation cycles , at different values of . The accumulation of these dots traces an intricate branched structure , which constitutes the bifurcation diagram ( Figure 7B ) . Bifurcation diagrams for the leader and for the laggard area are plotted in Figure 7B ( top panel , in orange and green color , respectively ) . We compare these bifurcation diagrams with the analogous diagrams constructed in the case of two unidirectionally coupled oscillating areas . Qualitatively similar bifurcation sequences are associated to the dynamics of the laggard area ( bidirectional coupling ) and of the driven area ( unidirectional coupling , Figure 7B , bottom panel , green color ) , for not too strong inter-areal couplings . In the case of unidirectional coupling , the peak amplitudes of the unperturbed driver area oscillations do not fluctuate at all . Therefore , the corresponding bifurcation diagram is given by a constant line ( Figure 7B , bottom panel , orange color ) . In the case of bidirectional coupling , the peak amplitudes of the leader area oscillations undergo fluctuations , but only with a tiny variance . Thus , the corresponding bifurcation diagram has still the appearance of a line , although now “thick” and curved ( zooming would reveal bifurcating branches ) . Note that , for unidirectional coupling , the structural connectivity is explicitly asymmetric . The periodic forcing exerted by the driving area is then known to entrain the driven area into chaos [70] . Such direct entrainment is the dynamical cause of chaos . On the other hand , for bidirectional coupling , the structural connectivity is symmetric . However , due to spontaneous symmetry breaking , the resulting effective connectivity is asymmetric and the system behaves as if the leader area was a driver area , entraining the laggard area into chaos being only negligibly affected by its back-reaction . Such effective entrainment can be seen as an emergent dynamical cause of chaos . Thus , the dynamics of a symmetric structural motif with asymmetric effective connectivity and of a structural motif with a matching asymmetric topology are equivalent . For a sufficiently strong inter-areal coupling , symmetry in the dynamics of the bidirectional structural motif is suddenly restored [60] , in correspondence with a transition to the mutual driving family of effective motifs ( Figure 5 ) . As a result , in absence of symmetry breaking , effective driving cannot anymore take place . Thus , for a too strong inter-areal coupling , the emergent anisotropy of effective connectivity is lost , and , with it , the possibility of a dynamic control of effective connectivity ( at least via the previously discussed strategies ) . Despite its name , Transfer Entropy is not directly related to a transfer of information in the sense of neuronal information processing . The TE from area to area measures indeed just the degree to which the knowledge of the past “LFP” of reduces the uncertainty about the future “LFP” of [43] , [71] . As a matter of fact , however , the information stored in neural representations must be encoded in terms of spikes , independently from the neural code used . Therefore , it is important to understand to which extent an effective connectivity analysis based on “macroscopic” dynamics ( i . e . TEs estimated from “LFPs” ) can pretend to describe actual “microscopic” information transmission ( i . e . at the level of spiking correlations ) . In order to address this issue , we first introduce a framework in which to quantify the amount of information exchanged by interacting areas . In the case of our model , rate fluctuations could encode only a limited amount of information , since firing rate oscillations are rather stereotyped . On the other hand , a larger amount of information could be encoded based on spiking patterns , since the spiking activity of single neurons is very irregular and thus characterized by a large entropy [44] , [58] . As illustrated by Figure 8A , a code can be built , in which a “1” or a “0” symbol denote respectively firing or missed firing of a spike by a specific neuron at a given oscillation cycle . Based on such an encoding , the neural activity of a group of neurons is mapped to digital-like streams , “clocked” by the ongoing network rhythm , in which a different “word” is broadcast at each oscillation cycle . Note that we do not intend to claim that such a code is actually used in the brain . Nevertheless , we introduce it as a theoretical construct grounding a rigorous analysis of information transmission . We focus here on the fully symmetric structural motif of areas of Figure 2C . We modify the network model considered in the previous sections by embedding into it transmission lines ( TLs ) , i . e . mono-directional fiber tracts dedicated to inter-areal communication ( see Figure 8B ) . In more detail , selected sub-populations of source excitatory neurons within each area establish synaptic contacts with matching target excitatory or inhibitory cells in the other area , in a one-to-one cell arrangement . Synapses in a TL are strengthened with respect to usual synapses , by multiplying their peak conductance by a multiplier ( see Methods section ) . Such multiplier is selected to be large , but not too much , in order not to affect the phase-relations between the collective oscillations of the two areas . Indeed , selecting a too large would lead to an in-phase-locking configuration in which collective dynamics is enslaved to the synchronous activity of source and target populations . As analyzed in the Supporting Figure S3 , a suitable tuning of ensures that source-to-target neuron communication is facilitated as much as possible , without disrupting the overall effective connectivity ( associated to the unperturbed phase-locking pattern ) . Note that such TL synapses are here introduced as a heuristic device , allowing to maximize the potential capacity of inter-areal communication channels [44] . However , due to the occurrence of consistent spike-timing relations in out-of-phase locked populations , it might be that spike-timing-dependent plasticity [72] lead to the gradual emergence of subsets of synapses with substantially enhanced weight [73] , which would play a role in inter-circuit communication very similar to TL synapses . The information transmission efficiency of each TL , for the case of different effective motifs , is then assessed by quantifying the Mutual Information ( MI ) [44] , [58] between the “digitized” spike trains of pairs of source and target cells ( see Methods ) . Since a source cell spikes on average every five or six oscillation cycles , the firing of a single neuron conveys of information per oscillation cycle . MI normalized by the source entropy H indicates how much of this information reaches the target cell , a normalized MI equal to unity denoting lossless transmission . As shown by Figure 8C–D , the communication efficiency of embedded TLs depends strongly on the active effective motif . In the case of unidirectional driving effective motifs ( Figure 8C ) , communication is nearly optimal along the TL aligned with the effective connectivity . For the misaligned TL , however , no enhancement occurs with respect to control ( i . e . pairs of connected cells not belonging to a TL ) . In the case of leaky driving effective motifs ( Figure 8D ) , communication efficiency is boosted for both TLs , but more for the TL aligned with the dominant effective direction . For both families of effective motifs , despite the strong anisotropy , the communication efficiencies of the two embedded TLs can be “swapped” within one or two oscillation cycles , by reversing the active effective connectivity through a suitable transient perturbation ( see Figure 6E–F ) . The considered structural motif acts therefore as a “diode” through which information can propagate efficiently only in one ( dynamically reconfigurable ) direction determined by effective connectivity .
We have shown that a simple structural motif of interacting brain areas can give rise to multiple effective motifs with different directionality and strengths of effective connectivity , organized into different families . Such effective motifs correspond to distinct dynamical states of the underlying structural motif . Beyond this , dynamic multi-stability makes the controlled switching between effective motifs within a same family possible without the need for any structural change . On the contrary , transitions between effective motifs belonging to different families ( e . g . a transition from a unidirectional to a leaky driving motif ) cannot take place without changes in the strength of the delay of inter-areal couplings , even if the overall topology of the underlying structural motif does not need to be modified . Each specific effective motif topology ( i . e . motif family ) is robust within broad ranges of synaptic conductances and latencies , however if parameters are set to be close to critical transition lines separating different dynamical regimes , transitions between different families might be triggered by moderate and unspecific parameter changes . This could be a potential role for neuromodulation , known to affect the net efficacy of excitatory transmission and whose effect on neural circuits can be modeled by coordinated changes in synaptic conductances [74] , [75] . Note that dynamical coordination of inter-areal interactions based on precisely-timed synchronous inputs would be compatible with experimental evidence of phase-coding [76]–[81] , indicating a functional role for the timing of spikes relative to ongoing brain rhythms ( stimulus-locked [82] , [83] as well as stimulus-induced or spontaneous [84] ) . Note also that the time of firing is potentially controllable with elevated precision [85]–[87] and has been found to depend on the phase of LFPs in local as well as in distant brain areas [37] . In general , control protocols different from the one proposed here might be implemented in the brain . For instance , phased pulses might be used as well to stabilize effective connectivity in the presence of stronger noise . Interestingly , the time periods framed by cycles of an ongoing oscillation can be sliced into distinct functional windows in which the application of the same perturbation produces different effects . Finally , in addition to “on demand” transitions , triggered by exogenous –sensory-driven– or endogenous –cognitive-driven– control signals , noise-driven switching between effective motifs might occur spontaneously , yielding complex patterns of activity during resting state [26] , [88] , [89] . As revealed by our discussion of spontaneous symmetry breaking and effective entrainment , an analysis based on TE provides a description of complex inter-areal interactions compliant with a dynamical systems perspective . It provides , thus , an intuitive representation of dynamical states that is in the same “space” as anatomical connectivity . Note that it is currently debated whether TE should be considered as a measure of effective connectivity in strict sense [13] , [15] , or , rather , of yet another type of connectivity beyond functional connectivity ( that could be dubbed causal connectivity [16] , [66] or directed functional connectivity ) . Our position is that TE constitutes , at least in the context of the present study , a measure of effective connectivity in proper sense . Indeed , as indicated by the analysis of Figure 8C–D , the connectivity motifs inferred by TE correctly represent characteristic dynamic mechanisms , like spontaneous symmetry breaking or asymmetric chaos [60] , enabling specifically associated modalities of inter-areal information transmission . Therefore , we can conclude that causality ( as inferred by TE ) follows dynamics ( by representing the action of corresponding dynamic mechanisms ) . TE constitutes thus a model-free approach ( although , non “parameter-free” , cfr . forthcoming section and Figure 9 ) to the effective connectivity problem , suitable for exploratory data-driven analyses . In this sense it differs from regression-based methods like usual implementations of Granger Causality ( GC ) [45] , [46] or from Dynamic Causal Modeling ( DCM ) [90] , which are model-driven [15] , [16] , [91] . Strategies like DCM , in particular , assume prior knowledge about the inputs to the system and works by comparing the likelihood of different a priori hypotheses about interaction structures . Such an approach has the undeniable advantage of providing a direct description of actual mechanisms underlying effective connectivity changes ( the stimulus-dependence of effective couplings is indeed modeled phenomenologically ) . However , it might be too restrictive ( or arbitrary ) when the required a-priori information is missing or highly uncertain . TE , on the contrary: does not require any hypothesis on the type of interaction; should be able to detect even purely non-linear interactions and should be robust against linear cross-talk between signals [92] . These features , together with the efficacy of TE for the causal analysis of synthetic time-series , advocate for a more widespread application of TE methods to real neural data [93]–[95] ( at the moment limited by the need of very long time-series [92] ) . Note that we do not intend to claim superiority of TE in some general sense . As a matter of fact TE is equivalent to GC , as far as the statistics of the considered signals are gaussian [71] . Furthermore , non-linear generalizations of GC and DCM [96]–[99] might be able to capture at a certain extent the complex self-organized dynamics of the neural activity models analyzed in the present study . However , a systematic comparison of the performance of different methods in capturing causal connectivity of realistic non-linear models of neural dynamics goes beyond the focus of the present study and is deferred to future research . We finally would like to stress , to avoid any potential confusion , that the structural motifs analyzed in the present study are well distinct from causal graphical models of neural activity , in the statistical sense proper of DCMs [90] , [100] . They constitute indeed actual mechanistic models of interacting populations of spiking neurons , with a highly non-linear dynamics driven by background noise . Connections in these models are model synapses , i . e . mere structural couplings , not phenomenological effective couplings . Thus , effective connectivity is not constrained a priori , as in DCMs , but is an emergent property of network dynamics , consistent with the existence of effective motif topologies different from the underlying structural topology . The effective connectivity analyses presented in this study were conducted by evaluating TEs under specific parameter choices . However , absolute values of TE depend on parameters , like , notably , the resolution at which “LFP” signals are quantized and the time-lag at which we probe causal interactions . As discussed in detail in the Methods section , estimation of TE requires the sampling of joint distributions of “LFP” values in different areas at different times . Such distributions are sampled as histograms , based on discrete multi-dimensional binning . In practice , each “LFP” time-series is projected to a stream of symbols from a discrete alphabet , corresponding to different quantization levels of the continuous “LFP” signals [101] . The actual number of used bins is a free parameter , although some guiding criteria for its selection do exist [43] . Concerning time-lag , our TE analysis ( conducted at the first Markov order [42] , following [41] , [94] ) describes predictability of “LFPs” at time based on “LFPs” at time . The used time-lag is once again a free parameter . To deal with this arbitrariness in parameter choices , we explore systematically the dependence of TE estimations from the aforementioned parameters , by varying both and in a wide continuous range . Figure 9 summarizes the results of this analysis , for three different effective motifs . Considering the dependency on time-lag , a periodic structure is clearly noticeable in the TE matrices reported in Figure 9 . TE values tend to peak in precise bands of , related to latencies between the oscillations of different areas . The analysis of the unidirectional driving motif ( Figure 9A ) , associated to leader-laggard periodic configurations , is particularly transparent ( and has a high pedagogic value ) . Two characteristic time-lags can be defined: a “short” lag , given by the time-shift from oscillation peaks of the leader area to oscillation peaks of the laggard area ; and a “long” lag , , given by the time-shift from laggard to leader oscillation peaks ( here , is an average oscillation period , common to both areas leader and laggard areas and ) . TE in the direction from leader to laggard , , peaks for the first time at a time-lag ( and then at lags , where is a positive integer ) . TE in the direction from laggard to leader , , peaks first at a time-lag ( and then at lags ) . If the “LFP” signals were deterministic and strictly periodic , the quantities and would be identical ( and diverging for infinite precision [42] ) . However “LFP” signals are only periodic on average and have a stochastic component , due to the joint effect of random network connectivity and noisy background inputs . This stochastic component is responsible for small cycle-to-cycle fluctuations in the amplitude of “LFP” oscillation peaks . As discussed more in depth in a next subsection , the efficiency with which fluctuations in the output of a local area can induce ( i . e . , can “cause” ) fluctuations of the output of a distant interconnected area depends on the instantaneous local excitability of this target area , which is undergoing a rhythmic modulation due to the ongoing collective oscillation [31] , [33] . As a result , TE can reach different peak values in different directions ( and , as a matter of fact , ) . Considering then the dependence on signal quantization , we observe that TE values tend to grow for increasing number of bins , i . e . for a finer resolution in tracking “LFP” amplitude variations . This can be once again understood in terms of the temporal structure of “LFP” signals . As just mentioned , dynamic correlations between small “LFP” amplitude fluctuations carry information relevant for causality estimation . This information would be completely lost by using a too small number of bins for TE evaluation , given that the largest contribution to the dynamic range of “LFP” signals is provided by its fairly stereotyped oscillatory component . Conversely , using a too large number of bins would lead to under-sampling artifacts ( therefore , we do not consider the use of more than quantization bins ) . By evaluating a threshold for statistical significance independently for each direction and combination of and , we find that , for weak inter-areal coupling , TE never goes above this threshold in the laggard-to-leader direction ( Figure 9A ) . We are also unable to find any choice of and such that , for intermediate inter-areal coupling , TE in the laggard-to-leader direction becomes larger or equal than TE in leader-to-laggard direction ( Figure 9B ) . Looking at matrices of the causal unbalancing ( see Methods , and Figure 9 , third column ) , we see indeed that , for weak and intermediate coupling strengths , effective connectivity is robustly asymmetric in the parameter regions in which causal interactions are statistically significant . Effective connectivity is on the contrary balanced for strong inter-areal coupling ( Figure 9C ) . We can thus summarize the previous statements by saying that absolute values of TE depend on the choices of and , but that the topology of the resulting effective motif does not ( at least in the wide range considered for this robustness analysis ) . Traditionally , studies about communication-through-coherence or long-range binding between distant cell assemblies have emphasized the importance of in-phase locking ( see , e . g . [35] , [102] ) . Although , as previously mentioned , in-phase locking ( as well as anti-phase locking ) can also arise in our models for different choices of coupling delays and inhibition strengths [60] , we decided in the present study to focus on out-of-phase lockings . The case of spontaneous symmetry breaking is indeed particularly interesting , because it underlie the emergence of a dominant directionality in the causal influences between areas reciprocally coupled with comparable strengths . Furthermore , spontaneous symmetry breaking is responsible for the multi-stability between effective connectivity configurations , thus opening the way to a self-organized control of inter-areal interactions [11] , [12] . In particular , our study confirms that the reorganization of oscillatory coherence might regulate the relative weight of bottom-up and top-down inter-areal influences [17] , [30] or select different interaction modes within cortical networks involving areas of similar hierarchical level , as in the case of motor preparation or planning [4] , [103] or language [104] . As a next step , we directly verified that “information follows causality” , since changes in effective connectivity are paralleled by reconfiguration of inter-areal communication modalities . Following [32] , [35] , we explain the anisotropic modulations of communication efficiency ( see Figure 8 ) in terms of a communication-through-coherence mechanism . In fact , because of the out-of-phase locking between rhythms , spikes emitted by neurons in a phase-leading area reach neurons in a phase-lagging area at a favorable phase in which they are highly excitable . Conversely , spikes emitted by neurons in a phase-lagging area reach neurons in a phase-leading area when they are strongly hyperpolarized by a preceding peak of synchronous inhibition . This same mechanism underlie also the anisotropy of “LFP”-based TE , since “LFP” fluctuations are the manifestation ( at least in our model ) of local population firing rate fluctuations . Therefore , by combining TE analyses of “LFP”-based effective connectivity with MI analyses of spike-based information transmission , we are able to establish a tight link between control of effective connectivity and control of communication-through-coherence , both of them being emergent manifestations of the self-organized dynamics of interacting brain rhythms . To conclude , we also note that similar mechanisms might be used beyond the mesoscale level addressed here . Multi-stabilities of structural motifs might be preserved when such motifs are interlaced as modules of a network at the whole-brain level [64] . Likewise , dynamic control of information routing between neuronal clusters [73] , [105] or even single cells might occur within more local microcircuits [106] , [107] . The previous discussions suggest that oscillations , rather than playing a direct role in the representation of information , would be instrumental to the reconfigurable routing of information encoded in spiking activity . Original formulations of the communication-through-coherence hypothesis [31] suggested that oscillatory coherence facilitates the transmission of local fluctuations of firing rate to a distant site , thus assuming implicitly a rate-based encoding of information in neuronal activity . However , more complex coding mechanisms based on patterns of precisely timed spikes might be achievable by biologically-plausible neuronal circuits [85] , [86] . As a matter of fact , our study reveals that the inherent advantages of “labelled-line” codes [51] , [108] ( in which the information about which local neuron is firing is preserved ) –i . e . , notably , an augmented information capacity with respect to “summed-population” codes– might be combined with the flexibility and the reliability offered by the communication-through-coherence framework . Indeed , as shown by the analyses of Figure 8 , suitable inter-areal phase relations make possible the transmission of information encoded in detailed spiking correlations , rather than just in population firing rate fluctuations . This is particularly interesting , since many cortical rhythms are only sparsely synchronized , with synchronous oscillations evident in LFP , Multi-Unit Activity or intracellular recordings but not in single unit spike trains [109]–[111] . Such sparse firing might possibly reflect population-coding of behaviorally-relevant information transcending rate-based representations [49]–[53] . Independently from the complexity of these hypothetic representations , our study shows that self-organized communication-through-coherence would have the sufficient potential to dynamically route the rich information that these representations might convey . It is very plausible that flexible inter-areal coordination is achieved in the brain through dynamic self-organization [11] as in our models . However , qualitatively different mechanisms than symmetry breaking might contribute to the generation of dynamic effective connectivity in other regimes of activity . Despite sparse synchronization , the level of coherence in our model neuronal activity is larger than in many brain oscillations . However , our results might be generalized to activity regimes in which synchronization is weaker . Phase-relations have been shown to impact effective connectivity even in essentially asynchronous regimes [112] . It would be interesting to understand whether the dominant directionality of effective connectivity can be controlled when out-of-phase locking is only transient [12] , [41] . Another open question is whether our theory can be extended to encompass the control of effective connectivity across multiple frequency bands [94] . This is an important question since top-down and bottom-up inter-areal communication might exploit different frequency channels , possibly due to different anatomic origins of ascending and descending cortico-cortical connections [113] . Finally , we are confident that our theory might inspire novel experiments , attempting to manipulate the directionality of inter-areal influences via local stimulation applied conditionally to the phase of ongoing brain rhythms . Precisely timed perturbing inputs could indeed potentially be applied using techniques like electric [114] or optogenetic [115] microstimulation , especially in closed-loop implementations with millisecond precision [116] , [117] .
Each area is represented by a random network of excitatory and inhibitory Wang-Buzsáki-type conductance-based neurons [118] . The Wang-Buzsáki model is described by a single compartment endowed with sodium and potassium currents . Note that results ( not shown ) of simulations performed with a more realistic ratio of excitatory and inhibitory neurons per population would lead to qualitatively similar results with small parameter adjustments ( using , for instance , parameters as in [69] ) . The membrane potential is given by: ( 3 ) where is the capacitance of the neuron , is a leakage current , is an external noisy driving current ( due to background Poisson synaptic bombardment ) , and and are respectively a sodium and a potassium current , depending non linearly on voltage . The last input term is due to recurrent interactions with other neurons in the network . Excitatory synapses are of the AMPA-type and inhibitory synapses of the GABA-type and are modeled as time-dependent conductances . A complete description of the model and a list of all its parameters are given in the Supporting Text S1 . “LFP” is defined as the average membrane potential over the cells in each area . Short-range connections within a local area from population to population are established randomly with probability , where and can be either one of the type ( excitatory ) or . The excitatory populations are allowed also to establish connections toward populations and in remote areas ( ) . Such long-range connections are established with a probability ( ) . For simplicity , however , we assume that and that . For each of the considered dynamical states , probabilities of connection are provided in the corresponding figure caption . First , a structural motif of interconnected random networks of spiking neurons is generated , as in the previous section . Then , on top of the existing excitatory long-range connections , additional stronger long-range connections are introduced in order to form directed transmission lines . In each area a source sub-population , made out of 400 excitatory neurons , and a non-overlapping target sub-population , made out of 200 excitatory and 200 inhibitory neurons , are selected randomly . Excitatory cells in the source populations get connected to cells in the target sub-populations of the other area via strong synapses . These connections are established in a one-to-one arrangement ( e . g . each source cell establishes a TL-synapse with a single target cell that does not receive on its turn any other TL-synapse ) . The peak conductance of TL-synapses is times stronger than the normal excitatory peak conductance . For the simulations with TL ( Figure 8 of the main paper ) , we set respectively for the unidirectional and for the leaky driving effective motifs . Such unrealistically strong peak conductances , whose purpose is to optimize information transfer by enhancing spiking correlations , can be justified by supposing that each source neuron establishes multiple weaker synaptic contacts with the same target neuron . The multiplier is selected to be as large as possible without altering the original out-of-phase locking relations between the two populations ( Figure S3A ) . Concretely , is tuned by raising it gradually until when a critical point is reached in which the populations lock in-phase ( Figure S3C ) . Then , is set to be just below this critical point ( Figure S3B ) . Each area is represented by a single rate unit . The dynamical equations for the evolution of the average firing rate in an area are given by: ( 4 ) Here , if , and zero otherwise . A constant current represents a background input , stands for the strength of intra-areal inhibition , for the strength of inter-areal excitation and and are the delays of the local and long-range interactions , respectively . We consider in this study only fully symmetric structural motifs of mutually connected areas . For each of the considered dynamical states , the values of , , and are provided in the figure caption . Given an oscillatory time-series of neuronal activity , generated indifferently by a rate or by a network model , a phase , for , is linearly interpolated over each oscillation cycle . Here denotes the start time of the oscillation cycle . Note that this definition does not require that the oscillation is periodic: this empiric phase “elastically” adapts to fluctuations in the duration of oscillation cycles ( see Supporting Figure S1A ) . The phase shift induced by a pulse perturbation ( see Supporting Figure S1B ) is described by the Phase Response Curve ( PRC ) ( see Eq . ( 2 ) and [47] ) . For the rate model , the PRC can be evaluated analytically if certain general conditions on the relation between the oscillation period and the local inhibition delay are fulfilled [60] . Analytical expressions for the PRC of the rate model , as plotted in Figure 6D ( left ) , are reported in the Supporting Text S1 . In the network model , it is possible to evaluate the phase-shift induced by a perturbation , by directly simulating the effects of this perturbation on the oscillatory dynamics . A perturbation consists of a pulse current of strength injected synchronously into all neurons of one area at a phase of the ongoing local oscillation . The induced phase-shift is estimated by comparing the phases of the perturbed and of the unperturbed oscillations , when a new equilibrium is reached after the application of the perturbation . In detail , since the “LFP” time-series are not strictly periodic and the phase relation is fixed only on average , the average time-lag between the perturbed and the unperturbed “LFPs” is measured by computing their crosscorrelogram over 50 oscillation cycles , starting from the 10-th cycle after the perturbation . This average time lag ( readable from the position of the crosscorrelogram peak ) is then translated into a phase-shift , by dividing it by the average period ( estimated through autocorrelation analysis of the perturbed and unperturbed time-series over the same observation window ) . Vanishingly small perturbations do not induce long-lasting phase-shifts . Therefore , moderately large perturbation strengths have to be used . In this case , the dependence of on is sensibly non-linear . As a consequence , we evaluate directly the resulting for the used perturbation strength , plotted in Figure 6D ( right ) . The qualitative shape of however does not depend strongly on . In particular , changes of affect the amplitude of the maximum phase-shift but not the perturbation phase for which it occurs . The curve is evaluated point-wise by applying perturbations at 100 different phases within a cycle . For each given phase , the perturbation is applied 100 times to 100 different cycles and the corresponding phase-shifts are averaged . Confidence intervals for are determined phase-by-phase by finding the 2 . 5-th and the 97 . 5-th percentile of the induced phase-shift distribution across these 100 trials . For simplicity , we focus in the following on the case of areas , although our approach can be extended to larger motifs . The instantaneous phase-difference between two areas and is given by . For vanishing inter-areal coupling , the time evolution of is described by Eq . ( 1 ) . The term is a functional of the phase response and of the limit cycle waveform of the uncoupled oscillating areas . For the rate model , is determined from analytic expressions of and of the rate oscillation limit cycle ( note that the dependence on is replaced by a dependence on after phase-reduction ) for . It can be expressed as , with: ( 5 ) The resulting expression is reported in the Supporting Text S1 and plotted in Figure 6B ( left ) . Given Eq . ( 1 ) , the phase shifts between the two areas and in stable phase-locked states correspond to top-down zero-crossings of the functional ( i . e . zeroes with negative tangent slope , ) . For the network model , the waveform of “LFP” oscillations can be determined through simulations . Since not all oscillation cycles are identical , the limit cycle waveform is averaged over 100 different cycles –as for the determination of – to yield an average limit cycle . Then , it is possible to evaluate a functional , where: ( 6 ) The functional is plotted in Figure 6B ( right ) for the used perturbation strength . Although Eq . ( 1 ) does not exactly hold for the network model , the top-down zero-crossings of the functional ( whose position only weakly depends on ) continue to provide an approximation of the phase shifts between the two areas and in stable phase-locked states . In particular it is possible to predict whether the stable lockings will be in-phase , anti-phase or out-of-phase . Phase intervals in which the application of a pulsed perturbation leads to a change of effective connectivity directionality are determined theoretically as shown below . For and in a given phase-locking state , the phase of the leader area can be written as and the phase of the laggard area as . The application of a pulse perturbation of strength to the laggard area shifts the phase of the ongoing local oscillation to , where holds for the rate model in the case of small perturbations . If the achieved transient phase-shift between the two areas , , is falling into the basin of attraction of an alternative stable phase-locking ( see Figure 6C ) , then a switching toward a different effective motif takes place . Considering the dynamics of the instantaneous phase-shift , determined by the functionals for the rate model and for the network model ( see Figure 6B ) , switching will occur when: ( 7 ) Here , we consider perturbations which induce a phase advancement , because the positive part of both the PRC in the rate model and the empiric in the network model is larger than the negative part ( see Figure 6D ) . For a fixed perturbation intensity , the condition ( 7 ) will be fulfilled only if when the phase of application of the perturbation falls within specific intervals , determined by the precise form of . These intervals are highlighted in green in Figure 6E and F . Analogous considerations can be done in order to determine the intervals for successful switching when perturbing the leader area ( see Supporting Figure S2 ) . Let us consider first a structural motif with areas . Let and be the “LFP” time-series of the two areas and , and let quantize them into discrete levels ( bins are equally sized ) . The continuous-valued “LFP” time-series are thus converted into strings of symbols and from a small alphabet [101] . Two transition probability matrices , and , where the lag is an arbitrary temporal scale on which causal interactions are probed , are then sampled as normalized multi-dimensional histograms over very long symbolic sequences . These probabilities are sampled separately for each specific fixed phase-locking configuration . Epochs in which the system switches to a different phase-locking configuration , as well as transients following state switchings are dropped . The evaluation of and is thus based on disconnected symbolic subsequences , including overall oscillation cycles . Then , following [42] , the causal influence of area on area is defined as the Transfer Entropy: ( 8 ) where the sum runs over all the three indices , and of the transition matrices . This quantity represents the Kullback-Leibler divergence [44] between the transition matrices and , analogous to a distance between probability distributions . Therefore , will vanish if and only if and coincide , i . e . if the transition probabilities between different “LFP” values of area do not depend on past “LFP” values of area . Conversely , this quantity will be strictly positive if these two transition matrices differ , i . e . if the past “LFP” values of area affect the evolution of the “LFP” in area . We also measure the causal unbalancing [93]: ( 9 ) which is normalized in the range . A value close to zero denotes symmetric causal influences in the two directions , while large absolute values of signal the emergence of asymmetric effective connectivity motifs . Considering now a structural motif with areas , equation ( 8 ) has to be modified in order to distinguish causal interactions which are direct ( e . g . toward ) from interactions which are indirect ( e . g . toward , but through ) . A solution allowing to assess only direct causal influences is partialization [42] , [71] . Indirect interactions from area to area involving a third intermediate area are filtered out by conditioning the transition matrices for the “LFP” activity of with resepect to the activity of the . Two conditional transition matrices , and , are then constructed and used to evaluate: ( 10 ) where the sum runs over all the four indices , , and . The effective connectivity in the panels C of Figures 3 , 4 and 5 is computed using pTE according to equation ( 10 ) . Absolute values of depend strongly on the time-lag and on the number of discrete levels . Nevertheless , we find that relative strengths of causal influences are qualitatively unchanged over broad ranges of parameters , as displayed in the Supporting Figure S1 . Furthermore the “plug-in” estimates of TE given by equations ( 8 ) and ( 10 ) suffer from finite-sampling biases , and a rigorous debiasing procedure is not yet known [43] . Therefore , for each value of and it is necessary to assess the significancy of the inferred causal interactions through comparison with suitably randomly resampled data [119] . To estimate the confidence intervals for the estimated TEs and the baseline for significancy we adopt a geometric bootstrap method [120] , guaranteed to generate resampled time-series with similar auto- and cross-correlation properties up to a certain lag . This is important , since “LFP” time-series have a strong oscillatory component , whose correlation structure has to be maintained under resampling . Each resampled time-series consists of a concatenation of blocks sampled from the original time-series . Each has the same length as the original . Every upward crossing , i . e . every time at which crosses from below its time-averaged value , is a potential start-time for a block . The first element of each block is obtained by selecting randomly one of these potential start-times . Then , the block consists of the oscillation cycles following the chosen start-time , where the random integer follows a geometric distribution , with and an average block length of ( we have taken oscillation cycles , longer than the mean correlation time for all the simulated “LFPs” ) . Randomly selected blocks are then concatenated up to the desired length . When considering a structural motif involving more areas , the “LFP” time-series of each area can be resampled jointly or independently . When resampling jointly , matching starting points and block-lengths are selected for each block of the resampled time-series of each area , leading to resampled multivariate time-series in which the structure of causal influences should not be altered . The distribution of over jointly resampled “LFP” time-series describes then for each directed pair of areas and the strength of the corresponding effective connectivity link , as well as the fluctuations of this strength . Conversely , when resampling independently the time-series , start-points and block-lengths of the resampled blocks are chosen independently for each area . This second procedure leads by construction to causally independent time-series . Any residual between directed pairs of independently resampled “LFPs” indicates therefore systematic biases , rather than actual causal influences . For each directed pairs of areas and , significance of the corresponding causal interaction can be assessed by comparing the bootstrapped distributions of and of . This comparison is performed in Figures 3 , 4 and 5 and in Supporting Figure S3D–E . Here , boxes indicate the median strength of for different directions and the corresponding confidence intervals , comprised between a lower extreme and and upper extreme , where and are respectively the first , the second and the third quartiles of the distribution of over jointly resampled time-series . Median values of and the corresponding confidence intervals , evaluated as before , are represented by horizontal dashed lines and a surrounding shaded band . When the distributions of and are not significantly different , a single baseline band is plotted . In this study , strengths and base-line for significancy of effective connectivity for each direction are validated based on , respectively , 500 jointly resampled and 500 independently resampled replicas . Note that geometric bootstrap can be applied to arbitrary signals , and does not depend on their strict periodicity . However it is precisely the strong periodic component of our signals that makes necessary the use of geometric bootstrap techniques . Indeed , conventional bootstrap , strongly disrupting signal periodicity , would lead to artificially low thresholds for statistical significance of TE ( not shown ) . We evaluate information transmission between pairs of mono-synaptically connected cells in different areas , linked by a TL-synapse ( TL pairs ) or by a normally weak long-range synapse ( control pairs ) . Inspired by [58] , spike trains are digitized into binary streams , where = 1 or 0 respectively when neuron fires or does not fire during the -th local oscillation cycle ( cycle counting is performed independently for each area and includes all the oscillation cycles following a common reference initial time ) . Note that neurons fire very sparsely and , due to the elevated degree of synchrony in our model , only in narrow temporal intervals centered around the peaks of the ongoing “LFP” oscillations . In particular , they fire at maximum once per oscillation cycle . Thus , this oscillatory spiking activity is naturally quantized in time and binning [58] is not required . For each considered directed pair of cells ( source cell , target cell ) , based on very long duration spike trains , we sample normalized histograms for three probability distributions: , and . When sampling the joint probability distribution we have to distinguish two cases: ( i ) If the presynaptic cell belongs to a leader area , i . e . the oscillation of the source area leads in phase over the oscillation of the target area of the considered synapse , then ; ( ii ) Conversely , if the presynaptic cell belongs to a laggard area , i . e . the oscillation of the target area leads in phase over the oscillation of the source area of the considered synapse , then . This means that we seek for spiking correlations only in pairs of spiking ( or missed spiking ) events in which the “effect” follows temporally its potential “cause” , since physical information transmission cannot occur backward in time . As for the estimation of TE ( see previous section ) , the probabilities , and are sampled separately for each specific phase-locking configuration of the ongoing “LFPs” . Epochs in which the system switches to a different phase-locking configuration , as well as transients following state switchings are dropped . The evaluation of these probabilities is thus based on disconnected spike train chunks , including overall oscillation cycles . Based on these probabilities , the Shannon entropy H of the spike train of the presynaptic neuron ( measuring the information content in its activity ) is evaluated as: ( 11 ) and MI between pre- and postsynaptic cells as: ( 12 ) MI is then normalized by the entropy of the pre-synaptic cell , in order to measure the relative efficiency of information transmission along each TL or control synapse . Statistics are taken over 400 pairs of cells per synapse set , i . e . one set of strong synapses per embedded TL , plus one set of ( control ) weak synapses . The box-plots in Figure 8C–D report median efficiencies of information transmission efficiencies ( for different active effective connectivities ) , as well as their confidence intervals , estimated non-parametrically from distribution quartiles , as discussed above for TE . Both MI and H are computed for ( finite ) spike trains of the largest available length . Following [58] , [121] , it is possible to correct these results for finite-size sampling bias ( see Supporting Figure S4 ) . MI and H are computed again , based on randomly selected shorter matching sections of the full length spike trains . The results of obtained for various shorter lengths are then plotted against the so-called inverse data fraction , where correspond then to estimations based on full length spike trains . Quadratic extrapolation to provides a debiased estimation of . Note that , in order to allow a more direct comparison with the non-debiased TE analysis , the results plotted in Figure 8C–D do not include any finite-size correction . As a matter of fact , as discussed in Supporting Figure S4 , finite size bias induces a small quantitative overestimation of information transmission efficiency ( from to ) , that does not affect qualitatively any of the results presented here .
|
The circuits of the brain must perform a daunting amount of functions . But how can “brain states” be flexibly controlled , given that anatomic inter-areal connections can be considered as fixed , on timescales relevant for behavior ? We hypothesize that , thanks to the nonlinear interaction between brain rhythms , even a simple circuit involving few brain areas can originate a multitude of effective circuits , associated with alternative functions selectable “on demand” . A distinction is usually made between structural connectivity , which describes actual synaptic connections , and effective connectivity , quantifying , beyond correlation , directed inter-areal causal influences . In our study , we measure effective connectivity based on time-series of neural activity generated by model inter-areal circuits . We find that “causality follows dynamics” . We show indeed that different effective networks correspond to different dynamical states associated to a same structural network ( in particular , different phase-locking patterns between local neuronal oscillations ) . We then find that “information follows causality” ( and thus , again , dynamics ) . We demonstrate that different effective networks give rise to alternative modalities of information routing between brain areas wired together in a fixed structural network . In particular , we show that the self-organization of interacting “analog” rate oscillations control the flow of “digital-like” information encoded in complex spiking patterns .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"physics",
"medicine",
"condensed-matter",
"physics",
"mathematics",
"neuroanatomy",
"computational",
"neuroscience",
"neurology",
"interdisciplinary",
"physics",
"neurological",
"disorders",
"biology",
"neuroscience",
"nonlinear",
"dynamics"
] |
2012
|
Dynamic Effective Connectivity of Inter-Areal Brain Circuits
|
Resistance ( R ) protein–associated pathways are well known to participate in defense against a variety of microbial pathogens . Salicylic acid ( SA ) and its associated proteinaceous signaling components , including enhanced disease susceptibility 1 ( EDS1 ) , non–race-specific disease resistance 1 ( NDR1 ) , phytoalexin deficient 4 ( PAD4 ) , senescence associated gene 101 ( SAG101 ) , and EDS5 , have been identified as components of resistance derived from many R proteins . Here , we show that EDS1 and SA fulfill redundant functions in defense signaling mediated by R proteins , which were thought to function independent of EDS1 and/or SA . Simultaneous mutations in EDS1 and the SA–synthesizing enzyme SID2 compromised hypersensitive response and/or resistance mediated by R proteins that contain coiled coil domains at their N-terminal ends . Furthermore , the expression of R genes and the associated defense signaling induced in response to a reduction in the level of oleic acid were also suppressed by compromising SA biosynthesis in the eds1 mutant background . The functional redundancy with SA was specific to EDS1 . Results presented here redefine our understanding of the roles of EDS1 and SA in plant defense .
Plants have evolved highly specific mechanisms to resist pathogens . One of the common ways to counter pathogen growth involves the deployment of resistant ( R ) proteins , which confer protection against specific races of pathogens carrying corresponding avirulence ( Avr ) genes [1] . Following recognition of the pathogen , one or more signal transduction pathways are induced in the host plant and these lead to the prevention of colonization by the pathogen . Induction of defense responses is often accompanied by localized cell death at the site of pathogen entry . This phenomenon , termed the hypersensitive response ( HR ) , is one of the earliest visible manifestations of induced defense reactions and resembles programmed cell death in animals [1]–[6] . Concurrent with HR development , defense reactions are triggered in both local and distant parts of the plant and accompanied by a local and systemic increase in endogenous salicylic acid ( SA ) levels and the upregulation of a large set of defense genes , including those encoding pathogenesis-related ( PR ) proteins [7]–[9] . The SA signal transduction pathway plays a key role in plant defense signaling ( see reviews in [10]–[12] ) . Arabidopsis mutants that are impaired in SA responsiveness , such as npr1 ( Nonexpressor of PR; [13]–[15] ) , or are defective in pathogen-induced SA accumulation , such as eds1 ( Enhanced Disease Susceptibility 1; [16] ) , eds5 ( Enhanced Disease Susceptibility 5; [17] ) , sid2 ( isochorishmate synthase; [18] ) and pad4 ( Phytoalexin Deficient 4; [19] ) , exhibit enhanced susceptibility to pathogen infection and show impaired PR gene expression . The EDS1 , EDS5 , PAD4 , NPR1 and SID2 proteins participate in both basal disease resistance to virulent pathogens as well as R protein-mediated resistance to avirulent pathogens [20] . Defense signaling mediated via a majority of R proteins , which contain Toll-interleukin1-like ( TIR ) domains at their N-terminal ends , is dependent on EDS1 [21] . Conversely , the NDR1 ( Non-race-specific Disease Resistance ) protein is required for many R proteins that contain coiled-coil ( CC ) domains at their N-terminal ends . However , several CC-nucleotide binding site ( NBS ) -leucine rich repeat ( LRR ) type of R proteins , including RPP8 , RPP13-Nd , HRT , and RPP7 , signal resistance via a pathway ( s ) that is independent of NDR1 [21] , [22]–[24] . Strikingly , the CC-NBS-LRR gene HRT , which confers resistance to Turnip Crinkle Virus ( TCV ) , is dependent on EDS1 [23] . Besides HRT , the only other CC domain-containing R protein that utilizes an EDS1-dependent pathway is RPW8 , which confers broad-spectrum resistance to powdery mildew [25] . However , RPW8 is not a typical NBS-LRR type of R protein; it contains an N-terminal transmembrane domain in addition to the CC domain . Although several components contributing to resistance against pathogens have been identified , the molecular signaling underlying R gene-mediated resistance still remains obscure . Furthermore , potential relationship ( s ) among different downstream components and how they relay information leading to resistance remains unknown . The EDS1 and PAD4 proteins are structurally related to lipase/esterase-like proteins although their lipase-like biochemical functions have not been demonstrated [16] , [19] . EDS1 interacts with PAD4 and SAG ( senescence associated gene ) 101 and the combined activities of these proteins are required for HR formation and to restrict the growth of virulent bacterial strains [26] . PAD4 and SAG101 also restrict the post-invasive growth of non-pathogenic fungi in Arabidopsis [27] . In addition to the major phytohormone-mediated defense pathways , fatty acid ( FA ) -derived signals have emerged as important mediators of defense signaling [28]–[35] . The Arabidopsis SSI2/FAB2-encoded stearoyl-acyl carrier protein-desaturase ( SACPD ) converts stearic acid ( 18∶0 ) to oleic acid ( 18∶1 ) . A mutation in SSI2 results in the accumulation of 18∶0 and a reduction in 18∶1 levels . The mutant plants show stunting , spontaneous lesion formation , constitutive PR gene expression , and enhanced resistance to bacterial and oomycete pathogens [29] , [36] . Characterization of ssi2 suppressor mutants has shown that the altered defense-related phenotypes are the result of the reduction in the levels of the unsaturated FA , 18∶1 [30] , [31] , [35] , [37]–[40] . The altered defense-related phenotypes in ssi2 plants can be rescued by restoring the 18∶1 levels via second site mutations in genes encoding a glycerol-3-phosphate ( G3P ) acyltransferase [ACT1 , 30] , a G3P dehydrogenase [GLY1 , 31] , and an acyl carrier protein [ACP4 , 35] . A mutation in act1 disrupts the acylation of G3P with 18∶1 resulting in the increased accumulation of 18∶1 , thereby restoring wild-type ( wt ) phenotypes in ssi2 plants . ACT1 preferentially utilizes 18∶1 conjugated to the ACP4 isoform in Arabidopsis [35] . Thus , a mutation in acp4 produces similar phenotypes as the act1 mutant and suppresses ssi2-mediated signaling by increasing 18∶1 levels [35] . A mutation in GLY1 also restores 18∶1 levels in ssi2 gly1 plants because it disrupts the formation of G3P from dihydroxyacetone phosphate [31] . Reduced availability of G3P in turn impairs the ACT1-catalyzed reaction resulting in accumulation of 18∶1 in ssi2 gly1 plants . Concurrently , increasing the endogenous G3P levels via exogenous application of glycerol reduces 18∶1 levels and induces ssi2-like phenotypes in wt plants [31] , [40] . This effect of glycerol is highly specific because ssi2-associated phenotypes are not induced upon glycerol treatment of act1 ( defective in the acylation of G3P with 18∶1 ) or gli1 ( defective in the phosphorylation of glycerol to G3P ) mutants [40] . Recently , we showed that a reduction in 18∶1 levels upregulates the expression of several R genes in an SA-independent manner [37] . Furthermore , we showed that pathogen resistance induced via this mode bypasses the requirement for components that are normally required for signaling downstream of R protein activation . For example , resistance to TCV mediated by the R gene HRT ( HR to TCV ) , requires the recessive locus rrt ( regulates resistance to TCV ) , SA , EDS1 and PAD4 [23] . Exogenous application of SA induces the expression of HRT and overcomes the requirement for rrt . However , exogenous SA is unable to induce HRT or confer resistance in pad4 background [23] . Interestingly , even though a reduction in 18∶1 levels also upregulates HRT expression to confer resistance to TCV , this mode of resistance is independent of PAD4 , SA , EDS1 and EDS5 , which are required for HRT-mediated resistance to TCV [37] . Remarkably , induction of R genes in response to reduced 18∶1 is conserved in plants as diverse as Arabidopsis and soybean [41] . Furthermore , this low 18∶1-mediated induction of defense responses was also demonstrated in rice recently [42] . Together , these studies strengthen the conserved role of 18∶1 in plant defense signaling . Here , we show that R gene expression induced in response to a reduction in 18∶1 levels and the associated defense signaling can be suppressed by simultaneous mutations in EDS1 and the genes governing synthesis of SA . We also show that EDS1 and SA function redundantly in R gene-mediated resistance against bacterial , viral and oomycete pathogens and that EDS1 also regulates signaling mediated by CC domain containing R proteins .
Signaling mediated by many R genes is known to require EDS1 and/or NDR1 . Previously , we have shown that ssi2 eds1 plants continue to express R genes at high levels , including those that are dependent on EDS1 for their signaling [37] . To determine if NDR1 played a role in ssi2-triggered phenotypes , we generated ssi2 ndr1 plants . The double-recessive plants segregated in a Mendelian fashion and all ssi2 ndr1 plants showed ssi2-like morphology in the F2 , F3 and F4 generations ( Figure 1A; Table S1 ) . Although the ssi2 ndr1 plants accumulated significantly less SA/SAG ( Figure 1C ) , compared to ssi2 plants , they showed ssi2-like PR-1 and R gene expression ( Figure 1D and 1E , Figure S1A ) . Exogenous glycerol application , which reduces 18∶1 levels , also induced R gene expression in eds1 and ndr1 plants ( data not shown ) . Together , these results suggest that R gene expression induced by low 18∶1 levels does not require EDS1 or NDR1 . The SA/SAG levels in ssi2 eds1 and ssi2 ndr1 plants were significantly higher compared to those in wt plants ( Figure 1C ) . To determine whether high SA in these genotypes was responsible for increased R gene expression , we generated ssi2 eds1 sid2 and ssi2 ndr1 sid2 plants . Interestingly , only the ssi2 eds1 sid2 plants showed wt-like morphology and did not develop visible or microscopic cell death ( Figure 1A and 1B ) . In contrast , ssi2 sid2 , ssi2 ndr1 , ssi2 ndr1 sid2 or ssi2 eds1 plants exhibited ssi2-like phenotypes . PR-1 gene expression was restored to wt-like levels in the ssi2 eds1 sid2 and ssi2 ndr1 sid2 plants , due to the sid2-derived reduction in SA levels ( Figure 1D ) . In contrast , expression of the SA-independent PR-2 gene was restored to basal levels only in ssi2 eds1 sid2 [43] , but not in ssi2 sid2 or ssi2 ndr1 sid2 plants ( Figure 1D , Table S2 ) . Most importantly , ssi2 eds1 sid2 showed basal expression of R genes , unlike ssi2 ndr1 sid2 plants ( Figure 1E and 1F; Figure S1A , S1B; Table S1 ) . R gene induction was further confirmed by comparing the transcript profiles of 162 NBS-LRR genes in ssi2 sid2 with that of wt plants using Affymetrix ATH1 GeneChips arrays . Twenty-one NB-LRR genes were specifically expressed at 2-fold or higher levels in ssi2 sid2 plants as compared to wt ( Col-0 ) or eds1 plants ( P<0 . 05 ) ( Table S2 ) . All 21 NB-LRR genes were expressed at low levels in ssi2 eds1 sid2 plants , further confirming the results from the RT-PCR analysis . Transcriptional profiling performed using Affymetrix arrays showed that the induction of several R genes ( RPM1 , RPS2 , RPP5 , RPS4 ) was lower than 2-fold in ssi2 or ssi2 sid2 compared to wt plants ( Table S2 , data not shown for ssi2 ) . To determine if this low-level induction translated to a significant increase in R protein levels , we analyzed the levels of RPM1 in ssi2 plants . Indeed , ssi2 plants accumulated significantly higher levels of the RPM1-Myc protein ( Figure 1G ) . To rule out the effects of the varied ecotypes of the ssi2 sid2 eds1 ( Nössen , Col-0 , Ler ) plants we introduced eds1-1 ( Ws-0 ecotype ) and eds1-2 ( Ler ecotype ) alleles in ssi2 sid2 and ssi2 nahG ( Nössen ecotype ) backgrounds ( Table S1 ) . All combinations of ssi2 with eds1-1/eds1-2 and sid2/nahG produced similar phenotypes ( data not shown ) . FA profiling showed that the ssi2 eds1 sid2 plants contained low 18∶1 levels , similar to ssi2 plants ( Table S3 ) . We thus concluded that EDS1 and SA function downstream of 18∶1 levels , but upstream of R gene expression . Furthermore , ssi2 eds1 sid2 plants were wt-like , even though neither ssi2 eds1 nor ssi2 sid2 were restored for defense signaling . Therefore , EDS1 and SA likely fulfill redundant functions in defense signaling induced in response to a reduction in 18∶1 levels . To further test the redundancy for EDS1 and SA , ssi2 eds1 sid2 plants were treated with SA or its active analog benzo ( 1 , 2 , 3 ) thiadiazole-7-carbothioic acid ( BTH ) . Application of SA or BTH induced lesion formation on ssi2 eds1 sid2 plants but not on wt , eds1 , sid2 , eds1 sid2 or EDS1 SID2 F2 plants ( Figure 2A and 2B , data not shown for eds1 sid2 and EDS1 SID2 ) . Also , application of SA or BTH induced R gene expression in ssi2 eds1 sid2 plants ( Figure 2C ) . Thus , application of SA restored ssi2-like phenotypes in ssi2 eds1 sid2 plants . Since glycerol application mimics the effects of the ssi2 mutation , we generated eds1 sid2 plants and evaluated them for their ability to induce R genes in response to glycerol . Exogenous application of glycerol lowered 18∶1 levels in all genotypes , but induced the expression of R genes only in wt , eds1 , sid2 and EDS1 SID2 F2 plants ( Figure 2D , Figure S1C ) . Only a marginal or no increase in R gene expression was observed in the eds1 sid2 plants ( Figure 2D ) . These results confirmed that EDS1 and SA function redundantly downstream of signaling induced by low 18∶1 levels , but upstream of R gene expression . We next evaluated the effect of simultaneous mutations in EDS1- and SA-signaling pathways on resistance to TCV in the ssi2 background . We reported previously that resistance to TCV is dependent on the R gene , HRT , and a recessive locus rrt [23] . However , the ssi2 mutation overcomes the requirement for rrt in HRT-containing plants [23] , [37] . Furthermore , the ssi2 mutation only confers resistance to TCV when HRT is present ( Figure 3A ) . The ssi2 mutation also overrides a requirement for EDS1 and SA and consequently ssi2 HRT eds1 as well as ssi2 HRT sid2 plants exhibit resistance to TCV [37] ( Figure 3A ) . Unlike HRT ssi2 , HRT ssi2 eds1 or HRT ssi2 sid2 plants , the HRT ssi2 eds1 sid2 plants showed susceptibility to TCV; ∼85% HRT ssi2 eds1 sid2 plants were susceptible to TCV as against ∼2–4% of HRT ssi2 sid2 or HRT ssi2 eds1 plants ( Figure 3A ) . TCV-induced expression of PR-1 is also independent of EDS1 and SA . However , TCV inoculation failed to induce PR-1 expression in HRT ssi2 eds1 sid2 plants , unlike in HRT ssi2 sid2 plants ( Figure 3B ) . These results showed that both EDS1 and SA have redundant functions in ssi2-mediated resistance to TCV in HRT plants . To determine the redundancy of EDS1 and SA in signaling mediated by CC-NBS-LRR R proteins , we tested the effects of mutations in EDS1- and/or SID2 on HR to TCV . Earlier , we showed that HRT-mediated HR to TCV and PR-1 gene expression is not affected by mutations in the EDS1 or SID2 genes [23] . Consistent with previous results , Di-17 ( HRT-containing resistant ecotype ) , HRT sid2 and HRT eds1 plants revealed discrete and similar-sized HR lesions on TCV-inoculated leaves ( Figure 3C and 3D ) . In comparison , HR in HRT eds1 sid2 plants was diffused and formed larger lesions ( Figure 3C and 3D ) . Increased lesion size in HRT eds1 sid2 plants correlated with increased accumulation of the TCV coat protein ( CP ) and TCV CP transcript ( Figure 3E and 3F ) . Analysis of PR-1 and PR-2 gene expression indicated that TCV-inoculated HRT eds1 sid2 plants accumulated lower levels of PR-1 and PR-2 transcripts , unlike Di-17 , HRT eds1 or HRT sid2 plants ( Figure 3G and 3H ) . In contrast to PR , HRT expression remained unaltered in HRT eds1 sid2 plants ( Figure 3H ) . Together , these results suggested that EDS1 and SA function redundantly in HRT-mediated signaling leading to HR formation and expression of PR-1 . The functional redundancy with SA was specific to EDS1 and did not extend to PAD4; HRT pad4 sid2 plants showed normal replication of the virus and wt-like HR and PR-1 gene expression ( Figure 3C–3G ) . A majority of CC-domain containing R proteins , including RPS2 , have been reported as not requiring EDS1 for resistance signaling [21] . To determine the effect of simultaneous mutations in EDS1 and SID2 on RPS2-mediated resistance , we compared defense phenotypes produced in single or double mutant plants with that of plants lacking a functional RPS2 gene . Since different alleles of RPS2 confer varying levels of resistance to Pseudomonas syringae ( containing AvrRPT2 ) [44] , we screened and isolated an EDS1 knockout ( KO ) mutant ( designated eds1-22 ) in the Col-0 background and crossed it into the sid2 background ( Col-0 ecotype ) . Inoculation with P . syringae expressing AvrRPT2 induced severe chlorosis on eds1-22 sid2 leaves ( Figure 4A ) . Similar results were obtained when P . syringae expressing AvrRPT2 was inoculated into eds1-1 sid2 double mutant plants ( Figure S2A ) . Interestingly , these phenotypes were very similar to those produced on plants lacking a functional RPS2 ( rsp2-101c ) , while eds1 and sid2 showed no or very mild symptoms , respectively ( Figure 4A , Figure S2A ) . The appearance of symptoms correlated with bacterial growth; eds1-22 sid2 plants and the rps2 mutant supported maximum growth of the pathogen , followed by sid2 plants ( Figure 4B ) . Similarly , the eds1-1 sid2 double mutant plants supported more pathogen growth compared to eds1-1 or sid2 plants ( data not shown ) . Together , these data suggest that the simultaneous loss of EDS1- and SA-dependent signals is required to mimic a phenotype produced by the loss of the cognate R gene , RPS2 . To determine if the loss of both EDS1- and SA-dependent signaling impaired resistance by affecting the RPS2 protein , we analyzed R protein levels in eds1-1 and sid2 single and eds1-1 sid2 double mutant plants . Analysis of RPS2 tagged with HA epitope at various times did not detect any significant changes in RPS2 levels in response to inoculation with P . syringae expressing AvrRPT2 ( Figure 4C ) . Therefore , RPS2 levels in mutant plants were analyzed at only 12 and 24 h post-pathogen inoculation . The RPS2-HA levels in eds1-1 , sid2 or eds1-1 sid2 plants were similar to that in wt plants ( Figure 4D ) . These results suggested that abrogation of resistance in eds1 sid2 double mutants was not due to a defect in the accumulation of the R protein . We next evaluated the effects of mutations in EDS1 and SID2 on RPP8-mediated resistance to Hyalopernospora arabidopsidis biotype Emco5 encoding Atr8 . RPP8 ( encodes a CC-NBS-LRR type R protein ) -mediated resistance signaling was previously reported to be independent of both EDS1 and SA [21] , [24] . As expected , RPP8 plants ( ecotype Ler ) inoculated with the Emco5 isolate showed localized HR and did not support growth of the pathogen ( Figure 5A ) . Consistent with earlier reports [21] , [24] , RPP8 eds1-2 plants also did not support the growth of Emco5 , although they did develop trailing necrosis ( Figure 5A and 5B ) . The presence of the nahG transgene did not alter HR formation or pathogen response in the RPP8 nahG plants ( Ler ecotype ) . In contrast , eds1-2 nahG plants were affected in both HR as well as resistance; eds1-2 nahG plants not only showed extensive trailing necrosis but also supported growth and sporulation of the pathogen ( Figure 5A–5C ) . Although RPP8 EDS1 nahG and RPP8 eds1-2 nahG plants showed contrasting phenotypes ( Figure 5A–5C ) , we still wanted to rule out the possibility that susceptibility of eds1 nahG plants was not due to the accumulation of catechol , which is formed upon degradation of SA by NAHG . Estimation of SA levels in Emco5 inoculated RPP8 ( Ler ) plants showed marginal increase in SA and no significant increase in SAG levels compared to mock-inoculated plants ( data not shown ) . This suggests that Emco5 inoculated nahG plants are unlikely to show a significant increase in catechol levels . In addition to this , we tested two independent lines of RPP8 eds1-2 sid2 ( in the ssi2 background ) plants and both showed increased susceptibility to Emco5 ( Figure 5D ) . In comparison , RPP8 eds1-2 or RPP8 sid2 genotypes did not support any growth or sporulation of the pathogen ( Figure 5D ) . Taken together , these results show that EDS1 and SA have redundant functions in RPP8-mediated resistance to H . arabidopsidis Emco5 . To determine the relation between EDS1- and SA-derived signaling , we compared PR-1 gene expression and resistance in plants that were either overexpressing EDS1 or were pretreated with SA . EDS1 overexpression was achieved by expressing EDS1 ( At3g48090 from the Col-0 ecotype ) under control of the CaMV 35S promoter in Col-0 plants ( Figure 6A ) . The 35S-EDS1 plants analyzed in the T2 and T3 generations showed wt-like morphology ( data not shown ) , wt-like expression of the PR-1 gene ( Figure 6A ) and accumulated wt-like levels of SA/SAG ( data not shown ) . In comparison , exogenous application of SA induced PR-1 and EDS1 gene expression [data not shown; 16] . Analysis of RPS4 ( encodes a TIR-NBS-LRR type R protein ) -mediated resistance showed that exogenous application of SA enhanced resistance to P . syringae ( expressing AvrRPS4 ) in wt as well as eds1-22 plants , although wt plants were more resistant to AvrRPS4 bacteria than the eds1-22 plants ( Figure 6B ) . Overexpression of EDS1 , on the other hand , did not alter the response to AvrRPS4 bacteria . Strikingly , exogenous application of SA on 35S-EDS1 plants enhanced resistance even more than in the SA-treated wt or eds1-22 plants . Together , these results suggest that EDS1- and SA-derived signaling contribute additively towards pathogen resistance . We next evaluated the effect of the eds1 sid2 mutations on basal resistance to virulent P . syringae , since both EDS1 and SID2 are known to contribute to basal defense as well . The eds1-1 , eds1-22 , sid2 and eds1 sid2 plants all showed enhanced susceptibility to virulent bacteria as compared to the respective wt ecotypes ( Figure 7A ) . Interestingly , unlike in the case of the avirulent bacteria , growth of virulent bacteria was similar in eds1 sid2 double mutant plants as compared to that in eds1 or sid2 single mutant plants . These results suggested that loss-of-function mutations in EDS1 and SID2 do not additively reduce basal resistance to virulent P . syringae . Similar to the results obtained with the bacterial pathogen , the loss of both EDS1- and SA-dependent signals did not additively lower basal resistance to TCV either ( Figure 7B ) . This further suggested that the redundant functions of EDS1 and SA might be relevant only for R gene-mediated signaling . Besides SID2 , mutations in FAD7 FAD8 , which catalyze desaturation of 18∶2 to 18∶3 on membrane glycerolipids , also lower the SA levels in ssi2 plants [40] . To test if fad7 or fad7 fad8 mutations produced a similar effect as sid2 , these mutations were mobilized into the ssi2 eds1 background . The ssi2 eds1 fad7 and ssi2 eds1 fad7 fad8 plants were bigger in size compared to ssi2 fad7 or ssi2 fad7 fad8 plants ( Figure S3A ) . The ssi2 eds1 fad7 fad8 were wt-like in morphology and showed no or greatly reduced cell death lesions ( Figure S3A , S3B ) . PR-1 expression was greatly reduced or abolished in ssi2 eds1 fad7 and ssi2 eds1 fad7 fad8 plants , respectively ( Figure S3C ) and correlated with their endogenous SA/SAG levels; the ssi2 eds1 fad7 and ssi2 eds1 fad7 fad8 plants showed greatly reduced or basal levels of SA and SAG , respectively ( Figure S3D , S3E ) . Expression of some R genes ( SSI4 , RPS2 , RPP5 ) was nominally or moderately reduced in ssi2 eds1 fad7 plants ( Figure S3D , S3E ) . By comparison , all R genes tested were expressed at basal levels in ssi2 eds1 fad7 fad8 plants ( Figure S3F ) . These results showed that presence of fad7 fad8 mutations restored the altered defense phenotypes of ssi2 eds1 plants . FA profiling did not detect any significant increase in 18∶1 levels in ssi2 eds1 fad7 and ssi2 eds1 fad7 fad8 plants , compared to ssi2 fad7 and ssi2 fad7 fad8 , respectively ( Table S4 ) . This suggested that restoration of defense phenotypes in ssi2 eds1 fad7 fad8 was not the result of restored 18∶1 levels , but rather the reduction of SA levels in the eds1 background . Mutations in EDS5 and PAD4 also lower SA/SAG levels in ssi2 plants [40] . To determine if mutations in these can substitute for sid2 triple mutants containing ssi2 eds1 pad4 and ssi2 eds1 eds5 were generated . The ssi2 eds1 pad4 plants were morphologically similar to ssi2 eds1 or ssi2 pad4 plants and showed spontaneous cell death and increased expression of PR-1 gene ( Figure 8A–8C ) . In comparison , ssi2 eds1 eds5 showed wt-like morphology , greatly reduced cell death and basal expression of PR-1 gene ( Figure 8A–8C ) . Quantification of endogenous SA levels showed that both ssi2 eds1 eds5 and ssi2 eds1 pad4 accumulated lower SA/SAG levels compared to ssi2 eds5 and ssi2 pad4 , respectively ( Figure 8D and 8E ) . However , while ssi2 eds1 eds5 plants accumulated basal levels of SA/SAG , the ssi2 eds1 pad4 accumulated significantly higher levels of SA/SAG compared to wt , ssi2 sid2 and ssi2 eds1 eds5 plants ( Figure 8D and 8E ) . Analysis of R gene expression showed greatly reduced levels in ssi2 eds1 eds5 plants but the ssi2 eds1 pad4 expressed ssi2-like levels of R genes ( Figure 8F , Figure S1D ) . Taken together , these results suggest that the suppression of SA levels was required for the normalization of defense phenotypes in the ssi2 eds1 background . Besides EDS1 , the SA signaling pathway is also regulated by PAD4 and EDS5 and via the physical association of EDS1 with SAG101 and PAD4 [17] , [19] , [45] . To determine if PAD4 , SAG101 or EDS5 also function redundantly with SA , we introduced the pad4 , sag101 and eds5 mutations in the ssi2 and ssi2 sid2 backgrounds . The ssi2 sag101 , ssi2 pad4 and ssi2 eds5 plants showed ssi2-like morphology , visible and microscopic cell death and constitutive PR-1 gene expression ( Figure S4A , S4B , S4C and Figure S5A , S5B , S5C ) . Consistent with these phenotypes , the ssi2 sag101 , ssi2 pad4 , ssi2 eds5 plants showed increased expression of R genes ( Figure S4D and Figure S5D ) and accumulated elevated levels of SA and SAG ( Figure S4E , S4F and Figure S5E , S5F ) . Notably , the SA levels in ssi2 sag101 plants were ∼6-fold lower than in ssi2 plants , suggesting that SAG101 contributed to the accumulation of SA in ssi2 plants . To determine if the reduced SA in the sag101 background could restore wt-like phenotypes in ssi2 eds1 plants , triple mutant ssi2 eds1 sag101 plants were generated . Although the ssi2 eds1 sag101 plants accumulated significantly lower levels of SA/SAG ( Figure S4E , S4F ) , these plants were only slightly bigger than ssi2 eds1 or ssi2 sid2 plants ( Figure S4A ) , showed spontaneous cell death ( Figure S4B ) and expressed PR-1 ( Figure S4C ) and R genes constitutively ( Figure S4D ) . We next analyzed the triple mutant ssi2 sag101 sid2 , ssi2 pad4 sid2 and ssi2 eds5 sid2 plants . All the triple mutants contained wt-like levels of SA and SAG ( Figure S4E , S4F and Figure S5E , S5F ) . The ssi2 sag101 sid2 plants were morphologically similar to ssi2 plants , showed spontaneous cell death and expressed R genes constitutively ( Figure S4A , S4B , S4C , S4D ) . In comparison , the ssi2 pad4 sid2 and ssi2 eds5 sid2 plants were bigger in morphology . However , plants of both genotypes showed cell death ( Figure S5A , S5B ) and expressed R genes constitutively ( Figure S5D ) . Together , these data suggest that the functional redundancy with SA was specific only to EDS1 and did not extend to PAD4 , SAG101 or EDS5 .
SA is long known as an essential modulator of R gene-derived signaling in pathogen defense . Several molecular components , including EDS1 , have been identified as essential effectors of SA-derived signaling [23] , [26] , [45] . Since SA upregulates expression of EDS1 , both SA and EDS1 are thought to function in a positive feedback loop and EDS1 is widely considered an upstream effector of SA [16] , [19] , [23] , [45] . Recent data has shown that EDS1 signals resistance via both SA-dependent as well as SA-independent pathways [46] . Strikingly , EDS1-dependent but SA-independent branch of EDS1 pathway still requires SA pathway for full expression of resistance [46] . In this study , we have characterized the relationship between EDS1 and SA . We show that the two components act in a redundant , and not necessarily sequential manner to regulate R gene expression induced in response to a reduction in the levels of the FA 18∶1 . Furthermore , EDS1 and SA also function redundantly in R gene-mediated defense against viral , bacterial and oomycete pathogens . It appears that the redundant functions of EDS1 and SA may have prevented their identification as required components for signaling mediated by CC-NBS-LRR R proteins . Indeed , RPS2-mediated signaling is fully compromised only in eds1 sid2 and not in the single mutant plants . Similarly , HRT-mediated signaling leading to HR formation and PR-1 gene expression is only affected in eds1 sid2 plants , while eds1 or sid2 plants behave similar to wt plants . Furthermore , RPP8-mediated resistance , which was previously reported not to require EDS1 or SA [21] , [24] , is compromised in plants lacking both EDS1 and SA . In contrast to their effect on R gene-mediated resistance , loss of both EDS1- and SA-dependent signals did not additively lower basal resistance to P . syringae or TCV . Together , these data suggests that the redundant functions of EDS1 and SA might be relevant only for R gene-mediated signaling . In contrast to SA application , overexpression of EDS1 was unable to confer increased resistance to the avirulent pathogen P . syringae . Furthermore , unlike SA , overexpression of EDS1 was not associated with the induction of PR-1 gene expression . These findings , together with the observation that SA was able to induce EDS1 expression and that SA application on wt plants resulted in higher resistance than that in eds1 , suggests that SA feedback regulates EDS1-derived signaling in a unidirectional manner ( Figure 9B ) . Thus , SA application induces both SA- and EDS1-derived signaling , the additive effects of which enhance resistance in wt plants much more than in eds1-22 plants . Furthermore , the combined effects of SA pretreatment and EDS1 overexpression induced much better resistance than the individual effects of each . This is consistent with a previous report that 35S-EDS1 plants induce rapid and stronger expression of PR-1 in response to pathogen inoculation [47] . The additive effects of EDS1 and SA was also supported by the observation that eds1 sid2 plants showed pronounced chlorosis upon inoculation with AvrRPS4 expressing pathogen , which is recognized by a TIR-NBS-LRR protein RPS4 ( Figure S2B ) . Since mutations in SA-independent branch of EDS1 pathway and sid2 have additive effects on R gene-mediated resistance [46] , it is possible that overexpression of EDS1 triggers signaling via both SA-dependent and/or -independent branches of EDS1 pathway . Although the Col-0 ecotype is thought to contain two functional alleles of EDS1 [26] , a KO mutation in At3g48090 was sufficient to compromise both basal and R gene ( RPS4 ) -mediated resistance . However , the Col-0 eds1-22 mutant consistently supported less growth of virulent or avirulent pathogens compared to eds1-1 or eds1-2 plants . This suggests that the second EDS1 allele in the Col-0 ecotype might also contribute towards the resistance response . This is consistent with another study where constitutive defense phenotypes due to the overexpression of the SNC1 gene , encoding a TIR-NBS-LRR R protein , are not completely suppressed by a mutation in eds1 in the Col-0 background but restored by the eds1 mutation in the Ws background [48] . The inability to accumulate SA together with a mutation in EDS1 was also required to suppress constitutive defense signaling resulting from the overexpression of R genes induced in response to reduced 18∶1 levels . Although eds1 or sid2 plants were entirely competent in inducing R gene expression in response to a reduction in 18∶1 , eds1 sid2 plants were not . Thus , ssi2 eds1 sid2 as well as glycerol-treated eds1 sid2 plants showed wt-like expression of R genes while ssi2 eds1 , ssi2 sid2 and glycerol-treated eds1 or sid2 plants showed increased expression of R genes . Moreover , treatment of ssi2 eds1 sid2 plants with exogenous SA restored R transcript induction and cell death in these plants . The fact that glycerol treatment is unable to induce R gene expression in eds1 sid2 plants supports the possibility that EDS1 and SA function upstream of , and not merely serve as a feedback loop in , R gene induction . Signaling induced by low 18∶1 levels continues to function in the absence of SA , suggesting a novel SA-independent role for EDS1 in defense signaling . Since ssi2 eds1 sid2 plants contain a mixed ecotypic background ( Nö , Ws/Ler , Col-0 , ecotypes ) , it is possible that ecotypic variations in various genetic backgrounds resulted in the restoration of ssi2-triggered defense phenotypes . Indeed , phenotypic variations amongst different Arabidopsis ecotypes have been associated with many physiological processes [48]–[51] . Moreover , certain alleles can express themselves only in specific ecotypic backgrounds [48] , [51] . However , since ssi2 EDS1 SID2 , ssi2 EDS1 sid2 or ssi2 eds1 SID2 plants ( F2 population ) always exhibited ssi2-like phenotypes , it is highly unlikely that ecotypic variations resulted in the restoration of phenotypes in ssi2 eds1 sid2 plants . The effect of ecotypic variations on the observed phenotypes can be further ruled out for the following reasons . First , the effects of different mutations were assessed in multiple backgrounds . For example , we used both eds1-1 ( Ws-0 ecotype ) and eds1-2 ( Ler ecotype ) alleles in ssi2 sid2 ( Nö , Col-0 ecotypes ) and ssi2 nahG ( Nö ecotype ) backgrounds and all combinations of ssi2 with eds1-1/eds1-2 and sid2/nahG produced similar phenotypes ( Table S1 ) . Second , all defense phenotypes were assessed over three generations using multiple progeny . Third , similar results were obtained when different ecotypic backgrounds were evaluated for their response to different pathogens . For example , eds1 nahG or eds1 sid2 backgrounds conferred increased susceptibility to H . arabidopsidis , P . syringae and TCV , even though only the genotypes used for TCV were of mixed ecotypic backgrounds . Fourth , F2 plants containing wild-type alleles behaved like wild-type parents . Finally , the effects of various mutant backgrounds on ssi2 phenotypes were also confirmed by glycerol application on individual mutants . Although glycerol treatment failed to induce R gene expression in eds1 sid2 plants , it did induce cell death . This is in contrast to the absence of a cell death phenotype in ssi2 eds1 sid2 leaves . One possibility is that the glycerol-triggered cell death is not due to a reduction in 18∶1 levels . However , significant overlap between ssi2- and exogenous glycerol-triggered signaling pathways lessens such a possibility [40] . An alternate possibility is that , while EDS1 affects a majority of the responses induced by low 18∶1 levels , the cell death phenotype is also governed by some additional molecular factor ( s ) . This is supported by the fact that ssi2 pad4 sid2 plants exhibit improved morphology and reduced cell death even though they are not restored for other defense-related phenotypes . Since the overexpression of R genes can initiate defense signaling in the absence of a pathogen [48] , [52] , it is possible that the induced defense responses in ssi2 plants are the result of increased R gene expression . This idea is supported by the fact that ssi2-related phenotypes can be normalized by restoring R gene expression to wt-like levels , irrespective of their 18∶1 levels . Thus , wt-like defense phenotypes are restored in suppressors containing high 18∶1 levels , such as ssi2 act1 , ssi2 gly1 or ssi2 acp4 [30] , [31] , [35] , as well as in suppressor containing low 18∶1 levels , such as ssi2 eds1 sid2 ( this work ) and restored in defective crosstalk ( rdc ) 2 ( unpublished data ) ( Figure 9A ) . We have also characterized additional ssi2 suppressors that show wt-like phenotypes even though they contain low 18∶1 levels and express R genes constitutively ( rdc3 , rdc4 ) . Together , these results suggest that the ssi2-associated phenotypes can be restored by normalizing R gene expression to wt-like levels either by increasing 18∶1 levels , impairing factors downstream of signaling induced by low 18∶1 levels , or impairing events downstream of R gene expression induced by low 18∶1 levels . In addition to 18∶1 levels or R gene expression , ssi2-related defense signaling could also be normalized by altering some factor ( s ) that function downstream of R gene induction . Indeed , our preliminary characterizations have identified additional ssi2 suppressors that yield wt-like phenotypes with regards to defense signaling but continue to express R genes at high levels . Reduced 18∶1 levels may induce defense signaling by directly regulating the transcription of activators or suppressors of defense gene expression . This is supported by the fact that 18∶1-mediated activation of a transcription factor induces the expression of genes required for neuronal differentiation [53] . Similarly , in Sacharromyces cerevisiae as well as mammalian cells , binding of 18∶1 to specific transcription factors induces the transcription of genes carrying 18∶1 responsive elements in their promoters [54] , [55] . On the other hand , expression of the oncogene HER2 is inhibited via the 18∶1-upregulated expression of its transcriptional repressor [56] . Reduced 18∶1 might also directly activate/inhibit/alter protein activities . For example , 18∶1 is known to activate the Arabidopsis phospholipase D [57] and inhibit glucose-6-phosphate transporter activity in Brassica embryos [58] . Indeed , we have also identified several Arabidopsis proteins for which enzymatic activities are inhibited upon binding to 18∶1 ( unpublished data ) . In conclusion , results presented here redefine the currently accepted pathway for SA-mediated signaling by showing that EDS1 and SA play a redundant role in plant defense mediated by R proteins and in signaling induced by low 18∶1 fatty acid levels . Further biochemical characterization should help determine if 18∶1 binds to EDS1 and if cellular levels of 18∶1 modulate the as yet undetected lipase activity of EDS1 .
Plants were grown in MTPS 144 Conviron ( Winnipeg , MB , Canada ) walk-in-chambers at 22°C , 65% relative humidity and 14 hour photoperiod . The photon flux density of the day period was 106 . 9 µmoles m−2 s−1 and was measured using a digital light meter ( Phytotronic Inc , Earth city , MO ) . All crosses were performed by emasculating the flowers of the recipient genotype and pollinatng with the pollen from the donor . All the genotypes and crosses analyzed in this work , their genetic background and number of single , double , or triple mutant plants studied are listed in Table S1 . In most cases , single , double , or triple mutant plants were obtained from more than one combination of crosses and showed similar morphological , molecular and biochemical phenotypes . F2 plants showing the wt genotype at the mutant locus were used as controls in all experiments . The wt and mutant alleles were identified by PCR , CAPS , or dCAPS analysis and/or based on the FA profile [30] , [31] , [38] , [40] . The EDS1 KO mutant in At3g48090 was , isolated by screening SALK_071051 insertion line , obtained from ABRC . The EDS1 KO was designated eds1-22 , based on the previous designation assigned to SALK_071051 T-DNA KO line [48] . The At3g48090 gene showed 98 . 8% identity at amino acid level to EDS1 allele from Ler ecotype . The homozygous insertion lines were verified by sequencing PCR products obtained with primers specific for the T-DNA left border in combination with an EDS1-specific primer . The eds1-22 lines did not show any detectable expression of EDS1 . Small-scale extraction of RNA from one or two leaves was performed with the TRIzol reagent ( Invitrogen , CA ) , following the manufacturer's instructions . Northern blot analysis and synthesis of random-primed probes for PR-1 and PR-2 were carried out as described previously [29] . RNA quality and concentration were determined by gel electrophoresis and determination of A260 . Reverse transcription ( RT ) and first strand cDNA synthesis were carried out using Superscript II ( Invitrogen , CA ) . Two-to-three independent RNA preparations were used for RT-PCR and each of these were analyzed at least twice by RT–PCR . The RT–PCR was carried out for 35 cycles in order to determine absolute levels of transcripts . The number of amplification cycles was reduced to 21–25 in order to evaluate and quantify differences among transcript levels before they reached saturation . The amplified products were quantified using ImageQuant TL image analysis software ( GE , USA ) . Gene-specific primers used for RT–PCR analysis are described in Table S5 . The leaves were vacuum-infiltrated with trypan-blue stain prepared in 10 mL acidic phenol , 10 mL glycerol , and 20 mL sterile water with 10 mg of trypan blue . The samples were placed in a heated water bath ( 90°C ) for 2 min and incubated at room temperature for 2–12 h . The samples were destained using chloral hydrate ( 25 g/10 mL sterile water; Sigma ) , mounted on slides and observed for cell death with a compound microscope . The samples were photographed using an AxioCam camera ( Zeiss , Germany ) and images were analyzed using Openlab 3 . 5 . 2 ( Improvision ) software . The asexual conidiospores of H . arabidopsidis Emco5 expressing Atr8 were maintained on the susceptible host Nössen ( Nö ) or Nö NahG . The spores were removed by agitating the infected leaves in water and suspended to a final concentration of 105 spores/mL . Two-week-old seedlings were sprayed with spore suspension and transferred to a MTR30 reach-in chamber ( Conviron , Canada ) maintained at 17°C , 98% relative humidity and 8 h photoperiod . Plants were scored at ∼10–14 dpi and the conidiophores were counted under a dissecting microscope . The bacterial strain DC3000 derivatives containing pVSP61 ( empty vector ) , AvrRpt2 or AvrRps4 were grown overnight in King's B medium containing rifampicin ( Sigma , MO ) . The bacterial cells were harvested , washed and suspended in10 mM MgCl2 . The cells were diluted to a final density of 105 to 107 CFU/mL ( A600 ) and used for infiltration . The bacterial suspension was injected into the abaxial surface of the leaf using a needle-less syringae . Three leaf discs from the inoculated leaves were collected at 0 and 3 dpi . The leaf discs were homogenized in 10 mM MgCl2 , diluted 103 or 104 fold and plated on King's B medium . Transcripts synthesized in vitro from a cloned cDNA of TCV using T7 RNA polymerase were used for viral infections [59] , [60] . For inoculations , the viral transcript was suspended at a concentration of 0 . 05 µg/µL in inoculation buffer , and the inoculation was performed as described earlier [56] . After viral inoculations , the plants were transferred to a Conviron MTR30 reach-in chamber maintained at 22°C , 65% relative humidity and 14 hour photoperiod . HR was determined visually three-to-four days post-inoculation ( dpi ) . Resistance and susceptibility was scored at 14 to 21 dpi and confirmed by northern gel blot analysis . Susceptible plants showed stunted growth , crinkling of leaves and drooping of the bolt . Total RNA isolated from four-week-old plants using TRIZOL as outlined above . The experiment was carried out in triplicate and a separate group of plants was used for each set . RNA was processed and hybridized to the Affimetric Arabidopsis ATH1 genome array GeneChip following the manufacturers instructions ( http://www . affymetrix . com/Auth/support/downloads/manuals/expression_analysis_technical_manual . pdf ) . All probe sets on the Genechips were assigned hybridization signal above background using Affymetrix Expression Console Software v1 . 0 ( http://www . affymetrix . com/Auth/support/downloads/manuals/expression_console_userguide . pdf ) . Data was analyzed by one-way Anova followed by post hoc two sample t-tests . The P values were calculated individually and in pair-wise combination for each probe set . The identities of 162 NBS-LRR genes were obtained from the Arabidopsis information resource ( TAIR; www . arabidopsis . org ) and disease resistance gene homolog databases ( http://niblrrs . ucdavis . edu/ ) . FA analysis was carried out as described previously [61] . For FA profiling , one or few leaves of four-week-old plants were placed in 2 ml of 3% H2SO4 in methanol containing 0 . 001% butylated hydroxytoluene ( BHT ) . After 30 minutes incubation at 80°C , 1 mL of hexane with 0 . 001% BHT was added . The hexane phase was then transferred to vials for gas chromatography ( GC ) . One-microliter samples were analyzed by GC on a Varian FAME 0 . 25 mm×50 m column and quantified with flame ionization detection . The identities of the peaks were determined by comparing the retention times with known FA standards . Mole values were calculated by dividing peak area by molecular weight of the FA . SA and SAG quantifications were carried out from ∼300 mg of leaf tissue as described before [23] . SA treatments were carried out by spraying or subirrigating 3-week-old plants with 500 µM SA or 100 µM BTH . For glycerol treatment , plants were sprayed with 50 mM solution prepared in sterile water . Total protein was extracted in buffer containing 50 mM Tris pH 8 . 0 , 1 mM EDTA , 12 mM β-mercaptoethanol and 10 µg ml−1 phenylmethylsulfonyl fluoride . Proteins were fractionated on a 10–12% SDS-PAGE to confirm the quality . An antigen-coated enzyme-linked immunosorbent assay was used to determine levels of TCV CP in the infected plants as described before [62] . For protein gel blot analysis , leaf tissue from 4-week-old plants was extracted with a buffer containing 50 mM Tris-HCl , pH 7 . 5 , 10% glycerol , 150 mM NaCl , 10 mM MgCl2 , 5 mM EDTA , 5 mM DTT , and 1× proteinase inhibitor ( Sigma ) . Protein concentrations were determined by the Bradford assay ( Bio-Rad , CA ) . For immunodetection , 10–50-µg protein samples were electrophoresed on 10–15% polyacrylamide gels and run in the presence of 0 . 38 M Tris and 0 . 1% SDS . Proteins were transferred from the gels to polyvinylidene difluoride membranes by electroblotting , incubated with primary anti-HA antibody ( Sigma ) and alkaline phosphatase-conjugated secondary antibody ( Sigma ) . Immunoblots were developed using color detection .
|
Salicylic acid and enhanced disease susceptibility 1 are important components of resistance gene-mediated defense signaling against diverse pathogens in a variety of plants . Present understanding of plant defense signaling pathways places salicylic acid and enhanced disease susceptibility 1 downstream of resistant protein activation . In addition , enhanced disease susceptibility 1 is primarily thought to function in the signaling initiated via Toll-interleukin 1-receptor type of resistance proteins . Here , we show that salicylic acid and enhanced disease susceptibility 1 serve redundant functions in defense signaling mediated by coiled-coil-domain containing resistance proteins that were thought to function independent of enhanced disease susceptibility 1 . Furthermore , resistance signaling induced under low oleic acid conditions also requires enhanced disease susceptibility 1 and salicylic acid in a redundant manner , but these components are required upstream of resistance gene expression . Together , these results show that the functional redundancy between salicylic acid and enhanced disease susceptibility 1 has precluded their detection as required components of many resistance protein–signaling pathways .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"biology/plant-biotic",
"interactions",
"plant",
"biology"
] |
2009
|
Enhanced Disease Susceptibility 1 and Salicylic Acid Act Redundantly to Regulate Resistance Gene-Mediated Signaling
|
Escherichia coli translation initiation factor 2 ( IF2 ) performs the unexpected function of promoting transition from recombination to replication during bacteriophage Mu transposition in vitro , leading to initiation by replication restart proteins . This function has suggested a role of IF2 in engaging cellular restart mechanisms and regulating the maintenance of genome integrity . To examine the potential effect of IF2 on restart mechanisms , we characterized its influence on cellular recovery following DNA damage by methyl methanesulfonate ( MMS ) and UV damage . Mutations that prevent expression of full-length IF2-1 or truncated IF2-2 and IF2-3 isoforms affected cellular growth or recovery following DNA damage differently , influencing different restart mechanisms . A deletion mutant ( del1 ) expressing only IF2-2/3 was severely sensitive to growth in the presence of DNA-damaging agent MMS . Proficient as wild type in repairing DNA lesions and promoting replication restart upon removal of MMS , this mutant was nevertheless unable to sustain cell growth in the presence of MMS; however , growth in MMS could be partly restored by disruption of sulA , which encodes a cell division inhibitor induced during replication fork arrest . Moreover , such characteristics of del1 MMS sensitivity were shared by restart mutant priA300 , which encodes a helicase-deficient restart protein . Epistasis analysis indicated that del1 in combination with priA300 had no further effects on cellular recovery from MMS and UV treatment; however , the del2/3 mutation , which allows expression of only IF2-1 , synergistically increased UV sensitivity in combination with priA300 . The results indicate that full-length IF2 , in a function distinct from truncated forms , influences the engagement or activity of restart functions dependent on PriA helicase , allowing cellular growth when a DNA–damaging agent is present .
Translation Initiation Factor 2 ( IF2; for a review , see [1] ) is an essential cellular protein that brings mRNA , the 30S ribosome , and the initiator fMet-tRNA together into the 30S initiation complex and then promotes association with the 50S ribosomal unit to form the 70S initiation complex [2]–[4] . We have previously identified it as an essential component for reconstituting bacteriophage Mu replication by transposition in vitro , a process in which IF2 makes way for initiation of DNA synthesis by the cellular restart proteins [5] . This finding raises the question whether IF2 could play an important function in the maintenance of genome integrity by regulating the engagement or activity of restart proteins . For bacteriophage Mu transposition in vitro [6] , IF2 plays a critical part [5] during the transition from strand exchange catalyzed by MuA transposase [7] , [8] to the assembly of the replisome by the host replication restart proteins [9] ( Figure 1; for a review , see [10] ) . IF2 binds to Mu DNA only upon disassembly of the oligomeric MuA transpososome that remains tightly bound to Mu ends after strand exchange [11]–13 . This process begins as ClpX weakens the transpososome assembly [14]–[16] and is completed by host factors which promote transition to replisome assembly [5] , [9] , [15] , [17] . Strand exchange creates a fork at each Mu end , creating a potential site for initiating Mu DNA replication . However , the Mu forks retain a block to initiation of DNA replication even after transpososome disassembly , and IF2 appears to play a key role in unlocking this complex [5] . Restart proteins are subsequently assembled , beginning with the displacement of the IF2 by PriA helicase . The reaction in vitro specifically requires the E . coli replication restart proteins PriA , PriC , and DnaT but not PriB , indicating that the mode of Mu replication reconstituted in this system is through the PriA-PriC restart system [18] , [19] . ( The PriA-PriC pathway is one of the two major cellular restart pathways , the other being the PriA-PriB pathway , which requires PriA , PriB , and DnaT [18] . ) Additionally , only truncated forms of IF2 ( IF2-2 and IF2-3; Mr of 79 . 7 and 78 . 8 k compared to 97 . 3 k for full-length IF2-1 ) , synthesized from two internal , in-frame start codons within the infB gene , have been found to be active in this in vitro system . Indeed , the role of the various IF2 forms in translation remains unclear . Full-length ( IF2-1 ) and truncated ( IF2-2/3 ) forms are present in nearly equimolar amounts under normal growth conditions [20] , [21] , and IF2-2/3 levels increase with respect to IF2-1 during cold shock [22] . Mutations that prevent expression of IF2-1 or IF2-2/3 elicit cold sensitivity [21] . However , even IF2 with one-third of its residues deleted from the N-terminal end has intact activities in vitro as translation factor and supports cell viability when present in excess [23] , [24] . IF2's role in Mu DNA replication by transposition in vitro raises the question whether it can influence or regulate the engagement of cellular restart mechanisms . The apparent function implied by the Mu replication system is that by binding to forked DNA templates , it may promote or regulate the action of restart proteins . IF2's molecular chaperone activity [25] potentially plays a function similar to ClpX , promoting remodeling of the nucleoprotein assembly at the Mu ends for the transition to a new complex [5] or plays a key part in the activation of enzymatic functions necessary for replication restart . Moreover , IF2's major function as translation factor as well as its possible function as a transcriptional activator [26] , [27] also indicate its potential to influence restart mechanisms by promoting expression of proteins needed for this process . Indeed , the role of IF2 in Mu replication may be an idiosyncrasy of Mu as a parasite exploiting host proteins to promote its own propagation; alternatively , it may reflect IF2's cellular role in regulating engagement of restart functions , a function that Mu exploits as a parasite . In this work , we examined whether IF2 function can affect specific pathways for replication restart by perturbing its function with mutations that prevent expression of IF2-1 or IF2-2/3 . Only truncated forms of IF2 have been found to be active in the reconstituted Mu replication system by the PriA-PriC pathway [5] . While this result does not necessarily indicate that only the truncated forms of IF2 may be involved in restart mechanisms ( the in vitro system may have lacked factors needed to engage IF2-1 ) , it nevertheless suggests functional differences between isoforms that may be examined in vivo . Here , we demonstrate that the loss of IF2-1 or IF2-2/3 results in different defects in restart mechanisms that cope with DNA damage during cell growth . In particular , the loss of IF2-1 elicits a phenotype that is analogous to a certain restart mutant . No matter the mechanism by which IF2 influences restart mechanisms , the results indicate a new function of IF2 in influencing the engagement of restart mechanisms , the relative levels of IF2 isoforms having the potential to affect the choice or course of the restart mechanism . We discuss the potential for IF2 to regulate maintenance of genome integrity with respect to cell physiology , suggesting a means for coordinating replication , recombination , and repair with translation status .
In the in vitro Mu replication system , binding of IF2-2 can be detected after strand exchange just prior to the binding of the restart protein PriA [5] . Since this is the major basis for suspecting that IF2 may serve a function that affects activity of restart functions , we wished to confirm that IF2 indeed binds at or near Mu ends in vivo when Mu development is induced . Chromatin immunoprecipitation ( ChIP ) analysis was conducted with extracts of induced lysogens expressing IF2 with an N-terminal S tag ( S-IF2 ) after extensive RNase treatment . Mu DNA was co-precipitated with S-IF2-1 and S-IF2-2 in induced GTN373 ( a thermoinducible Mu lysogen ) at 35 min postinduction ( Figure 2A , S-IF2-1 and S-IF2-2 ) , using antibody against the S tag . In contrast , relatively little Mu DNA was precipitated with S-IF2-2 upon inducing the isogenic lysogen that has a clpX knockout mutation ( Figure 2A , S-IF2-2 ClpX− ) and thus cannot support Mu replication [28] . This result parallels findings in vitro that the omission of molecular chaperone ClpX from the reaction system does not permit binding of IF2-2 to Mu DNA and the initiation of Mu replication [5] , [15] . As it appeared that Mu ends were being enriched in immunoprecipitations when cells were undergoing Mu replication , we repeated the ChIP with 5-fold less antibody to ascertain whether bound S-IF2-2 in induced GTN373 is concentrated around Mu ends . In the immunoprecipitated samples , the Mu ends sequences were enriched over the center sequences ( 18 kb from either end ) as well as host DNA ( Figure 2C ) . In the control PCR amplification of total DNA , the Mu end and center sequences were amplified to the same extent . Mu PCR products were produced at higher levels than the host thrA PCR product at 35 min postinduction , reflecting the replication of Mu during lytic development . IF2 does have some nonspecific DNA binding activity [27] . Thus , the enrichment of Mu end sequences with respect to Mu center sequences by immunoprecipitation is the best indicator of preferred IF2 binding at or near Mu ends although the enrichment of Mu end sequences with respect to host DNA is also clear in this analysis . To ensure that the anti-S tag antibody was specifically precipitating Mu DNA bound to S-IF2-2 , we compared the co-precipitation of Mu DNA ( 35 min postinduction ) in induced lysogens expressing S-IF2-2 and untagged IF2-2 ( Figure 2B ) . When the IF2-2 had no S tag , no more Mu DNA was captured in the immunoprecipitation than in the no-antibody control . The results indicate that not only truncated IF2-2/3 but also full-length IF2-1 bind at or near Mu ends upon induction of Mu development , corroborating the role IF2 plays in vitro in promoting initiation of Mu DNA replication by restart proteins . In vitro , IF2 makes way for the binding of PriA [5] , which binds to forked DNA structures [29] , [30] such as the Mu fork , and PriA subsequently displaces IF2 from Mu DNA . The ChIP analysis by itself can only indicate a preponderance of IF2 binding around the Mu ends and does not rule out the possibility that IF2 binds at nearby sites . Nevertheless , these results together with the role IF2 plays in vitro strongly suggest that there are IF2 molecules bound at the Mu fork during lytic development . The role played by IF2 in Mu replication raises the question whether IF2 function can regulate the engagement or activity of restart functions . We constructed a series of strains with infB alleles that only allow expression of full-length IF2-1 or the truncated forms IF2-2/3 to examine their effect on restart functions . The infB alleles were introduced into the chromosome where a transposon vector was inserted , and then the natural infB allele was knocked out by introduction of the del ( infB ) 1::tet allele , which precisely deletes the natural cistron for IF2 ( Figure 3A–3B ) . To prevent the expression of IF2-1 , we deleted sequences around the translation initiation start site for IF2-1 . Sequences from 14 nucleotides upstream of the IF2-1 start codon to 32 nucleotides upstream of the IF2-2 start codon were deleted ( Figure 3B ) ; this is known to permit expression of the truncated IF2 forms while eliminating IF2-1 expression [21] . The resulting allele , denoted as infB ( del1 ) to indicate that the deletion prevents expression of IF2-1 , supports the synthesis of only IF2-2 and IF2-3 . Expression of the truncated IF2 forms were prevented by changing the start codons of IF2-2 and IF2-3 , gug to guc ( g474c ) and aug to acg ( t494c ) ; these mutations have previously been shown to eliminate expression of the truncated forms while leaving a functional IF2-1 [21] . We shall refer to this allele as infB ( del2/3 ) to indicate that the mutations prevent expression of IF2-2 and IF2-3 even though del2/3 is not a deletion mutation . The resulting infB del1 , del2/3 , and wild-type ( wt ) alleles were introduced into the transposon site as part of the nusA infB operon ( <nusAinfB> to signify that this is encoded within the transposon ) . The natural infB allele could be readily knocked out by introducing the del ( infB ) 1::tet allele when the operon in the transposon had infB ( wt ) , infB ( del1 ) , and infB ( del2/3 ) alleles . ( The procedures for verifying deletion of the natural infB allele as illustrated in Figure 3C and for verifying infB alleles by PCR and sequencing will be described under Materials and Methods . ) The <infB ( del2/3 ) > and especially the <infB ( del1 ) > strains display some measure of cold sensitivity , growing very slowly at 25°C and below , consistent with previous reports about strains with analogous alleles [21] . We determined that the strain with the single copy <infB ( del1 ) > as sole allele was highly sensitive to MMS whereas the strains with <infB ( wt ) > and <infB ( del2/3 ) > as sole alleles were not ( Figure 4A and Figure S1A ) . The results indicate that the del1 mutation causes the inability to grow in the presence of MMS . The question is whether this is due to a general deficiency in repair , recombination , and restart functions , resulting from a generally deficient translation initiation function , or whether there is any specificity of the defect . We should note that the <infB ( del1 ) > strain ( ArgA− ) was at least moderately proficient in homologous recombination measured by P1 transduction , although the frequency of Arg+ transductants was reduced approximately 5 fold compared to <infB ( wt ) > and <infB ( del2/3 ) > strains ( Figure S1B ) . While some reduction in homologous recombination frequency may be part of the phenotype of this strain , the reduction seen here is modest compared to the 20–50 fold reduction in P1 transduction demonstrated for the priA knockout strain [31] . To determine whether it is indeed IF2-1 that is needed to maintain MMS-resistance , we complemented the <infB ( del1 ) > allele of strain GTN1156 with the infB ( del2/3 ) allele , harbored as part of a nusA infB operon on a plasmid with a pSC101 replicon , pSPCnusAinfB ( del2/3 ) . While the empty plasmid vector could not confer MMS-resistance and homologous recombination proficiency , IF2-1 expressed from pSPCnusAinfB ( del2/3 ) did restore high viability on MMS plates ( Figure 4A ) . In contrast , IF2-2/3 expressed from the plasmid-borne infB ( del1 ) allele only partially restored viability on MMS plates ( Figure 4A ) . While the multicopy infB ( del1 ) allele did increase dramatically the viable count on MMS plates , the colonies grew up very slowly , and the viable count on these plates was still 5–10 fold lower than that of the strain with the multicopy infB ( del2/3 ) allele ( Figure 4A ) . These results illustrate functional differences between IF2-1 and IF2-2/3 in promoting recovery after MMS treatment . They also indicate that IF2-2/3 when expressed from a multicopy vector may compensate for the lack of IF2-1 , albeit inefficiently . To confirm that it was not just the DNA segment deleted in the infB ( del1 ) allele but the full-length IF2-1 protein that was needed for complementation , we introduced IF2 G domain mutations , infB ( c1227a ) or ( c1501a ) , which result in the IF2 D409E and D501N alterations , respectively , into pSPCnusAinfB ( del2/3 ) . The infB ( D409E ) allele is an example of a viable G mutant that is functional at 37°C [32] whereas infB ( D501N ) is a recessive allele that is lethal as a single-copy gene [33] . Introduction of pSPCnusAinfB ( del2/3 , D409E ) into GTN1156 , but not pSPCnusAinfB ( del2/3 , D501N ) , restored MMS-resistance ( Figure 4A ) . IF2-1 must therefore be providing the function needed for viability in MMS . The level of homologous recombination in the <infB ( del1 ) > mutant , examined by P1 transduction , could also be increased by supplying the various infB alleles on the plasmid vector ( Figure S1C ) . Due to the relatively modest effect on homologous recombination , this aspect of the infB ( del1 ) mutant was not further examined . The <infB ( del1 ) > strain , which produces only IF2-2/3 and has extremely low viability in the presence of 6 mM MMS , attains high viability when complemented with the plasmid-borne infB ( del2/3 ) allele , which restores IF2-1 production ( Figure 4A and 4B ) . This indicates that the multicopy infB ( del2/3 ) allele is dominant over the <infB ( del1 ) > allele . The inability of pSPCnusAinfB ( del2/3 , D501N ) to restore efficient growth of the <infB ( del1 ) > strain in MMS could indicate the inactivation of a necessary function of IF2-1 by the D501N mutation . Alternatively , the <infB ( del1 ) > allele may be dominant negative over the multicopy infB ( del2/3 , D501N ) allele in terms of supporting growth in MMS . Although the D501N mutation is lethal when present as a single-copy infB allele , this mutation is recessive to the wild-type allele [33] . We therefore tested whether the multicopy infB ( del2/3 , D501N ) allele on the plasmid could support viability by itself . The natural infB allele in strain GTN932 that bears plasmids pSPCnusAinfB ( del2/3 ) , pSPCnusAinfB ( del2/3 , D409E ) , or pSPCnusAinfB ( del2/3 , D501N ) could readily be knocked out , leaving the infB on the plasmid as the sole allele in the cell . This allowed us to test whether or not IF2-1 ( D501N ) , expressed from multicopy <infB ( del2/3 , D501N ) > , is defective in a function that IF2-1 provides but IF2-2/3 fails to perform . Although the strain with the multicopy infB ( del2/3 , D501N ) as the sole allele grew relatively slowly , requiring at least twice the incubation time as the other two strains for growth , it was clearly viable and also retained significant viability on MMS plates , comparable to viability of analogous strains with infB ( del2/3 ) and infB ( del2/3 , D409E ) as sole alleles ( Figure 4B ) . That is , the multicopy infB ( del2/3 , D501N ) is able to support high viability in MMS so long as the <infB ( del1 ) > allele is absent . Introduction of the D501N mutation to the multicopy infB ( del2/3 ) allele thus results in loss of dominance over <infB ( del1 ) > , not in the loss of a function needed to maintain viability in MMS . These results suggest that IF2-2/3 , at levels produced from the single-copy <infB ( del1 ) > allele , is performing a function in a way that aggravates problems which the cells encounter during growth in MMS , outcompeting IF2-1 ( D501N ) that is able to carry out the function appropriately to maintain viability . In other words , IF2-2/3 does not necessarily lack the capacity to perform the IF2-1 function . Rather , it appears to carry it out in a way that dramatically reduces viability . That is , the recessive properties of infB ( D501N ) with respect to the infB ( del1 ) allele , including its ability to support resistance to MMS as the sole multicopy allele , suggest that MMS sensitivity of the <infB ( del1 ) > strain is not simply due to a general deficiency in translation initiation function when only IF2-2/3 is present . We next determined whether the MMS sensitivity of the <infB ( del1 ) > mutant reflected deficiency in the levels of repair or restart proteins in these mutants . In the analysis described above , MMS resistance was measured by growth of cells on plates containing MMS . By this analysis , cells must not only survive initial exposure to the DNA-damaging agent but also grow into colonies in its presence . We also measured the ability of strains exposed to MMS to recover and grow in the absence of MMS in order to assess their capacity to repair DNA lesions and restart DNA replication . Strains that are defective in genes such as priA , recA , and polA that participate in DNA repair or replication restart are known to be quite sensitive as measured by initial exposure for 15 min in MMS and plating without MMS to determine the number of survivors; the alkA tag mutant , which is defective in a major mechanism for repairing alkylated bases ( base excision repair ) , is also sensitive to MMS by this criteria [34] . The <infB ( del1 ) > mutant was as resistant to MMS as infB ( wt ) strains by this criteria ( Figure 4C ) , with MMS resistance comparable to strains with natural infB , <infB ( wt ) > , and <infB ( del2/3 ) > alleles; in contrast the recA938 mutant was highly sensitive by this criteria ( Figure 4C ) . It should be noted that when cells were deficient in both the PriA-PriB and PriA-PriC pathway ( deficient in both PriB and PriC ) , they had very low viability even without MMS treatment ( Figure S2A ) . As restart mutants tend to have very low viability even without MMS , we measured MMS sensitivity of a del ( dnaT ) 759::kan mutant with a dnaC ( a491t ) suppressor mutation , which greatly increases cell viability . Even with the suppressor mutation , the dnaT knockout strain was significantly more sensitive to the 15-min MMS treatment ( Figure S2B ) than the <infB ( del1 ) > mutant . These results indicate that levels of repair and restart factors in the <infB ( del1 ) > strain are sufficient for the recovery of DNA replication and cell growth after DNA damage by MMS . However , there was a 1000-fold reduction in viability of the <infB ( del1 ) > mutant on 6 mM MMS plates ( Figure 5A ) . That is , the <infB ( del1 ) > strain is proficient in repairing DNA damage and resuming DNA replication after the 15-min exposure in MMS , but it is severely defective in its ability to sustain growth in MMS . Thus , the <infB ( del1 ) > mutant is not able to cope with the sustained damage to DNA during cell growth . This could indicate that a repair or restart factor , although not deficient , is sufficiently low such that it cannot keep up with constant DNA damage inflicted on MMS plates; alternatively , it is possible that the regulation of repair and restart processes are not appropriate for efficiently supporting DNA replication under these conditions . Introduction of the sulA::Mud ( lac , Ap , B::Tn9 ) allele greatly restored viability of the <infB ( del1 ) > mutant in MMS ( Figure 5A ) . The sulA gene , which is a component of the SOS system induced by DNA damage , is a cell division inhibitor [35] . In mutants such as the priA null strain , which has a constitutively induced SOS system , the high expression of sulA results in loss of viability , which can be largely restored by sulA mutations [36] . It is important to note that the sulA::Mud ( lac , Ap , B::Tn9 ) allele did not fully restore viability to the <infB ( del1 ) > mutant . Moreover , scorable colonies on MMS plates required incubation for over 36 hours at 37°C whereas infB ( wt ) colonies readily arose in 16 hours . That is , the sulA mutation did not fully revert <infB ( del1 ) > to the wild-type phenotype . Interestingly , the priA300 mutant had a phenotype much like the <infB ( del1 ) > mutant , resistant to MMS when exposed to MMS and plated in its absence but highly sensitive when plated on 6 mM MMS plates ( cf . Figure 4C with Figure 5A ) . The priA300 allele encodes for a helicase-deficient PriA that is fully proficient in primosome and replisome assembly by the PriA-PriB pathway [19] , [37] . The priA300 mutant has previously been shown to have essentially a wild-type phenotype unless that mutation is combined with mutations affecting other restart functions such as priB; wild-type properties of the priA300 mutant include homologous recombination proficiency and relatively high UV resistance [19] , [38] . As with the <infB ( de11 ) > mutant , the sulA::Mud ( lac , Ap , B::Tn9 ) allele could restore viability of the priA300 mutant in MMS ( Figure 5A ) . In addition , priA300 was epistatic with the infB ( del1 ) allele , causing no significant increase in MMS sensitivity ( Figure 4C and Figure 5A ) . In contrast , the <infB ( del1 ) > del ( priB ) 302 combination ( GTN1117 ) was synergistic , reducing viability to 0 . 010±0 . 002% on the MMS plates . The priB knockout alone did not have such a severe effect; the <infB ( wt ) > del ( priB ) 302 strain ( GTN1133 ) had a viability of 43±8% on MMS plates . In addition , knockout of priC did not increase UV sensitivity; the <infB ( wt ) > del ( priC ) 752::kan strain ( GTN1059 ) had a viability of 84±7% on MMS plates . These results indicate that the PriA-PriC pathway , which requires PriA helicase , is not solely responsible for allowing cell growth in the presence of MMS and that the PriA-PriB pathway most likely makes a significant contribution to mechanisms dependent on PriA helicase as well . We shall further examine the interactions of priA300 and del ( priB ) 302 with the <infB ( de11 ) > and <infB ( de12/3 ) > alleles by UV sensitivity . The epistatic relationship between the infB ( del1 ) and priA300 alleles suggests that the loss of IF2-1 specifically affects the activity or engagement of factors in restart pathways that require PriA helicase . Despite the high MMS sensitivity of the <infB ( del1 ) > strain , it did not resemble the priA knockout mutant in terms of having constitutively high levels of SOS induction ( Figure 5B ) . Expression from the sulA::lacZ SOS reporter was significantly lower than the strain with wild-type priA and infB and the priA300 strain . The latter strain had moderate basal levels of SOS induction , which was significantly less than that of the priA knockout . Treatment of the wild-type and priA300 strains with 18 mM MMS elicited moderate increases in SOS expression; in contrast , treatment of the <infB ( del1 ) > strain elicited over a 10-fold increase in SOS expression , consistent with the role of SOS induction reducing the strain's viability upon MMS treatment , Although the <infB ( del1 ) > strain was sensitive to growth in MMS , it was slightly more resistant to UV light than the <infB ( wt ) > strain ( Figure 6A ) . In fact , the <infB ( del2/3 ) > mutant , which was found to be the most MMS-resistant , was slightly more UV sensitive than the <infB ( wt ) > strain ( Figure 6A ) . These results do not rule out the possibility that the del1 and del2/3 mutations impair or knock out restart mechanisms engaged after UV irradiation . As there are multiple restart pathways in the cell , the PriA-PriB and PriA-PriC pathways being the two major ones [18] , the del1 or del2/3 mutation may predominantly affect only one pathway but not the other . To test this possibility , we examined the effect of the infB alleles in combination with priB or priC knockout alleles . It is well established that the knockout of priB or priC has little to no effect by itself [39] in contrast to the priA or dnaT knockouts , which affects both major restart pathways and elicits high sensitivity to DNA-damaging agents and low viability [18] , [36] , [40] . As expected , neither the priB nor priC knockout had any effect on UV sensitivity when introduced into the parent strain ( GTN932 ) used to construct the various <infB> mutants ( Figure 6D ) . While the del ( priC ) 752 allele had absolutely no effect on single-copy <infB ( wt ) > , <infB ( del1 ) > , and <infB ( del2/3 ) > strains ( Figure 6C; cf . with Figure 6A ) , the del ( priB ) 302 clearly had a synergistic effect with the infB ( del1 ) mutation to elicit relatively high UV sensitivity ( Figure 6B ) . This finding that the priB knockout , but not the priC knockout , is synergistic with the <infB ( del1 ) > allele to increase UV sensitivity indicates that the loss of full-length IF2-1 diminishes the PriA-PriC pathway for recovery after UV irradiation . Introduction of a pBAD24-priB plasmid into the del ( priB ) 302 <infB ( del1 ) > strain ( GTN1117 ) , allowing the expression of PriB driven by the PBAD promoter with arabinose as inducer , increased its UV resistance to levels comparable to the del ( priB ) 302 <infB ( wt ) > strain ( GTN1133; Figure 7A ) , confirming that the deficiency of GTN1117 can be reversed by expressing PriB . This indicates that the activity of repair and restart proteins needed for recovery after UV irradiation in GTN1117 , which has the <infB ( del1 ) > allele , is comparable to that in GTN1133 , which has the <infB ( wt ) > allele . Therefore , the increased UV sensitivity of GTN1117 with respect to GTN1133 is most likely due to some type of deficiency in the PriC-dependent pathway . We were unable to measurably increase UV resistance by expressing PriC from pBAD24-priC ( Figure 7A ) . Indeed , PriC in its active form must be present in GTN1117 . When the chromosomal priC was knocked out in pBAD24-priC/GTN1117 ( GTN1566 ) , expression of PriC from the plasmid vector became essential for viability with or without pre-treatment with MMS ( Figure S2A ) , viability being less than 0 . 1% in the presence of glucose . In the presence of arabinose , viability of GTN1566 with or without MMS treatment was comparable to the strain with an intact chromosomal priC . That is , active PriC can be expressed from pBAD24-priC or the chromosomal priC gene in the <infB ( del1 ) > genetic background , and supplementation of PriC expression in GTN1117 from the plasmid cannot restore any measure of UV resistance . These results suggest that its relatively high UV sensitivity is not caused by a deficiency in PriC , PriA , and DnaT . Although the del ( priB ) 302 <infB ( del1 ) > strain ( GTN1117 ) has high UV sensitivity , its ability to recover after a 15-min exposure to MMS was comparable to the wild-type control ( Figure 4C ) . Moreover , Mu plating efficiency on this strain is not dramatically reduced , indicating that the PriA-PriC pathway can promote Mu replication in the absence of IF2-1 ( Figure S3 ) , a result consistent with properties of Mu replication in vitro [5] . In general , the Mu plating efficiencies on the various <infB ( wt , del1 , or del2/3 ) > strains , whether in the PriB+PriC+ , del ( priB ) 302 , or del ( priC ) 752::kan genetic backgrounds , were nearly the same . These results indicate that restart proteins needed to promote Mu replication by the PriA-PriC pathway are present at sufficient levels to support lytic development . They also suggest that the defect of the infB ( del1 ) allele is not a deficiency in restart activity needed for recovery but rather in the regulation of restart activity needed to maintain replication in the presence of the DNA-damaging agent . Although the effect is not as much as in the <infB ( del1 ) > background , the del ( priB ) 302 allele also did significantly increase UV sensitivity when introduced into the <infB ( wt ) > background ( cf . the solid square data points in Figure 6A and 6B ) whereas it had essentially no effect in the natural infB ( wt ) background ( Figure 6D ) . This may reflect a small change in relative or absolute levels of full-length and truncated IF2 when the infB allele is expressed from the transposon site , a change that has no discernible effect unless specific restart mechanisms are inactivated as with the del ( priB ) 302 mutation . Interestingly , in the <infB ( wt ) > background both the priA300 ( Figure 7B ) and the del ( priB ) 302 ( Figure 6B ) allele increased UV sensitivity to the same level . Like the del ( priB ) 302 , the priA300 allele is known to have little effect on UV sensitivity [19] , and indeed we found essentially no effect of the priA300 allele in the GTN932 background ( Figure 6D ) , which has the natural infB ( wt ) allele . As we described above , the priA300 and <infB ( del1 ) > alleles both independently elicit sensitivity to growth in MMS , and the two mutations are epistatic for this trait , consistent with a model in which PriA helicase and IF2-1 function in the same pathway to maintain efficient growth in MMS . In the UV sensitivity analysis , the infB ( del1 ) allele was also found to be epistatic with priA300 , not being able to elicit further UV sensitivity in the priA300 background ( Figure 7B ) . That is , loss of IF2-1 attenuates pathways dependent on PriA helicase such as the PriA-PriC pathway . In contrast , the infB ( del2/3 ) allele was synergistic with priA300 to increase UV sensitivity ( Figure 7B ) . The results indicate that loss of IF2-2/3 from the infB ( del2/3 ) allele results in deficiency of a restart pathway that is distinct from the IF2-1/PriA helicase-dependent pathway . Mu plating efficiency on the three priA300 strains with the <infB ( wt ) > , <infB ( del1 ) > , and <infB ( del2/3 ) > were essentially the same , the titer obtained on the latter two strains being greater than 90% of the titer on the priA300 <infB ( wt ) > strain . Thus , as with the del ( priB ) 302 <infB ( del1 ) > combination , which also synergistically contributes to high UV sensitivity , the priA300 <infB ( del2/3 ) > combination does not lead to an inability to initiate Mu replication by the available host restart machinery . What is notable about the UV sensitivity analysis is that the combination of priA300 <infB ( del2/3 ) > or del ( priB ) 302 <infB ( del1 ) > mutations does not produce the extremely severe phenotype of the priA300 del ( priB ) 302 combination , which elicits a phenotype analogous to the priA knockout [19] . That is , loss of IF2-1 or IF2-2/3 does not result in the inability to promote replication restart by the respective pathways they influence , but rather the loss of each IF2 isoform affects some mechanism needed to maintain high viability when the restart mechanism is engaged after DNA damage . However , under normal growth conditions or if cells are allowed to recover after MMS treatment or UV irradiation without the presence of DNA damaging agents , there is little effect of knocking out IF2 isoforms , and a mild effect is seen when these mutations are combined with the restart mutation del ( priB ) 302 or priA300 , which by itself has little effect under normal growth conditions . We examined the cell morphology of the various infB mutants to examine whether there is an increased incidence of sporadic SOS induction , leading to filamentation [41] of a small fraction of the cells in the population . The strains with the single del ( priB ) 302 or <infB ( del2/3 ) > mutant had essentially wild-type morphology ( Figure S4A ) , 100% of cells being 0–8 µm in length when at least 40 , 000 cells were analyzed . Cells with the single <infB ( del1 ) > or priA300 mutation ( GTN1114 and GTN1298 , respectively ) tended to be longer in size , with a higher incidence of moderate sized filaments ( examples of moderate filaments are indicated by white arrows ) . In a sample of 1000 cells , 1% of the cells were in the 8–30 µm range for GTN1114 and GTN1298 . Consistent with the relatively low basal levels of SOS expression measured for the <infB ( del1 ) > mutant at the macroscopic level ( Figure 5B ) , the level of its filamentation was quite low compared with that of the priA knockout mutant ( Figure S4B ) , but the moderate filamentation suggests an increased incidence of sporadic SOS induction . What was notable for the synergistic <infB ( del1 ) > del ( priB ) 302 combination ( GTN1117 ) was that it gave rise to a low but significant frequency of very large filaments greater than 30 µm ( Figure S4A ) . The incidence of filaments over 30 µm in size was found to be 0 . 13% in a screening of 33 , 000 total cells , most of these large filaments ( 0 . 10% of total cells ) being over 50 µm in length . Only one other combination of an infB allele with the priA300 , del ( priB ) 302 , or wild-type restart functions ( Figure S4A ) yielded any filaments over 50 µm in 100 , 000 cells screened . The mutant with the synergistic <infB ( del2/3 ) > priA300 combination ( GTN1297 ) produced filaments greater than 30 µm at a significantly lower frequency of 0 . 02% in a screening of 100 , 000 cells , of which only 3 were greater than 50 µm . Filaments in the 30–50 µm range also arose with the single <infB ( del1 ) > or priA300 mutants ( GTN1114 and GTN1298 , respectively ) but with a frequency of no more than one in 40 , 000 cells . No filaments of greater than 30 µm were detected with the <infB ( wt ) > ( GTN1050 ) , <infB ( del2/3 ) > ( GTN1115 ) , del ( priB ) 302 ( GTN1133 ) , and the <infB ( del1 ) > priA300 ( GTN1323 ) strains when at least 100 , 000 cells were examined . The results indicate that the <infB ( del1 ) > del ( priB ) 302 mutant ( and , to a lesser extent , the <infB ( del2/3 ) > priA300 mutant ) has an increased incidence of very high SOS induction ( leading to the formation of giant filaments ) in a small fraction of the cell population growing in LB , suggesting a reduced capacity to cope with accidents that might occur during DNA replication for normal cell growth . However , these mutants clearly do not have the characteristics of extensive SOS induction as with a priA knockout strain such as GTN430 ( Figure S4B; 3% of cells producing filaments greater than 30 µm in a sample of 4000 cells , 2% greater than 50 µm ) . The characteristics of strains such as the <infB ( del1 ) > or priA300 mutant are more akin to a priA knockout strain that has acquired a suppressor mutation in dnaC ( Table 1 ) . GTN412 , which is a Mucts62 lysogen , can support Mu replication upon thermoinduction to yield a high level of infective centers , has a high level of viability on MMS plates , and has a relatively low level of expression from its SOS reporter gene ( dinD::lacZ ) . Introduction of the priA knockout decreased viability on MMS and formation of Mu infective centers by several orders of magnitude . The presence of a suppressor mutation in dnaC ( GTN522 ) did diminish cell filamentation ( Figure S4B; the number of filaments >30 µm are reduced to 0 . 05% from 3% , measured in a sample of 15 , 000 cells ) , reduce the level of SOS induction as indicated by the dinD::lacZ reporter , and restore the ability to form Mu infective centers , but this strain retained the severe sensitivity to growth in the presence of MMS , a central feature of the both the <infB ( del1 ) > and priA300 mutants . This is consistent with the ability of the dnaC suppressor mutation to bypass the requirement for PriA to initiate DNA synthesis at forked DNA structures [38] , [42]; however , without PriA the mechanism for promoting replication restart and promoting high viability in the presence of MMS ( the IF2-1/PriA helicase-dependent pathway ) appears to be compromised . In the same way , a priA300 or <infB ( del1 ) > mutant may be able to promote replication restart by a less preferred pathway , which may permit replication restart to proceed but does not do so in a way that supports high viability during growth in the presence of MMS .
The present work indicates a special relationship between the PriA helicase function and IF2-1 ( see Table 2 ) . Both the PriA helicase function and IF2-1 are required to allow cells to grow with maximal viability in the presence of MMS . Nevertheless , neither of these mutants display the severe characteristics of the priA knockout , having UV resistance that is comparable to wild type and being able to recover from MMS treatment with very high viability provided that it can do so in the absence of MMS . The defect of the priA300 mutant , previously shown to have nearly a wild-type phenotype [19] , is a surprising new phenotype , being defective in the ability to grow in the presence of MMS but not in its ability to recover from MMS treatment . Even more surprising is the finding that the loss of the IF2-1 function elicits the same phenotype . Another characteristic which indicates that the infB ( del1 ) allele affects some aspect of replication restart is the suppressing effect of knocking out sulA , a mutation that greatly increases viability of both the infB ( del1 ) and priA300 mutant on MMS plates . Moreover , MMS treatment of inf ( del1 ) mutant promotes an especially high level of SOS induction compared to the level promoted in wild type . A relationship between full-length and truncated IF2 isoforms and replication restart functions is further indicated by UV sensitivity analysis . Both the <infB ( del1 ) > and <infB ( del2/3 ) > exhibit UV resistance comparable to wild type , but the combinations of <infB ( del1 ) > del ( priB ) 302 and <infB ( del2/3 ) > priA300 significantly enhance UV sensitivity . Moreover , the <infB ( del1 ) > del ( priB ) 302 mutant ( and , to a lesser extent , the <infB ( del2/3 ) > priA300 mutant ) display an increased frequency of sporadic SOS induction , indicated by the increased frequency of very long filaments over 30 µm . Clearly , the general population of these cells do not display the same high level of SOS induction of the priA knockout cells at the macroscopic level . The sporadic nature of filamentation is consistent with the thinking that these cells are mostly proficient in coping with accidents of DNA replication which may arise during normal growth conditions , unlike the priA knockout that copes with such accidents poorly . One would expect that only a small minority of cells would need to cope with a large number of DNA lesions during growth in LB unless a DNA-damaging agent such as MMS is present . The combination of <infB ( del1 ) > del ( priB ) 302 and <infB ( del2/3 ) > priA300 alleles may sufficiently attenuate the major pathways that lead from DNA damage to replication restart , thus manifesting a modest but significant increase in sensitivity to UV irradiation . The epistatic relationship between the priA300 and infB ( del1 ) alleles revealed by both UV sensitivity and viability on MMS plates indicates that IF2-1 and PriA helicase function in common pathways as proposed in Figure 8A . This includes mechanisms in both the PriA-PriB and PriA-PriC pathway , for neither the priB or priC knockout has the severe effect of priA300 for growth on MMS plates . What remains of the major restart pathways when PriA helicase is inactive are mechanisms in the PriA-PriB pathway that can operate in the priA300 background [19] . Thus , the effect of the infB ( del2/3 ) allele in this genetic background ( increased UV sensitivity and increased incidence of sporadic cell filamentation ) suggests that IF2-2/3 plays a role in this pathway . However , we have yet to find a phenotype for the infB ( del2/3 ) allele alone , comparable to MMS sensitivity of the infB ( del1 ) mutant , and whether IF2-2/3 is a key participant in PriA helicase-independent restart mechanisms ( Figure 8A ) remains be determined . Finally , the characteristics of the <infB ( del1 ) > and priA300 mutants and especially the infB ( del1 ) del ( priB ) 302 double mutant are more like the priA knockout with a suppressor mutation in dnaC rather than the priA knockout with no suppressor . The <infB ( del1 ) > and priA300 mutants , like the priA knockout with suppressor , do not exhibit the extreme sensitivity to UV irradiation , the massive cell filamentation , and the inability to support Mu replication that is characteristic of the priA knockout . Nevertheless , all of these mutants are not able to grow efficiently on media containing 6 mM MMS , their viability on MMS plates being approximately 0 . 1% or less . For the priA knockout , the dnaC suppressor allows replication restart to proceed , but the bypass of the restart proteins compromises maintenance of high cell viability when DNA replication proceeds during relatively high rates of DNA damage . Similarly , replication restart mechanisms can still operate in the <infB ( del1 ) > mutant , and the lack of IF2-1 may bypass the preferred pathway that maintains high cell viability during growth in the presence of MMS . As IF2-1 and IF2-2/3 share 726 common residues , IF2-1 having 157–164 extra residues at the N-terminal end , it is quite conceivable that IF2-2/3 can replace IF2-1 in the IF2-1/PriA helicase-dependent pathway , allowing replication restart to proceed but lacking the function need to maintain high cell viability . The ability to grow under conditions that damage DNA at elevated levels could provide cells with the selective advantage that conserves the function of restart proteins despite the fact that suppressor mutations can bypass the need for these proteins . For example , the fact that the helicase motif of PriA is highly conserved among diverse bacteria [43] has been puzzling in light of the fact that its inactivation by the priA300 mutation seemed to have little effect on the cell phenotype , but the ability of cells with active PriA helicase to grow under conditions that damage DNA at a relatively high rate would indeed be a selective advantage that would conserve this motif . The phenotype of the <infB ( del1 ) > mutant raises the question of what IF2-1 could be doing to influence cellular recovery after DNA damage by a PriA helicase-dependent pathway . First , IF2-1 and IF2-2/3 could have different preferences for mRNAs such that IF2-1 specifically promotes the translation of factors needed to support this pathway . Such a mechanism would be novel as such a role of the various IF2 isoforms in promoting differential gene expression has yet to be described . Second , IF2 may act as a transcription factor and the various IF2 isoforms may have different activity in this regard such that IF2-1 is specifically needed to regulate expression of genes needed for PriA helicase-dependent pathways . The finding that IF2 can selectively promote transcription of rRNA by RNA polymerase in vitro [26] and the identification of a region in the carboxy terminal region of IF2 with nonspecific DNA binding activity [27] have prompted the proposal that IF2 has activity influencing transcription . Third , IF2 has been shown to have molecular chaperone activity [25] . The IF2 isoforms may ensure that specific factors in their respective pathways are active when required . We have previously proposed a role of IF2 as a chaperone performing a function much like ClpX ( Figure 1B–1C ) where IF2 binds to a Mu end and prepares the DNA template for assembly of restart proteins , a process beginning with displacement of IF2 from DNA by PriA helicase . The analysis of this present work cannot definitively establish that any one of these possibilities is the basis for IF2's influence on cellular restart mechanisms; however , we favor the third mechanism in which IF2 acts as molecular chaperone , based on the role of IF2 in bacteriophage Mu replication in vitro [5] , [17] , the phenotype of the infB ( del1 ) mutant , and the relationship of this allele with priA300 . A key question regarding the function of IF2-1 is , why does its loss lead to a severe decrease in viability during growth on MMS despite the fact that the cell remains proficient for supporting replication restart ? We suspect that the loss of the preferred IF2 isoform for a restart mechanism , loss of PriA helicase activity , or the complete loss of PriA in the presence of a dnaC suppressor results in the inability to fine-tune the progression of restart pathways , a level of regulation that becomes essential when cells must grow under conditions that damage DNA at a high rate . If we speculate that the role of IF2 in Mu replication in vitro is applicable for cellular restart mechanisms , we can illustrate the type of regulation that IF2 might exert ( Figure 8B ) . An important difference between IF2-1 and the truncated forms IF2-2/3 for the assembly of restart proteins at stalled forks may be the mechanism by which they respond to a hypothetical go-ahead signal for restarting DNA replication . When DNA damage is accumulating at a relatively high rate , a mechanism that regulates restart by preventing re-establishment of the replication fork until the template is relatively free of DNA damage may ensure efficient DNA replication in the presence of a DNA-damaging agent . For example , restarting DNA replication before the DNA is relatively free of lesions will only result in the stalling of the fork again , causing delay in establishing a productive replication fork and thus inducing a high level of SOS response that may become toxic . These considerations are reminiscent of the findings of Flores et al . [44] , who determined that priA300 greatly diminishes viability of the holDG10 mutant . The holD gene encodes the Psi unit of DNA polymerase III holoenzyme , and the mutant Psi causes frequent replication fork stalling . That is , the effect of priA300 becomes discernible only when the rate of replication fork stalling becomes high . As noted by Flores et al . [44] , the deficiency in PriA helicase caused by the priA300 mutation may lead to the inability to promote duplex opening on the DNA substrate for DnaB helicase loading and replisome assembly [29]; alternatively , another function of PriA besides the helicase could be inactivated by the priA300 mutation , leading to the inability to cope with frequent fork arrest in the holDG10 mutant . The PriA function needed to sustain high viability of the holDG10 mutant may be related to the pathway in which both IF2-1 and PriA helicase play a role . When cells must grow in the presence of MMS , the action of PriA helicase to displace IF2-1 may play a critical function to ensure maximal cell viability , or conceivably , the inactivation or attenuation of another function by priA300 may prevent what we call the IF2-1/PriA helicase pathway from operating optimally . This example underscores the possibility that PriA helicase as well as IF2-1 play multiple roles for replication restart , some of which may be part of their mutual participation in the IF2-1/PriA helicase pathway and some of which may not . PriA helicase may play important roles in duplex opening for DnaB loading as well as displacement of IF2 to initiate replication restart , but only the latter may be essential for the IF2-1/PriA helicase pathway . The role of IF2 isoforms in influencing replication restart mechanisms has important implications for how replication restart and the maintenance of genome stability may be regulated with respect to cell physiology . As a translation factor , IF2 has a strong influence on cell growth and progression through the cell cycle while responding to cellular signals such as the alarmone ( p ) ppGpp [45] , which is an indicator of nutritional deprivation . Depending upon the physiological status , how replication restart is carried out can be critical in determining cell viability , and IF2 may respond to cellular signals to determine the conditions for restart . The IF2 function in translation is a highly conserved one found in all living cells [46] , [47] . Its role in influencing pathways for maintaining genome integrity prompts the question whether this general function has been conserved in other organisms to play some function in coordinating replication , recombination , and repair functions with respect to growth conditions .
All experimental analysis was conducted with derivatives of GTN932 ( Hfr del ( gpt-lac ) 5; see Table S1 ) , an E . coli K-12 strain that is a derivative of PK191 [48] . We have conducted PCR and sequencing analysis to verify that this line of strains have wild-type relA , not the relA1 allele [49] as sometimes reported for PK191 strains . The del ( priB ) 302 and priC303::kan alleles from JC19272 [39] , priA2::kan from PN104 [36] , del ( priC ) 752::kan from JW0456-1 , del ( dnaT ) 759::kan dnaC ( a491t ) from JW4336-2 , and del ( argA ) 743::kan from JW2786-1 [50] were introduced into bacterial strains by P1vir transduction as previously described [39] . Inheritance of del ( priB ) 752::kan was screened by PCR analysis with primers PriBupper and PriBlower ( Table S2 ) . The priA300 was introduced by P1 transduction , first transferring the metB1 allele by selecting for the closely linked btuB3191::Tn10 from CAG5052; the priA300 was then transferred from SS97 by selecting for Met+ transductants ( tetracycline-sensitive transductants were chosen ) [19] , which were screened by PCR amplification with primers PriA-Nseq and PriA-Cseq and sequenced with revPriA820 primer . The sulA::Mud ( lac , Ap , B::Tn9 ) from SS97 [18] or dinD1::Mud1 ( lac , Ap ) from PN104 [36] was introduced into strains by P1 transduction and selection on ampicillin plates; transductants were screened for disruption of the sulA or dinD genes with primers sulAupper and sulAlower or dinDupper and dinDlower , respectively . The clpX::kan strain was constructed as previously described [15] . The del ( infB ) 1::tet allele was constructed by first integrating a single copy nusAinfB operon into a random site of the host chromosome as part of the EZ-Tn5 transposon . The natural infB cistron was precisely excised and replaced with a tetR cistron from pACYC184 [51] , using recombineering methods [52] to generate the del ( infB ) ::tet allele . As recombination events at the natural infB site were very difficult to isolate , we created a PCR template to generate the del ( infB ) ::tet allele , with approximately 1-kb of DNA from upstream and downstream of infB to flank the tet cistron . This template on the pGEM-Teasy vector ( Promega ) was amplified using PfuUltra High Fidelity DNA polymerase ( Stratagene ) using the primers nusLower and rbfUp2 , and the PCR product was used to transform heat-induced DY330 flgJ::<nusAinfB-kan> . The various flgJ::<nusAinfB-cat> alleles were constructed by introducing infB mutations into the nusAinfB operon harbored on the EZ-Tn5 transposon . The transposon was from the pMOD-6<KAN-2/MCS> purchased from Epicentre , and it was introduced into DY330 as a transpososome according to the instructions of the manufacturer . The transposon was determined to be integrated in the flgJ gene by a single primer PCR and sequencing method [53] . For introduction of various infB alleles at the transposon site , the transposon was modified by recombineering [52] . Heat-induced DY330<KAN-2/MCS> was transformed with a PCR product made by amplifying the cat gene of pACYC184 with primers DelMOD6Cat and lowerKanCat ( see Table S2 ) . The resulting strain DY330<del ( kan ) ::cat ) > , which is chloramphenicol-resistant and kanamycin sensitive , serves as the strain for introducing various alleles at this site . PCR products for introducing the nusA infB operon at the transposon were made using pMOD-6<KAN-2/MCS> constructs as template . The nusA infB operon , amplified from the E . coli chromosome using PfuUltra High Fidelity with primers argRmetYp2 and IF2BamHI , was cloned between the SphI and XbaI site of pMOD-6<KAN-2/MCS> ( promoter side of the operon is proximal to the SphI site ) . Various infB mutations were introduced into the resulting plasmid . The operon was then amplified using primers lowerMod6Tn and antiSqRP , and the PCR product was used to transform heat-induced DY330<del ( kan ) ::cat ) > , selecting transformed cells on LB plates containing 25 µg/ml kanamycin and screening for chloramphenicol sensitivity . To construct versions of these flgJ::<nusAinfB> alleles that encode chloramphenicol rather than kanamycin resistance , heat-induced DY330 flgJ::<nusAinfBkan> strains were transformed with PCR products made by amplifying the cat gene of pACYC184 with primers upperKanCat and lowerKanCat . This inactivates the kan gene while leaving intact the nusAinfB contained within transposons . The resulting constructs were always verified by sequencing as described below . We could readily knock out the natural infB allele of a strain with the <nusAinfB ( wt , del1 or del2/3 ) > cassette by introducing the del ( infB ) ::TetR allele . As the expression of tetracycline resistance was relatively feeble from this allele , introduction of the knockout was most conveniently done by co-transduction with the closely linked argG; Arg+ transductants of a del ( argG ) 781::kan recipient strain co-inherited the del ( infB ) ::TetR allele at a frequency greater than 80% , provided that an infB allele which supports cell viability was provided from another site . Even when the second infB function was supplied by pSPCnusAinfB ( del2/3 , D501N ) , greater than 80% of the Arg+ transductants coinherited del ( infB ) 1::tet allele , indicating that the multicopy infB ( del2/3 , D501N ) can maintain cell viability ( the presence of the D501N mutation in the sole infB allele was verified by sequencing ) . When the second nusAinfB operon was present on the chromosome , it was introduced into the transposon inserted in flgJ . The various flgJ::<nusAinfB-cat> alleles were constructed by recombineering methods in DY330 as described above and transferred to other strains by P1vir transduction . The nusAinfB operon contained within the transposon includes all three ArgR binding sites ( see Figure 3A ) and extends to the stop codon for infB . As the nusAinfB operon in the transposon lacks downstream genes such as rbf in the natural operon , the infB alleles at the natural site and the transposon in flgJ can be separately amplified for DNA sequencing ( Figure 3A–3C; primers p1 and p2 for the natural site and p1 and p4 for the transposon site ) . Thus , the presence of infB at the natural site could readily be detected by primers ( p1 and p2 ) annealing to sites flanking infB to yield a 4 . 7-kb band ( Figure 3C , lanes 1 , 3 , and 9 ) , confirmed by 2 . 8-kb band yielded by one primer ( p3 ) annealing within infB and one ( p2 ) downstream of the gene ( lanes 2 , 4 , and 9 ) . ( See the list of primers in Table S2 . ) Knockout of the natural infB , in contrast , could be detected with the formation of a 3 . 2-kb band with primers p1 and p2 ( lanes 5 and 7 ) and no bands ( lanes 6 and 8; cf . with lanes 2 , 4 , and 10 ) with p3 and p2 . We found this to be the best method for constructing strains with various single-copy infB alleles , for the replacement of the wild-type infB allele at the natural site proved to be very difficult . As constructed strains were suspected to be potential restart mutants , their dnaC allele was sequenced to determine whether any suppressor mutations have accumulated there [39] . None of the mutants we isolated had as severe a phenotype as the priA null mutant , and no suppressor mutations in dnaC were detected . All pSPCnusAinfB plasmids with various infB alleles were constructed using the pBAD43 plasmid vector ( a gift from Dr . Jonathan Beckwith , Harvard Medical School ) [54] . This plasmid is a relatively low copy plasmid , having a pSC101 plasmid origin and conferring spectinomycin resistance . The nusAinfB operon , amplified by PCR using primers p1nusAinfB and IF2BamHI ( see Table S2 ) and PfuUltra High Fidelity DNA polymerase , was inserted into the NsiI-BamHI site of the pBAD43 vector . The ara and PBAD sequences required for arabinose-based gene expression by this plasmid were deleted by digestion with NsiI-BamHI and replaced with the nusAinfB operon , which begins downstream of the metYp2 promoter , including the last 5 nucleotides of the Fis binding site and ending with the stop codon for infB . As a vector control for the pSPCnusAinfB plasmids , pBAD43 was used . Construction of pBAD24 plasmids [55] that express IF2-1 , IF2-2 , and S-tagged IF2-2 ( S-IF2-2 ) has been described previously [5] . The plasmid for expressing S-IF2-1 was similarly constructed by amplifying the infB gene using primers Stag-IF2-1 and IF2BamHI , which introduce the S-tag coding sequence . The coding sequence was ligated into the NdeI-BamHI site of a pBAD24 vector whose NcoI site has been modified to an NdeI site . The priB and priC genes were cloned into pBAD24 , amplifying these genes using the NdeI-priB/PstI-priB and NdeI-priC/PstI-priC oligonucleotides and ligating into the NdeI/PstI site of the pBAD24 vector . Site-specific mutagenesis was carried out using the QuikChange Lightning Multi-Site-Directed Mutagenesis Kit purchased from Stratagene , using primers listed for this purpose in Table S2 . The infB ( del1 ) deletion was generated by amplifying the nusAinfB operon harbored on a plasmid vector with 5′-phosphorylated primers delIF2-1UP and delIF2-LOW ( see Table S2 ) , with PfuUltra High Fidelity and circularizing the linear PCR product with T4 DNA ligase . All mutations were verified by sequencing . ChIP analysis was conducted by modification of previously published procedures [56] , [57] . The major change was the incubation of cell lysate with 50 µg/ml RNase A at 37°C for 30 min just before the immunoprecipitation step . Additional details are described in Protocol S1 . Sensitivity of strains to MMS was measured both by direct plating on LB plates containing 6 mM MMS and by 15 min exposure to 0–18 mM MMS , the latter based on the procedure by Nowosielska et al . [34] . β-galactosidase activity was measured according to the procedure of Miller [58] . Mu was plated on LB plates at 37°C with 10 mM magnesium sulfate on a background of indicator cultures . Mu infective centers from thermoinducible lysogens were plated on a background GTN932 indicator at 42°C . Mucts62 lysogens were grown at 30°C . Cultures of priA2::kan strains were maintained in Davis minimal medium ( Difco ) containing glucose , thiamine , proline , and histidine , and the viable count was determined on plates containing the same media . All results from measuring MMS and UV sensitivity , homologous recombination proficiency , enzyme assays , and Mu plating efficiency are indicated with error expressed as the standard deviation from the mean ( at least three independent experiments; the number of independent experiments is indicated ) . See Protocol S1 for additional details .
|
Translation Initiation Factor 2 ( IF2 ) is a bacterial protein that plays an essential role in the initiation of protein synthesis . As such , it not only has an important influence on cellular growth but also is subject to regulation in response to physiological conditions such as nutritional deprivation . Biochemical characterization of IF2's function in replicating movable genetic elements has suggested a new role in the maintenance of genome integrity , potentially regulating replication restart . The parasitic elements exploit the cellular replication restart system to duplicate themselves as they transpose to new positions of the chromosome . In this process , IF2 makes way for action of restart proteins , which assemble replication enzymes for initiation of DNA synthesis . For the bacterial cell , the restart system is the means by which it copes with accidents that result in arrest of chromosomal replication , promoting resumption of replication . We present evidence for an IF2 function associated with restart proteins , allowing chromosomal replication in the presence of DNA–damaging agents . As the IF2 function is a highly conserved one found in all organisms , the findings have implications for understanding the maintenance of genome integrity with respect to physiological status , which can be sensed by the translation apparatus .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"genetic",
"mutation",
"enzymes",
"microbiology",
"escherichia",
"coli",
"prokaryotic",
"models",
"model",
"organisms",
"dna",
"replication",
"microbial",
"evolution",
"cell",
"growth",
"dna",
"molecular",
"genetics",
"dna",
"synthesis",
"microbial",
"physiology",
"gene",
"expression",
"dna",
"amplification",
"biology",
"molecular",
"biology",
"biochemistry",
"bacterial",
"physiology",
"cell",
"biology",
"nucleic",
"acids",
"protein",
"translation",
"genetics",
"dna",
"repair",
"dna",
"recombination",
"molecular",
"cell",
"biology",
"bacterial",
"evolution",
"genetics",
"and",
"genomics"
] |
2012
|
A New Role for Translation Initiation Factor 2 in Maintaining Genome Integrity
|
Dengue virus ( DENV ) infections are preferentially diagnosed by detection of specific IgM antibodies , DENV NS1 antigen assays or by amplification of viral RNA in serum samples of the patients . The type-specific immunity to the four worldwide circulating DENV serotypes can be determined by neutralization assays . An alternative to the complicated neutralization assays would be helpful to study the serotype-specific immune response in people in DENV hyperendemic areas but also in subjects upon DENV vaccination . In consecutive samples of patients with DENV-1- 4 infection type-specific antibodies were detected using an immune complex binding ( ICB ) ELISA . During incubation of serum samples and enzyme- labeled recombinant envelope domain III ( EDIII ) antigens immune complexes ( ICs ) are formed , which are simultaneously bound to a solid phase coated with an Fc–receptor ( CD32 ) . After a single washing procedure the bound labeled ICs can be determined . To further improve type-specific reactions high concentrations of competing heterologous unlabeled ED III proteins were added to the labeled antigens . Follow-up serum samples of 64 patients with RT-PCR confirmed primary DENV-1 , -2 , -3 or -4 infections were tested against four enzyme-labeled recombinant DENV EDIII antigens . Antibodies to the EDIII antigens were found in 55 patients ( sensitivity 86% ) . A complete agreement between the serotype detected by PCR in early samples and the serotype-specific antibody in later samples was found . Type-specific anti-EDIII antibodies were first detected 9–20 days after onset of the disease . In 21% of the samples collected from people in Vietnam secondary infections with antibodies to two serotypes could be identified . The data obtained with the ICB-ELISA show that after primary DENV infection the corresponding type-specific antibodies are detected in almost all samples collected at least two weeks after onset of the disease . The method will be of value to determine the distribution of the various type-specific anti–DENV antibodies in DENV endemic areas .
Dengue fever is a highly prevalent arthropod-borne viral disease with 2 . 5 billion people in tropical or subtropical areas at risk for infection . The clinical picture of dengue may vary considerably from mere fever to severe shock syndrome . The annual number of infections is estimated to several hundred million [1] , [2] . As four DENV serotypes exist , humans can be exposed to DENV infections several times . While dengue fever is usually associated with a rather low mortality , dengue hemorrhagic fever may give rise to serious and sometime lethal complications . It has been shown by several studies that dengue hemorrhagic fever is frequently but not always due to secondary DENV infection [3]–[5] . Therefore the detection of serotype-specific IgG antibodies would be of value to determine the immunological anti-DENV profile of an individual but also of a larger population in endemic countries . Knowing the serotype-specific antibody response , the risk of secondary infections with a new serotype can be predicted . Information on serotype-specific antibodies may also help to monitor the immune response after successful DENV vaccination [6] , [7] . Early after onset of acute DENV infection the serotype involved can be detected by RT-PCR [8]–[11] , or by NS1 antigen detection [12] , [13] . However , several weeks after onset of infection both methods will no longer give positive results . In contrast , even years after human infection , serotype-specific IgG antibodies can be detected by the plaque reduction neutralization test ( PRNT ) . However , up to several months after primary and even more after secondary infection subtype cross-reactivities are observed by PRNT [14] , [15] . Moreover , the PRNT is both time consuming and also difficult to handle , because the four different DENV strains have to be propagated in a BSL2 laboratory [16] and due to various technical details a standardization may be difficult to achieve [14] , [17] . Meanwhile it has been shown for many flaviviruses that upon acute infection type-specific antibodies to the domain III of the viral envelope ( EDIII ) are produced . EDIII is regarded as putative receptor binding domain [18] , [19] , and strong neutralizing antibodies against individual flaviviruses localize to EDIII [20] , [21] . Monoclonal antibodies binding to the EDIII of DENV envelope glycoprotein are the most efficient inhibitors of virus adsorption to Vero cells [20] , whereas monoclonal antibodies that cross-react with other flaviviruses localize primarily to the envelope domain II [22]–[24] . A humanized monoclonal antibody to loop A of DENV-2 EDIII was able to protect mice from lethal infection [25] . However natural variation of the DENV genotypes may influence the binding of neutralizing antibodies to EDIII proteins [26] . In addition neutralization is not confined to the EDIII [24] , [27] , [28] . Of 40 monoclonal antibodies recognizing DENV-2- infected cells , 14 bound to yeast cells displaying the DENV-2 EDIII domain . Most of these antibodies directed to the envelope EDIII domain are serotype-specific , but minor DENV cross-reacting and flavivirus cross-reacting epitopes were also identified on the EDIII protein [24] , [29]–[32] . Due to the low homology with other flaviviruses the EDIII domain of DENV has been frequently used as antigen to increase the specificity of DENV IgG [27] , [33] , [34] The anti-EDIII assays were considerably more specific than assays using larger DENV antigens or complete virions but had a sensitivity of only 80% in detecting acute DENV infections , which may be explained by the small number of epitopes on the EDIII protein [23] , [35] . A detailed analysis of the EDIII type-specific immune response both in primary and secondary DENV infections has been presented by Midgley et al . 2011 [36] . They used an indirect ELISA with the four EDIII antigens coated on solid phase . The inhibition was carried out by pre- incubating different dilutions of the serum samples with different dilutions of the four EDIII antigens . Without inhibition 25 samples of children with confirmed DENV-1 , -2 and -3 primary infection were studied . In two patients detailed results of the competitive assay were presented . 17+10 samples of children with DENV-1-4 secondary infection were also analyzed both with a competitive ELISA and with PRNT . While with primary infections the type-specific results were in agreement with PCR subtyping , in secondary infections only the memory response to the preceding DENV infection was found , supporting the theory of the “antigenic sin” . Recently , homodimeric EDIII proteins were applied for the discrimination of serotype-specific DENV IgM antibodies , but only IgM antibodies of DENV-1 infected patients showed clear subtype-specific reactivity [37] . All these investigations show that EDIII proteins may also be of diagnostic significance for the detection of serotype-specific antibodies to DENV . Since our earlier immunoblot technique to detect anti-DENV EDIII activity in serum samples of patients with DENV infection [38] could not differentiate between type-specific and cross-reacting anti-EDIII antibodies , we developed a more sensitive and even more specific immune-complex binding ( ICB ) ELISA , which has already been shown to detect antibodies to recombinant WNV and tick-borne encephalitis virus ( TBEV ) EDIII antigens with high specificity and sensitivity [39] , [40] . The ICB assay is controlled by two specific immunological mechanisms: first the antibodies to be determined react with the horseradish peroxidase ( POD ) -labeled antigen in the wells of microtiter plates and simultaneously the aggregated antibodies attach to an appropriate Fc-receptor such as FcγRIIa ( CD32 ) coated onto the solid-phase plates [41] . As an advantage of the ICB ELISA , POD-labeled and unlabeled , closely related antigens can be combined to competitively suppress unwanted cross-reacting antibody activities . Thus , by combining TBEV POD- labeled EDIII antigen with an excess of unlabeled WNV EDIII antigen , highly specific antibody assays could be established [40] , [41] . A similar combination of POD-labeled and unlabeled DENV EDIII antigens was introduced here , to establish epitope-specific antibody assays . As we had collected numerous consecutive serum samples of patients , for whom the DENV serotype had been determined by RT-PCR , we were able to assess the anti-EDIII response in DENV infected patients and to evaluate the serotype-specificity of the four ICB ELISAs by using the respective recombinant DENV EDIII proteins as antigens . In addition , we studied the distribution of DENV type-specific antibodies in serum samples of healthy subjects from Vietnam , a hyper-endemic area with many secondary infections [42] , [43] .
From all patients in this study informed written consent for the detection of antibodies had been obtained prior to processing the samples . In minor age patients ( <18 years ) informed consent was given on his or her behalf by a parent . The collection of serum samples of patients with dengue fever was approved by the Ethics Committee of the Ärztekammer Hamburg ( WF-024/11 ) . All data on human subjects were analyzed anonymously . Clinical investigations have been conducted according to the principles expressed in the Declaration of Helsinki . Studies on sera from healthy Vietnamese people had been approved by the Scientific Council of Education , Training and Ethics of Hué Medical School on September 11 , 1998 [44] . During the last ten years consecutive samples of 64 European tourists with acute dengue fever ( 20–55 years of age m/f: 1 . 2 ) had been collected during our routine diagnostics [31] . All samples had been aliquoted and stored at −20°C . In early samples no antibodies to DENV were detected , while the DENV serotype could be identified using a real time RT-PCR protocol [9] . According to the PCR results 25 , 15 , 17 and 7 sera were obtained from DENV-1 , DENV-2 , DENV-3 and DENV-4 infected patients , respectively . Subsequent serum samples taken at least one week after onset of fever were available from all 64 patients . In all secondary samples IgM and IgG antibodies to DENV were demonstrated by indirect immunofluorescence ( IIF ) using virus infected Vero cells [38] . In addition in 2012 , 5 mL serum per patient could be obtained from seven patients with PCR confirmed DENV infection during the convalescent phase . Thus sufficient material was available for DENV neutralization assays . 12 samples of patients with WNV infection ( 8 North Americans and 4 Germans ) , 5 samples of Japanese encephalitis virus ( JEV ) vaccinees , 11 of yellow fever virus ( YFV ) vaccinees and 20 samples ( 10 clinical cases and 10 vaccinees ) of subjects with antibodies to TBEV were derived from our routine diagnostics . The presence of antibodies to the respective viruses had been confirmed using our highly specific ELISAs [39] , [41] and IIF . To test the specificity of the DENV ICB ELISA serum samples of 88 subjects of our routine diagnostics ( 25–59 years , m/f: 1 . 5 ) were included . The samples diluted 1∶10 in phosphate buffered saline ( PBS ) had no antibodies to DENV as shown by IIF . In addition , 87 serum samples of healthy Vietnamese people ( 18–65 years m/f: 1 . 5 ) had been obtained during studies on amoebiasis and dengue fever at the city of Hue in 1999 . 71 samples contained antibodies to DENV as shown by IIF ( titer 1∶>20 ) . For the detection of the four DENV serotypes in early serum samples of infected patients a real time RT-PCR protocol with four primer-probe sets in a single reaction mixture was applied [9] . The RT-PCR assays were run on a LightCycler 480 System ( Roche , Mannheim , Germany ) . A mixture of all four DENV RNAs served as positive and inhibition controls . DENV micro-neutralization [14] , [17] , [27] , [31] was performed in Vero E6 monolayers grown in 96-well plates . The cells were infected with the same DENV strains that had been used for cloning and expression of the ED3 proteins . DENV immune sera were heat- inactivated , serially diluted in twofold steps ( 1∶20 to 1∶10 , 240 ) in nutrition medium and the dilutions were preincubated in triplicate with 50 plaque forming units determined for each DENV strain in a final volume of 100 µl for 1 h at 37°C . The serum/virus mixtures were added to the cells for 1 h at 37°C . Then the wells were emptied and filled with cell culture medium ( MEM; 2% FCS ) to a total volume of 250 µl containing 0 . 75% methylcellulose . After incubation for 96 h at 37°C in 5% CO2 the cells were washed , fixed with 5% formalin in PBS and permeabilized with 0 . 5%Triton ×100 in PBS . The cells were immunostained with a cross-flavi mouse monoclonal antibody ( 2F1 ) , followed by POD-labeled anti-mouse antibody and precipitating TMB . Dilutions were plotted against spot counts in triplicate and 50% plaque reduction neutralization was calculated by nonlinear dose–response regression analysis ( Prism 6 Package , GraphPad Software , Inc . , San Diego , CA ) . For the production of the four recombinant DENV EDIII proteins the following dengue strains were used: DENV-1 West Pac , Genbank Accession–No . U88535; DENV-2 New Guinea C , Genbank-Accession-No . AF038403; DENV-3 , H87 Genbank-Accession-No . M93130; DENV-4 Thai1978 , Genbank-Accession-No . U18441 . The four EDIII antigens of DENV-1-4 ( aa 297–400 ) of the E protein were cloned in expression vector pQE30 with an N-terminal His- tag [38] . Upon IPTG-induced expression of the His-tag fusion proteins , the transformed cells ( E . coli JM109 , Promega , Mannheim , Germany ) were dissolved in 8M urea and the EDIII proteins purified by nickel affinity chromatography as described earlier [38] . The antigens were eluted using a pH gradient . After renewed binding to fresh Ni–NTA and matrix- associated refolding a second elution with PBS containing 0 . 25 M imidazole , 5% glycerol was performed . The material was stored in 20% glycerol at −20°C . Purity of refolded antigens ( about 0 . 4–0 . 8 mg/mL ) was controlled by SDS-PAGE and Coomassie staining . The extracellular part of CD32 ( FcgRIIa H131 ) [45] was cloned into expression vector pJC45 [46] using amplification with the primers 5′ACGCATATGGGACTTGAAGTCCTCTTTCAGGGACCCGGGCAAGCTGCA GCTCCCCCAAAG-3′ and 5′-ACCGGAATTCTTAGATCCCCATTGGTGAAGAGC-3′ and restriction enzymes NdeI and EcoRI , respectively . After heat shock transformation in E . coli BL21 ( pAPlacIQ ) , IPTG-induced expression of the His-tag fusion protein was performed at 18°C over night . Cells were lysed by lysozyme and sonification [47] ( modified ) . Washed inclusion bodies were solubilized in 6M guanidine hydrochloride , 50 mM Tris-HCl , pH 7 . 8 and the recombinant protein was purified by Ni-NTA under denaturing conditions . Protein fractions were combined , refolded by rapid dilution and concentrated in centrifugal filters . The His-tag was cleaved off by digestion with 3C protease ( GE Healthcare , Stockholm , Sweden ) at 4°C over night . To remove the protease and cleaved off tags further purification was performed using gel filtration chromatography with an ÄKTA pure FPLC system ( GE Healthcare , Stockholm , Sweden ) . Purity was visualized by SDS-PAGE with silver staining and with Western blotting using an anti-human CD32 polyclonal antibody ( R&D Systems , Minneapolis , USA ) . Purified CD32 was bound to microtiter plates ( Immunolon , Nunc , Wiesbaden , Germany ) at a concentration of 5 µg/mL in PBS containing 1 mg/mL NaN3 and was incubated on the plates for at least three days at 4°C before use . Sealed microtiter plates could be stored at 4°C for two months or at −20°C in the presence of 25% glycerol for at least one year . The antigens were directly labeled with POD as described earlier [39] . 4 mg POD type IV ( Sigma Aldrich , Taufkirchen , Germany ) in 0 . 5 mL distilled water were activated using 0 . 1 mL 0 . 1 M sodium periodate . After dialysis against 1 mM acetic acid buffer ( pH 4 ) 1 mg of recombinant DENV EDIII antigen were added at pH 9 . 3 . The mixture was threefold concentrated . After incubation at 4°C over night , the antigens were diluted 1∶10 in PBS containing 1% BSA , 1% fetal calf serum and 20% glycerol . Addition of NaBH4 was not necessary . The POD-labeled , 1∶10 diluted DENV-1 , -2 , -3 and -4 EDIII antigens ( DeP1-4 ) could be stored at −20°C for at least two years without any loss of activity . All DeP1-4 antigens were applied in the ICB ELISA at a dilution of 1∶≥16 , 000 ( 1–2 nM of antigen ) using a diluent containing 1% BSA , 1% fetal calf serum and 1% detergent ( Triton ×100; Sigma Aldrich , Taufkirchen , Germany ) in PBS . To increase the serotype-specific reactions , i . e . to eliminate antibody cross-reactions due to cross-reacting epitopes on all DePs , an excess of heterologous , unlabeled DENV EDIII antigens ( De1-4 ) was added to the labeled DeP1-4 . Using the ICB ELISA with DeP1 , DEP2 , DeP3 and DeP4 as antigens and the respective positive control sera the amount of unlabeled De 1 , 2 , 3 and 4 antigens for complete autologous inhibition was determined . A 100-fold excess of unlabeled to labeled antigen was usually sufficient to suppress the reaction with the positive control sera . Besides , using varying concentrations of the respective autologous De1 , 2 , 3 and 4 antigens the inhibitory strength of newly produced De batches could be quantified and compared . Similarly , a 100-fold excess of De3 added to DeP1 , 2 , 4 ( w/w ) and of De1 added to DeP3 antigen was able to competitively block subtype-DENV cross-reactivity . For instance the addition of a final dilution of 1∶160 of De3 ( 1 mg/mL ) to the DeP1 ( containing 1 mg/mL De ) diluted 1∶16000 was sufficient to obtain complete inhibition of subtype reactivity . Moreover , to avoid false positive reactions due to antibodies to the His-tag , to POD or E . coli proteins , POD-labeled EDIII TBEV antigen ( TBEP ) [40] was added to the DePs . The competitive addition of the labeled antigen required the irreversible inactivation of the enzymatic activity of POD . To this end 1 mL of the labeled TBEP was incubated with 0 . 1 mg NaN3 and 0 . 035% H2O2 [48] for at least three days at 4°C followed by dialysis over night against PBS containing 10% glycerol . The inactivated antigen ( iTBEP ) did no longer show enzymatic activity at a dilution of 1∶100 ( no reaction after 10 minutes with TMB substrate solution ) . The antigenicity of iTBEP was unaffected by the NaN3 treatment , as was documented by competitively adding iTBEP to TBEP using the TBEV ICB ELISA [41] . A complete inhibition of the reaction of the positive control was seen in the presence of iTBEP up to a dilution of 1∶400 . The ICB ELISAs were performed as described earlier for the detection of antibodies to the EDIII antigens of WNV and TBEV [41] . Prior to testing the microtiter plates coated with CD32 were rinsed three times with washing buffer ( 100 mM Tris , 150 mM NaCl , 0 . 05% Tween 20 ) . Protein blocking of the plates was not required . Human sera were diluted 1∶10 in PBS containing 0 . 5 mL/L ProCline 300 ( Bellafonte , PA , USA ) and 10 mg/L phenol red for better visualization . In each assay a positive , and two negative control serum samples were included to monitor interassay variation . To control for false positive results , one of the negative samples contained anti-HIS-tag antibodies , which are frequently found in sera of patients with falciparum malaria [40] . Aliquots of positive control sera for each of the four assays were stored at a dilution of 1∶20 at −20°C . Each 25 µl of diluted serum and 25 µl of the labeled ED3 antigen ( DeP; diluted 1∶≥16000 ) were added to the wells of the microtiter plate and the mixture was incubated at 4°C over night . Finally , plates were washed three times with the washing buffer and stained with 50 µl TMB ( KPL , Gaithersburg , MD USA ) per well at room temperature for 10 min . After adding 100 µl 1N sulphuric acid optical density ( OD ) was read at 450/620 nm . For each of the four ICB ELISAs the cut-off values ( mean OD+3σ ) were calculated using the 88 negative samples . In addition , for each ICB ELISA accuracy , optimum sensitivity and specificity were determined using the Two Graph Receiver Operating Characteristic ( TG-ROC ) curve analysis ( MedCalc statistical software B-8400 Ostend , Belgium ) . The OD values of 110 negative samples ( from 88 blood donors and from 22 subjects of Table 1 ) served as true negatives ( criterion 0 ) and the OD signals of the 25+15+17+7 RT- PCR positive samples of the tourists as true positives ( Criterion 1 ) . The results of the ICB ELISA are presented as positive/negative ( P/N ) ratio by dividing the OD of the sample by the OD of the cut-off value . P/N values≥1 were considered as positive . Intra- and interassay coefficients of variation ( CVs ) were calculated for each ICB ELISA ( ( Prism 6 Package , GraphPad Software , Inc . , San Diego , CA ) using our four positive control sera in rows of sixfold , tested on three consecutive days . DENV-1 West Pac , Genbank Accession–No . U88535; DENV-2 New Guinea C , Genbank-Accession-No . AF038403; DENV-3 , H87 Genbank-Accession-No . M93130; DENV-4 Thai1978 , Genbank-Accession-No . U18441
Samples from 88 patients , who had never entered DENV endemic areas and did not report any flavivirus vaccinations , were tested using the four DENV EDIII ICB ELISAs and IIF ( Table 1 ) . All 88 samples were negative in IIF , but five samples were positive in the four ICB ELISAs ( Table 1: third column ) . These false positive reactions were probably due to antibodies to the His-tags [40] , to traces of E . coli contaminations or to the POD in the DePs . When all samples were retested using the DeP1-4 antigens in the presence of a 100-fold excess of the iTBEP EDIII antigen ( see material and methods ) , all 88 samples were negative ( forth column; + iTBEP . The addition of iTBEP did not significantly alter the reaction of anti-DENV positive serum samples . Therefore all further DENV ICB-ELISAs were performed in the presence of a 100-fold excess of iTBEP antigen . Although we knew from earlier experiments that using the ICB ELISA with EDIII proteins as antigens flavivirus cross-reacting antibodies rarely occur [41] , we tested 12 samples of patients with WNV infection , 20 samples of subjects with anti-TBEV antibodies , five samples of JEV- and 11 samples of YFV vaccinees with both the ICB ELISA and IIF ( Table 1 ) . All 48 samples were negative ( P/N ratio≤1 ) with all four DENV serotype-specific ICB ELISAs ( specificity 100% ) . In contrast , using anti-DENV IIF , 26/136 samples ( Table 1 , last column ) were false positive ( all 12 anti-WNV samples , 7 sera with anti-TBEV antibodies , 3 sera of JEV vaccinees and 4 samples of YFV vaccinees ) . Over several years consecutive serum samples of 64 tourists with acute DENV infections had been obtained . All patients had primary infections ( no anti-DENV IgG antibodies in the early PCR positive sample [49] ) . To confirm the DENV type-specific results subsequent samples were available 6–15 days after onset of fever . In 55 of these 64 subsequent samples antibodies to at least one of the four DeP antigens were found ( overall sensitivity 86% ) . However , out of 12 samples taken more than 10 days after the onset of the disease 11 were positive . Obviously , antibodies to EDIII come up late during the course of the disease . However , they seem to persist for prolonged periods , because in two PCR positive patients with DENV-1 and -2 infection , where diagnostic samples were obtained three and six months after onset of infection , the initial antibody reactivity to the DeP antigens had not decreased . To obtain additional information on the accuracy , sensitivity and specificity of each of the four ELISAs a Two Graph Receiver Operating Characteristic ( TG-ROC ) analysis was applied ( see supporting information , figure S1 ) . The analysis was carried out using the OD values of the 110 negative samples ( criterion 0 ) and the 64 DENV PCR positive samples ( criterion 1 ) . The cut-off levels used in this study based on mean OD+3σ of the 88 blood donor samples resulted in sensitivities for the DeP1-4 ICB ELISAs between 80% and 100% and specificities between 94% and 100% . But due to the relatively small number of positive samples a wide range for the 95% confidence limits ( Cl ) was obtained . The accuracy of all four tests was between excellent and good ( ROC>0 . 85 ) . DENV serotype-specific reactions using the 55 anti-DeP positive samples are shown in figure 2A–D . The tests were run without competition by De antigens . 24 , 11 , 13 and 7 samples were positive with DeP1 , 2 , 3 and 4 antigens , respectively . The reactivity ( P/N ratio ) of the 24 samples of PCR-confirmed DENV-1 infected patients with the DPe1 antigen ranged between 1 . 3 and 23 . 2 ( figure 2A ) . The majority of the 24 sera did only weakly react with the other DeP antigens . Only in sera with strong anti-DeP1 reactivity DENV major subtype cross-reactions , preferentially with DeP3 , were observed . However , the reactivity ( P/N ratio ) with the homologous DeP antigen exceeded by far the signals observed with heterologous DeP2 , 3 , or 4 antigens ( figure 2A: patient 8; encircled data points ) . Likewise , in 11 patients with proven DENV-2 infection ( figure 2B ) , in 13 patients with DENV-3 infection ( figure 2C; patient 5 with strong cross-reactivity; boxed data points ) and in 7 patients with DENV-4 infections ( figure 2D ) the antibody reactions with homologous DeP clearly exceeded that obtained with the heterologous antigen . Overall , cross-reacting responses were seen in about 25% of the samples . The 220 OD values of the 55 positive samples are shown in Table S2 . To further reduce DENV subtype cross-reactions observed with some highly reactive serum samples , competitive ICB ELISAs were performed . Examples of the competitive assays are shown in figure 3A and B using sera of two patients , one with DENV-1 ( encircled in Fig . 2A ) and the other with DENV-3 infection ( boxed in Fig . 2C ) . Using the DENV-1 positive serum a P/N ratio of 21 . 1 and a cross-reactivity ( P/N ratio 11 . 5 ) was seen with DeP1 and DeP3 , respectively , without addition of competitors ( Fig . 3A; no comp . ) . Addition of unlabeled De3 protein did not significantly alter the positive signal of the DENV-1 serum with DeP1 , but minimized the cross-reactive signal with DeP3 . A similar result is shown in figure 3B for an exemplary DENV-3 serum . A complete suppression of the cross-reaction with DeP1 on competition with De1 was observed , while the strong reaction with DeP3 was unchanged . Based on this assay principle the exact amount of competitor needed to completely suppress type- and subtype-specific reactivity could be determined . For most unlabeled antigens an excess of De over DeP of 100-fold ( w/w ) was sufficient . By combining highly diluted , labeled DeP1-4 antigens with low dilutions of unlabeled heterologous EDIII antigens ( for example DeP1 , DeP2 and DeP4 combined with a 100-fold excess of De 3 , and DeP3 with an 100-fold excess De1 ) all subtype-specific cross-reactions , as shown in figure 2A–C , could effectively be reduced to P/N ratios≤1 without altering the type-specific reactions by more than ±10% ( see also figure 4 and supporting information , figure S2 ) . In conclusion , using the ICB ELISA the serotype-specific antibody response in all 55 primary DENV infections was in full agreement with the serotype as determined by RT-PCR . In four patients with PCR confirmed DENV-1 , -2 and -3 infection consecutive serum samples could be collected . The antibody response of a DENV-1 infected patient to all four DePs on days 3 , 9 and 26 is shown in figure 4 A . The ICB ELISA with competition ( filled symbols ) and without competition ( open symbols ) was applied . The strong cross-reactivity with DeP3 ( open diamonds ) could almost completely be removed by adding De3 to DeP1 ( filled diamonds: anti-DeP1+De3 ) , while the addition of heterologous antigen did not reduce but rather enhanced the autologous antibody response to DePs . In a patient with DENV-2 infection ( figure 4B ) almost no subtype-reactivity with DEP1 , 3 , 4 was seen . Again , the addition of De3 to DeP2 did slightly increase the type-specific response on day 200 . In figure 4C and D antibody reactions of two consecutive samples of another DENV-1 and of a DENV-3 infected patient are presented . As stated before , type-specific antibodies were only detected after the first week after onset of the disease , while in samples taken during the first week high IIF antibody titers to DENV were already present . The OD values used for figure 4 A–D are listed as supporting information in Table S3 . Sufficient amounts of serum for PRNT testing could be obtained from seven DENV infected tourists . A comparison of the P/N ratios , IIF titers and PRNT titers of these serum samples are shown in Table 2 . Both the PRNT titers and the results of the competitive CB ELISA confirmed the serotype obtained by RT-PCR , but the PRNT showed strong cross-neutralization especially with the DENV-2 samples . The results obtained with the seven samples taken several weeks or months after onset do not support a close correlation between IIF- , PRNT titers and ICB ELISA-derived P/N ratios . To learn more about the distribution of DENV serotypes in an endemic region , additional sera of 87 adult healthy Vietnamese people were analyzed using the four DENV ICB ELISAs . No PCR data were available . 71 of the 87 samples ( 82 , 6% ) were positive by IIF , but false positive reactions due to cross-reactions with other flavivirus antigens could not be excluded . 57 of the 71 IIF-positive samples ( 80% ) reacted with at least one of the four DeP- antigens . Without competition 35 samples showed DENV subtype cross-reactions with three to four DeP antigens . Therefore all 57 samples were retested in the presence of an excess of heterologous antigen , i . e . DeP1 , DeP2 and DeP4 antigens were combined with a 100-fold excess of De3 while DeP3 was combined with the same excess of De1 . Using the competitive version of the ICB ELISA 45 of the 57 subjects ( 79% ) produced positive reactions ( P/N ratios>1 ) with one antigen only . 14 , 14 , 11 and 6 samples reacted with DeP1 , DeP2 , DeP3 and DeP4 , respectively ( figure 5 ) . Therefore the most prevalent serotype- specific antibodies were directed to DENV-1 ( DeP1 ) and DENV-2 ( DeP2 ) followed by DENV-3 ( DeP3 ) and DENV-4 ( DeP4 ) . Provided that the detection of two or more DENV type-specific reactions in a serum sample mean secondary infection , 12 subjects ( 21% ) with secondary infections were found . Four reacted with both DeP1 and 3; three with DeP3 and 4; two with DeP1 and 4; three with both DeP1 and 2 . To exclude residual cross-reactions the 12 samples were also tested at a serum dilution of 1∶100 . The P/N ratios obtained with all four antigens are shown in Table S3 . Remarkably , among the subjects with secondary reactions 5 of 12 showed an anti– DeP4 response . Reactions with three or four antigens were not observed .
The detection of serotype-specific anti-DENV antibodies can be of value to characterize the complex anti-DENV immunity not only in individuals but also in larger groups in endemic areas . DENV type-specificity can be assessed by methods like RT-PCR or PRNT . However , this is only possible in samples taken either early ( RT-PCR ) or several months after infection ( PRNT ) to avoid cross-reactions . Therefore we looked for a simple but reliable alternative , which would allow the detection of DENV serotype-specific antibodies in samples of both patients and healthy subjects . As we have shown here , the ICB ELISA , using four POD-labeled EDIII proteins ( DeP1-4 ) as antigens , could reliably detect DENV serotype-specific antibodies in reconvalsescent patients with primary infection . Our data indicates , that serotype- specific IgG antibodies , reacting with DeP antigens and simultaneously with CD32 , are detected only approx . two weeks after onset of fever , but obviously persist for several years , similar to DENV neutralizing antibodies . Possibly low avidity antibodies , produced shortly after onset of the disease , are not detected using the ICB ELISA , since CD32 preferentially binds high avidity antigen-antibody complexes [45] . The delayed antibody detection may partially explain the relatively low sensitivity ( 55 of 64 cases were detected ) with samples taken shortly after onset of the disease . In acute infections a sensitivity of 87% to 93% was obtained by combining an NS-1 antigen test with IgM and IgG detection [50] , [51] . Using various EDIII constructs as antigens for IgG or IgM antibody detection [23] a sensitivity of about 80% and specificity of almost 100% was obtained [34] . The small number of epitopes on the EDIII protein may explain why anti- EDIII antibodies are generally present at low levels in human immune sera [23] . In contrast to publications using EDIII proteins as antigens for the rapid diagnosis of acute DENV infections recent efforts were focused on the detection of DENV sero-type specific antibodies [37] . Midgley et al . [36] have performed competitive ELISAs , where multiple dilutions of serum and EDIII proteins were preincubated before the mixtures were applied to EDIII coated microtiter plates . Their data convincingly show that during primary DENV infection the maximum activity is directed against the homologous EDIII antigen of the infecting serotype . Our data on primary infections are in full agreement with their findings . However , using the ICB ELISA we found less cross-reactivity in our samples of patients with primary infection . Without competition only 15 of the 55 sera of our patients reacted with heterologous EDIII antigens and the cross-reactivity mainly confined to DeP1 and DeP3 antigens . The reason for this difference is unclear but may be due to different strategies used for both ELISAs . In contrast to indirect ELISAs the ICB ELISA preferentially detects high avidity antibodies via CD32 binding . Also the EDIII proteins bound to aggregated POD molecules may show an altered antigenicity . Finally , the presence of iTBEP in the ICB ELISAs , to avoid background reactions , may play a role . On the EDIII domain , type-specific , subcomplex-specific and even a flavivirus-cross-reacting epitope have been recognized . The type-specific epitopes are located on the lateral ridge of the EDIII , as had been shown using mouse monoclonal antibodies [29] , [31] . The flavivirus cross-reacting epitope , residing at the AB loop of EDIII , is predicted to have limited accessibility on the mature virus [29] . This may partially explain , why flavivirus cross-reactivity was not observed in our earlier investigations [41] . The interpretation of the ICB assay may be more difficult with samples of subjects with secondary DENV infection and due to repeated stimulation of the immune system antibodies to cross-reacting epitopes may play a major role . Thus , for studies in DENV-endemic countries with many secondary infections the competitive variant of ICB-ELISA should be applied , using the addition of heterologous EDIII ( De1-4 ) antigens to eliminate cross-reacting antibodies directed to the subtype-specific epitopes on the DePs . As CD32 detects human and mouse antibodies equally well [41] , [45] mouse monoclonal antibodies to subtype-specific EDIII epitopes might help to exclude residual cross-reactivities of the competitive ICB assays . The low background reactions seen with the ICB ELISA depend on the use of highly labeled DeP antigens , which could be used at dilutions of 1∶≥16000 . The high dilution was essential for the competitive assays , because the competitor proteins had to be added at a rather low dilution of only 1∶160 ( 100–200 -fold excess ) . However , due to the strong labeling of the antigens non-specific antibodies to E . coli proteins , His-tags or even the POD itself may occasionally be detected [41] . These non-specific reactions were suppressed successfully by adding similarly labeled but enzymatically inactive EDIII proteins such as iTBEP . In the presence of non-specific antibodies the addition of competitive antigens like De1-4 or iTBEP may occasionally result in large amounts of unlabeled immune complexes competing with the labeled ones for CD32 . To avoid blocking of CD32 by unlabeled immune complexes , a high concentration of CD32 ( 5 µg/mL ) was applied to the solid-phase . Therefore we had to start our own production of recombinant human CD32 to test hundreds of samples using four antigens per sample both with and without competition . The results on serum samples collected in Vietnam in 1999 suggest that all four DENV serotypes have been circulating in that area for many years . Reports on the presence of all four DENV serotypes in Vietnam [42] , [43] , [52] , [53] support our findings but the low proportion of only 21% secondary infections is in contrast to earlier data obtained with PRNT on the presence of secondary infections in Thailand [53] , [54]} . Accordingly , our preliminary results obtained with samples of healthy Vietnamese subjects need further confirmation . In particular , a strong production of subtype-specific antibodies during DENV co-infections or subclinical infections [5] , [43] has to be taken into account . The data of Midgley et al [36] suggest that during secondary DENV infection no type-specific antibodies to the new serotype can be detected , while only the primary type-specific and subtype-specific response is stimulated . Therefore several DENV infections in the Vietnamese subjects may have been misinterpreted as primary infections , when only the anamnestic response was detectable . However , we could identify two different type-specific responses in 21% of the DENV infected Vietnamese subjects . The age of our subjects may have played a role , who in contrast to children studied by Midgley et al . had a greater chance to experience numerous DENV infections , which may eventually trigger additional type-specific antibodies . To prove the existence of different DENV type-specific responses in humans , additional adult patients with secondary DENV infection have to be studied . Patients showing a positive RT-PCR and containing high IgG antibody levels but no IgM antibodies to DENV [8] , [49] are highly indicative of secondary infection . Follow-up samples of such patients can possibly be obtained during investigations in hyper-endemic areas . The ICB ELISA can be applied for high throughput testing , because only a single incubation step is required . The method may help to extend our knowledge of the epidemiology of DENV infections in tropical areas .
|
Infections with four different dengue viruses are threatening 2 . 5 billion people in tropical countries . Since most antibodies to these four viruses are cross-reacting , a type-specific ELISA would be valuable to study the immune response to the circulating viruses in patients but also in healthy subjects in endemic counties . Therefore a novel DENV immune complex binding ( ICB ) ELISA was developed to detect serotype-specific antibodies to all four dengue virus serotypes in human serum samples . The tests use labeled recombinant EDIII antigens of the four DENV strains . Numerous samples of patients with RT-PCR confirmed dengue fever were assessed by the new method . In samples of 55 patients with primary dengue fever full agreement between the serotype detected by RT-PCR and the serotype-specific antibody based on the ICB ELISA was obtained . The type-specific antibodies were not observed before the second week of illness . Our data suggest that using the ICB ELISA in healthy adult subjects in an endemic region ( Vietnam ) both primary and secondary infections can be identified . The method may help to analyze the distribution of the four dengue viruses in the tropics .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2013
|
Detection of Serotype-Specific Antibodies to the Four Dengue Viruses Using an Immune Complex Binding (ICB) ELISA
|
Riboswitches sense cellular concentrations of small molecules and use this information to adjust synthesis rates of related metabolites . Riboswitches include an aptamer domain to detect the ligand and an expression platform to control gene expression . Previous structural studies of riboswitches largely focused on aptamers , truncating the expression domain to suppress conformational switching . To link ligand/aptamer binding to conformational switching , we constructed models of an S-adenosyl methionine ( SAM ) -I riboswitch RNA segment incorporating elements of the expression platform , allowing formation of an antiterminator ( AT ) helix . Using Anton , a computer specially developed for long timescale Molecular Dynamics ( MD ) , we simulated an extended ( three microseconds ) MD trajectory with SAM bound to a modeled riboswitch RNA segment . Remarkably , we observed a strand migration , converting three base pairs from an antiterminator ( AT ) helix , characteristic of the transcription ON state , to a P1 helix , characteristic of the OFF state . This conformational switching towards the OFF state is observed only in the presence of SAM . Among seven extended trajectories with three starting structures , the presence of SAM enhances the trend towards the OFF state for two out of three starting structures tested . Our simulation provides a visual demonstration of how a small molecule ( <500 MW ) binding to a limited surface can trigger a large scale conformational rearrangement in a 40 kDa RNA by perturbing the Free Energy Landscape . Such a mechanism can explain minimal requirements for SAM binding and transcription termination for SAM-I riboswitches previously reported experimentally .
Riboswitches reveal the versatility of Ribonucleic Acid ( RNA ) folding , and its remarkable biological impact . They are folded mRNAs that sense cellular metabolite levels and control expression of downstream genes [1]–[4] . Design of altered or novel riboswitches has been suggested for bioengineering applications [5]–[9] . Riboswitches also represent an important target for the design of novel antibacterials [10]–[12] . Riboswitches contain an aptamer , which recognizes and binds the metabolite . This binding triggers conformational rearrangement of the expression platform , which controls gene expression . Like other transcriptional riboswitches , the SAM-I riboswitch secondary structure is rearranged upon ligand binding [2] , [13] , [14] . The P1 and terminator ( T ) helices form in the ligand-bound state ( Figure 1 ) . This bound state is called the transcription OFF state since the terminator stops transcription . Without ligand the antiterminator ( AT ) helix forms , preventing formation of P1 and T helices , and allowing transcription ( the transcription ON state ) . SAM-I and other riboswitches raise the question–how can a small molecule binding to a limited contact surface cause a major folding rearrangement of a much larger RNA ? Addressing this question requires consideration of conformational dynamics . X-ray studies of riboswitches have largely focused on the ligand-bound aptamer , truncating the expression domain to suppress conformational dynamics [15]–[18] . Such dynamic behavior is problematic for high resolution structure determination . All-atom MD simulations are a major workhorse to tackle conformational dynamics [19]–[25] . Such methods have been applied to riboswitches , working largely with aptamer X-ray coordinates . These studies have revealed further insights into ligand recognition [26]–[30] , the role of ions [25] , [31] , and contrasted dynamic properties in the liganded and unliganded states [27] , [29] , [30] , [32] , [33] . Larger-scale , slower dynamic processes have required coarse-grained modeling or directed simulations using biased force fields [34] , [35] . Until recently , however , all-atom MD simulations using unbiased force fields have been generally limited to time scales less than microseconds . The birth of a specialized machine designed for MD simulation-Anton [36] , [37] has increased the timescale limitation up to 200 times compared to simulations with conventional High Performance Computing ( HPC ) machines . Recent advances in software development and RNA structure modeling have improved the building of RNA models [38]–[48] . Together with enhanced sampling techniques [49]–[52] these modeling tools extend the accessible conformational space beyond that available within even the extended MD timescales . Here we employ all-atom MD simulations to observe the direct effects of ligand binding on the equilibrium between alternative SAM-I riboswitch base pairing configurations . A large gap remains between the timescale required for strand migration ( perhaps ms-seconds [53] , [54] ) and even the extended Anton MD timescale . We bridge this gap by bypassing the “nucleation” step in strand migration to simulate propagation-presumably a more rapid step . To generate “pre-nucleated” starting models for intermediate states , we capitalize on the recent advances mentioned above for sampling of RNA conformations . We focus on the relationship between SAM binding and P1 helix propagation , or strand migration from an AT to a P1 helix ( also termed the “switching” event [29] or “conformational collapse” [55] ) . We observed a strand migration event in the presence of SAM converting 3 AT helix base pairs ( characteristic of the unbound riboswitch ON state ) to competing P1 helix base pairs ( characteristic of the OFF state ) . Overall , our simulations predict that SAM perturbs the reduced Free Energy Landscape ( FEL ) in a manner that favors conformations with expanded P1 helix base pairing and reduced AT pairing within the competing region , for certain starting geometries . Based on this simulation , we propose a mechanism for ligand-induced conformational switching which is consistent with reported requirements for SAM-I riboswitch function .
For SAM binding to fully convert an AT helix to a P1 helix may require at least milliseconds , judging from NMR measurements on an analogous strand-switching RNA [53] , or longer based on a strand displacement assay [54] . We reasoned that the most rapid effect of SAM binding on the riboswitch would take place if the ligand bound to an intermediate conformation , hybridizing elements of the ON and the OFF state . Figure 2 shows a schematic of the strategy that we used to generate a starting configuration for our MD simulation . In the ON state , one strand of the P1 helix pairs with a downstream segment of the expression domain ( removed in the crystallized RNAs ) to form the AT helix . We initiated our simulation with a truncated segment fixing a partial P1 helix ( two base pairs ) , and a partial AT helix ( seven base pairs ) . In between a 4 nucleotide competition region can form either a P1 or AT helix . We call this “hybrid” construct 6P1_11AT , since it has the potential to form up to 6 P1 base pairs and up to 11 AT base pairs . Our simulations start with three of the four switching base pairs as AT helix , and a boundary nucleotide residue ( U110 ) is positioned equally close to its putative AT or P1 binding partners . This choice of starting structure allowed us to 1 ) Work with a segment that was shown experimentally to bind SAM , 2 ) Include the minimal nucleated P1 helix known to bind SAM , and 3 ) Maximize the potential number of AT base pairs with the potential to switch to P1 pairing ( see “Details of MD simulations” in Supplementary Information ( text S1 ) ) . We used MC-Sym [40] , [56] to sample the placement of the AT helix in the 3D structures and the geometry of the boundary region with the nucleated P1 . Previously we showed experimentally that SAM binds to hybrid constructs [57] . Though reduced in affinity , the SAM binding to the hybrids has similar dependence on Mg2+ , and similar sensitivity to mutations as with the aptamer . Therefore we assumed that the folding of the portion of the SAM/hybrid riboswitch complex outside of the strand switching region , henceforth referred to as the “aptamer core” , is similar to that in the X-ray structure of the aptamer domain . Since the AT helix approximates a canonical A form geometry , the critical local region to be sampled is the three nucleotide segment A109 , U110 and A111 . These three nucleotides act as a hinge to bridge the partial P1 helix and the nearly complete AT helix . Additionally , an explicit triplet constraint was applied on the three nucleotides highlighted in purple in Figure 2A and 2B ( A4 , U110 and A136 ) . Two adenosines compete for base pairing with a U ( Figure 2B ) . The scripts used to generate the models can be found in the SI . An overview of the outcome from MC-Sym sampling is shown in Figure 2 using the pseudo-dihedral angle [58] . Monitoring of the pseudo-dihedral angle ( Figure 2C ) indicates that MC-Sym has focused on the populated geometries according to the known structures , but also has sampled exhaustively the full range of geometries ( Figure 2D ) . There is a region ( between 80 and 170 degrees ) that is rarely sampled due to steric clash with the P3 helix coordinates ( Figure 2D ) . Therefore , the results demonstrate that MC-Sym can sample a wide range of the conformational space , while placing the AT helix without steric clashes . Three criteria were used for selecting MC-Sym generated models for MD simulations: 1 ) Calculated potential energy should be favorable , 2 ) The SAM binding pocket must be accessible and 3 ) Coaxial stacking should be present between the P1 and AT helices . The latter constraint was based on experimental observations that SAM binding at µM affinity was detected for RNA constructs which allow the potential for such stacking ( “3P1_10AT” ) , but not for those which do not ( “3P1_9AT” ) [57] . For reasons explained in supplementary information , we used the Amber99bsc0 force field with the generalized Born ( GB ) implicit solvent model to calculate free energy . Two ( model 51 and model 55 ) out of the top five ranked in terms of free energy satisfied all the three criteria . Figure S1A shows calculated free energies , while Figure S1B highlights the coaxial stacking for these two models as measured by internucleotide vdW energies . The local geometry of the switching region is displayed schematically for these two structures in Figure 3A and global folds are shown in Figure 3B . The main difference between these two models is that the unpaired 5′ strand of the P1 helix is placed in the two different grooves of the AT helix–in the minor groove of the AT helix for model 51 , and in the major groove for model 55 ( Figure 3A ) . The geometry sampled in model 51 and 55 resembles an RNA triple helix composed of poly ( U ) -poly ( A ) -poly ( U ) from a crystal structure [59] . With limited experimental data , these two models are rationalized as potential models for the intermediate or “transition state” between ON and OFF state . Table 1 lists MD trajectories included in this study , using model 51 and 55 and the X-ray coordinates ( 3NPB ) [60] as starting models . Different trajectory evolutions are observed for model 51 with or without SAM . Strikingly , formation of a complete P1 helix ( all 6 Watson-Crick base pairs ) is observed at ∼1 . 3 µs for the simulation in the presence of SAM ( see Movie S1 ) . Figure 4A displays the time evolution of RMSD for individual base pairs with reference to that in the P1 helix of the X-ray structure . A small RMSD value ( deep blue ) indicates that the geometry of the nucleobases in a single base pair is close to that observed for the Watson-Crick base pair in the crystal structure . Monitors of classical Watson-Crick hydrogen bonding presence for the base pairs in the P1 helix and the AT helix are presented in Figure S2 . As is apparent from Figure 4A and Figure S2 , the time at which the P1 helix was completely formed can be located as indicated with the red arrow in Figure 4A . The lifetime of this conformation spans from frame 6544 to frame 6635 ( 18 . 2 ns ) . The top 4 base pairs in the P1 helix ( base pair 1 to 4 ) maintain the P1-like conformation corresponding to the crystal structure during the remaining simulation in the presence of SAM . Additionally , the electrostatic interactions between the sulfur atom of SAM and the carbonyl oxygen atoms of two U nucleotide residues persist through out the simulation ( Figure 4B ) as observed in the repeated simulation on the aptamer domain of the yitJ SAM-I riboswitch ( 3NPB in the presence of SAM in Table 1 ) . The short life span of the fully formed P1 helix is linked to fraying of the closing base pair ( base pair 6 ) . The two participating nucleotides flip to a cross-strand stacking conformation . The adjacent base pair ( base pair 5 ) is disrupted shortly after the loss of the closing base pair , but reappears at 1 . 9 µs for 300 ns ( altogether 2250 out of 9097 snapshots after the strand migration event display this base pair ) . The two bases remain proximal ( Figure 4C ) , however , and flip between states involving alternative hydrogen bonding patterns ( Movie S1 , Figure S3 ) . A similar plot for the AT helix pairs shows that the destabilization of AT base pairs 1 to 3 precedes complete P1 formation ( Figure 4A , Figure S2 ) . The disruption of this AT region is not due to the deficiency in modeling the AT helix since the simulation on the same model without SAM maintains the geometry close to a standard A-form helix for 2 ( base pair 2 and 3 ) out of these three base pairs . Moreover , the 2 base-pair starting partial P1 helix is unstable in the absence of SAM in model 51 ( Figure 4A , Figure S2 ) . During the interval leading up to the strand migration event , P1 helix base pairs 4–6 show a slowly rising trend in RMSD relative to the X-ray coordinates ( Figure 4A , Figure S3 ) . Thus the strand migration event is preceded by a fluctuation in which the corresponding nucleotide residues explore a “transition state” . This fluctuation coincides with the loss of AT helix base pairs 2 and 3 , which otherwise block P1 helix propagation through P1 base pairs 5 and 6 . The RMSD for P1 helix base pairs 3–6 goes down , in some cases dramatically , at the time of the strand migration . The two terminal base pairs drift towards configurations which show only a slightly smaller RMSD relative to X-ray coordinates than at the start . The RMSD of backbone atoms relative to the X-ray coordinates , however , decreases and remains low after the strand migration event ( Figure S3 ) . Complete P1 formation does not take place in model 55 within the time scale ( 1 . 467 µs ) accessible so far for this simulation when SAM is present . Base pair 3 in the P1 helix gets trapped in a state with base pair geometry close to an AU Hoogsteen base pair ( U•A cis W . C . /Hoogsteen and class XXIII according to reference [61] ) ( Figure S4 ) . In addition , this state is stabilized by a new hydrogen bond interaction between A4 and SAM , which is not sampled during the simulation of the aptamer domain ( the construct for the X-ray study ) in the presence of SAM ( Figure S4B , C ) . In Figures 5 and 6 we visualize the conformational pathways observed for the various trajectories for model 51 . The overall fraction of hydrogen bonds in Watson-Crick base pairs from the P1 helix and from the AT helix are used as generalized coordinates . Figure 5 displays the conformational trajectories for a simulation started from model 51 only , and a second simulation from model 51 in complex with SAM . The results suggest that starting model 51 locates at a branch point in the FEL . The formation of a stable AT helix ( high probability for AT Helix Hydrogen Bonding-vertical axis ) is favored in the absence of SAM , while the presence of SAM allows model 51 to navigate to other transient states and eventually leads to sampling of the conformation with a complete P1 ( high probability of P1 helix formation-horizontal axis , Figure 5 ) . However , the event of complete P1 helix formation is short-lived ( Figure 5 , Figure 6A , B ) . Therefore , a third simulation restarted from a snapshot with complete P1 at frame 6615 of the first trajectory in the presence of SAM ( the snapshot with the lowest RMSD relative to the X-ray coordinates ) was performed to evaluate the stability of this conformation ( Figure 6 C , D ) . A 1 . 767 µs trajectory starting from frame 6615 in simulation of model 51 with SAM , only samples the bottom part of a deep energy funnel populated by an ensemble with complete P1 helix ( Table 1 , Figure 6E ) . Interestingly , P1 helix base pairing is also relatively stable ( persists through the simulation ) when frame 6615 with a complete P1 helix is used as the starting coordinates for a simulation without SAM present ( Table 1 , Figure 6F ) . When snapshot 9974 , with a five base pair P1 helix , is used as starting coordinates , the 5 P1 helix base pairs again remain relatively stable over the course of the trajectory ( Table 1 , Figure 6G ) . In this case , however , opening of some individual P1 helix base pairs is observed , particularly towards the end of the trajectory with SAM absent ( Table 1 , Figure 6H ) . In the latter trajectory , at least one P1 helix base pair reverts to AT base pairing . In SAM-I aptamer X-ray structures a Mg2+ ion observed near the SAM binding site and phosphate moieties in J1/2 and J3/4 [60] , [62]–[64] . We monitored the contact distances between this Mg+2 and phosphates in J1/2 in the various trajectories of the SAM-I riboswitch aptamer and hybrid starting models with and without SAM . Our previous simulation on another aptamer of SAM-I riboswitch—metF from T . tengcongenesis [31] indicated cooperativity between this Mg2+-phosphate coordination complex and SAM-leading to stabilization of tertiary interactions . Similarly , this effect was also observed in simulations of model 51 and 55 in the presence of SAM ( Figure 7 ) . The presence of SAM is correlated with the maintenance of short magnesium contacts with J1/2 , while these contact distances increase during the simulations without SAM . Contact distances between Mg2+ and phosphates in J3/4 are almost constant in the presence and absence of SAM . For the yitJ aptamer the correlation between the presence of SAM and short Mg2+ contact distances with J1/2 is still maintained ( Figure S5 , but contact distances with phosphates on J3/4 begin to increase in the absence of SAM . Additionally , for restarted simulations of frame 6615 and 9974 , the contacts of this Mg2+ ion with J1/2 are still maintained even in the absence of SAM ( Figure S6 ) . Overall , these results confirm the stable coordination between the Mg2+ ion and J3/4 in the absence of SAM , and the tendency of SAM contact to stabilize an additional coordination with J1/2 . Movies S2 , S3 , S4 , S5 , S6 , S7 also highlight base moieties attached to nucleotide A7/9 in J1/2 and A80/82 in J3/4 . Our earlier study also observed transient formation of a non-adjacent dinucleotide stack between nucleotide bases in J1/2 and J3/4 in simulations with and without SAM [31] . Of the ∼12 X-ray SAM-I riboswitch coordinate sets [60] , [62]–[64] all except one ( pdb id 3GX3 , with SAH bound ) show the two nucleotide bases pointing to the same region outside the helix , with the respective bases within 3–7 angstroms proximity . In this study we again observed transient formation of dinucleotide stacking with and without SAM for model 51 and the aptamer , but with alternating stacking geometries ( Movies S2 , S3 , S4 , S5 , S6 , S7 ) . Predominantly the two nucleotides were positioned with favorable stacking energies ( Figure S7 ) but little effect was observed from SAM binding .
Overall , we can summarize the results with model 51 MD trajectories as the following: 1 ) In the absence of SAM , 2–3 starting P1 helix base pairs appear to be unstable , whereas a long AT helix remains stable up to the terminal base pair; 2 ) In the presence of SAM , a strand migration event is observed after ∼1 . 3 µs leading to transient formation of a full 6 base pair P1 helix , at the expense of competing AT helix base pairs; 3 ) Terminal base pairs within the fully formed P1 helix form transiently in the original simulation , but appear relatively stable in a new trajectory using the snapshot with fully-formed P1 helix as the starting point; 4 ) The fully formed P1 helix is also relatively stable in a trajectory which starts with the same snapshot even in the absence of SAM . 5 ) In a trajectory starting with 5 P1 base pairs with SAM , hydrogen bond contacts corresponding to the five base pairs appear slightly more stable than they do in one starting from the same snapshot without SAM . By contrast , trajectories starting with model 55 result in a conformation in which the competing base pair at the boundary between P1 and AT base pairs forms a non-Watson-Crick pair , while other P1 and AT base pairs are stable . Taken together , these findings indicate that SAM binding promotes P1 helix base pairing at the expense of AT helix pairing , but with qualifications . Certain starting geometries , such as that in which the 5′ nucleotides reside near the major groove of the AT helix ( as in model 55 ) , may be slow to convert to the P1 helix-forming conformation . Our original simulation of model 51 in the presence of SAM resulted in 4 stable P1 helix base pairs . Thus , SAM binding may have its strongest direct stabilization of P1 helix base pairs near the SAM binding site . All of these results are consistent with experimental evidence . A minimum length of P1 helix is necessary for SAM binding [13] , [65] , though the presence of a partial AT helix can restore µM SAM binding with a P1 helix as short as 2 base pairs [57] . The latter study indicated that SAM binding affinity increases in model systems as the P1 helix is extended and the AT helix shortened . There are also indications that P1 helix dynamics are reduced by SAM binding [17] , [54] , [65] for truncated aptamers . Earlier we proposed that SAM contacts with J1/2 and indirect stabilization of Mg2+ contacts with J1/2 enhance P1 helix formation [31] , and that the contacts with J1/2 block formation of competing conformers [66] . Our simulation suggests that additional enhancement of P1 helix formation arises through direct contact with SAM . The importance of these electrostatic SAM-P1 helix contacts for mediating the ligand binding specificity has been established experimentally [63] . As observed in our earlier simulations [31] , direct contacts between SAM and the key G11 nucleotide within the P1 helix are persistent throughout these extended timescale simulations . In addition , we observed shorter contact distances between a bound Mg2+ and at least two electronegative functional groups on J1/2 in the presence of SAM during the simulations starting with model 51 and with the aptamer , than in the absence of SAM ( Figures S4 , S5 , S6 ) . The Mg2+ ion site which we have monitored here , observed in the original X-ray structures , is suspected to form an inner sphere coordination complex [31] , [67] . Movies shown in supplementary materials ( Movies S2 , S3 , S4 , S5 , S6 , S7 ) vividly illustrate the interplay between SAM , Mg2+ , and the backbones of the J1/2 and J3/4 junctions . Movies without SAM show the Mg2+ surrounded by phosphates from J3/4 , with particularly stable coordination with phosphates 81 and 83 ( 83 and 85 in the aptamer ) . In the presence of SAM , G9/11 O6 is anchored in a bridging position between the Mg2+ and phosphate groups in J1/2 . A recent study identified a cooperative effect between Mg2+ and SAM in SAM-I riboswitch folding , and proposed a role for the same core Mg2+ in pre-organizing folding intermediates for SAM binding [55] . Movies S2 , S3 , S4 , S5 , S6 , S7 provide a striking illustration of a potential mechanism to explain this cooperativity . This coordination complex could induce a reorientation of the P1 helix , as reported [65] , by fixing the position of J1/2 . Favorable non-adjacent dinucleotide stacking between nucleotide bases in J1/2 and J3/4 is observed in model 51 with SAM , but with altered geometry in the absence of SAM ( Movies S2 , S3 ) . Altogether , these observations leave an open question as to the role that non-adjacent dinucleotide stacking may play in pre-positioning J1/2 and J3/4 in a manner that is favorable to aptamer formation and P1 helix formation specifically . The simulation of model 51 in this study shows that the stabilization of the partial P1 helix by SAM anchors the 5′ strand of the P1 helix in an orientation that enables this single strand region to compete over the AT helix . This model is reminiscent of an NMR study on a small RNA system showing that the stabilization of a pre-formed helical region by a tetra loop increases the rate of conversion between two different hairpin folds [53] . In the riboswitch , SAM stabilization of the nucleated P1 helix may play a similar role . Simulations on the same starting coordinates in the absence of SAM displayed the loss of all P1 helix base pairing . Conformations with three or fewer base pairs in the P1 helix may not be stable enough to prevent the formation of the AT helix in the absence of SAM . When a snapshot with fully formed P1 helix ( frame 6615 ) was used as the starting structure an MD trajectory displayed a relatively stable P1 helix even in the absence of SAM . When frame 9974 with 5 P1 helix base pairs was used as the starting structure , all 5 P1 helix pairs remained stable in the presence of SAM . With these starting coordinates , however , as the simulation time approached 1 µs , the P1 helix began to show some instability in the absence of SAM . Therefore , differing degrees of shift of the conformational equilibrium amongst a series of conformational intermediates toward the OFF state with SAM facilitate the SAM-I riboswitch function as a dimmer switch . The most dramatic SAM binding effect is on a hybrid conformer with minimal P1 helix base pairing . In the biological context , it is proposed that SAM binding takes place soon after the transcription of the aptamer-forming segment [68] . The conformation is then locked before full transcription of the antiterminator , a mechanism similar to that indicated for other transcriptional riboswitches [69] , [70] . Such a mechanism is highly sensitive to the concentrations of reaction components-a recent report indicated that nucleotide levels dramatically alter the degree of kinetic control of a lysine riboswitch within active transcription complexes [71] . For the yitJ SAM-I riboswitch , moreover , partial AT helix formation can take place in the non-overlapping region , even in the presence of a full P1 helix . Our simulations indicate that the presence of SAM would prevent strand invasion by this partial AT and dissociation of the P1 helix in this scenario . High resolution structures of riboswitch aptamers with and without ligand have led to the proposal that many fold according to the “conformational capture” mechanism [72] , [73] . Typically this mechanism is described as selection by the ligand of a single bound conformer amongst a range of conformations being sampled by the unliganded substrate ( Figure 8A ) . Inclusion of a portion of the expression domain , however , leads to more dramatic effects of ligand on RNA folding . Secondary structure calculations predict that an equilibrium Boltzmann ensemble for the yitJ SAM-I riboswitch includes some hybrid conformations with partial P1 and partial AT helix [57] , [66] . A “capture” of these intermediates , according to simulations here , would facilitate rapid propagation of a longer P1 helix . This event would free a sufficient segment of the 3′ strand of the AT to nucleate the downstream Terminator sequence . By contrast , our simulations indicate that in the absence of SAM the AT helix could displace the nucleated P1 helix within the intermediates . A more precise description of conformational capture in this scenario would be selection of a region of conformational space by the ligand , which then chaperones the RNA towards the aptamer configuration ( Figure 8B ) . In panel B of the figure , free energy is now the relative free energy of the total system , including ligand as well as RNA plus solvent and ions . Conformations that can bind SAM have reduced free energy relative to those for which RNA and SAM are not in contact , and the reduction in free energy for each conformer is proportional to favorable free energy of binding . We hypothesize that the aptamer folding rate would be accelerated by this mechanism because during the Levinthal sampling process the FEL region that can initiate aptamer formation is widened . Simulations of RNA folding kinetics [74] concluded that a strand migration pathway would lead to the fastest transition rate for an inter-conversion between two hairpins . Kinetic folding studies for a number of riboswitches [75]–[79] indicate that P1 helix formation takes place during later stages of the folding pathway . Our simulations and the experimental findings in our previous study [57] therefore raise the possibility of a role for SAM in accelerating P1 helix formation , by facilitating strand migration as the aptamer folding pathway . In this scenario , SAM binding could still facilitate aptamer formation after a portion of the expression domain has been transcribed . In vitro kinetics of SAM-I riboswitch folding and transcription termination would then be highly sensitive to mutations in the expression domain , as has been reported for a transcriptional lysine riboswitch [71] . SAM effects on P1 vs . AT length could be tested through NMR measurements on partially labeled SAM-I riboswitch hybrid constructs , or through NMR methods designed to detect minor conformers [80] , [81] . In previous work we showed that altered base pairing in the unliganded yitJ SAM-I riboswitch as compared to the bound state extends beyond the P1/AT helix switch [57] , [66] . The question of the impact of the ligand on the riboswitch conformation is therefore related to large-scale alterations in RNA folding . Although great advancement has been achieved to speedup MD simulations , a complete simulation of folding/unfolding for RNA of this size ( ∼40 kDa ) is still not possible . The P1/AT helix switching event alone may take ms or longer , according to data on analogous model systems [53] . This is because the nucleation of the transient state takes up most of the folding time . The propagation step can be faster than the overall folding rate by four orders of magnitude [82] . Therefore , MC-Sym was used to sample a discrete conformational space aiming to identify candidate transient state models with atomic details to bypass the most time-consuming part of the simulation . In the SI , we discuss measurements from the literature for folding and conformational switching for a range of RNAs . Overall , considering the literature data and estimating a rate constant based upon snapshots observed in the model 51 trajectory as transition states ( Tables S1 and S2 ) , the microsecond regime appears plausible for the strand migration within the three base pair stretch simulated in this study . Extension of the MD timescale to the microsecond regime by using Anton now appears to make some intermediate steps of strand migration accessible . The adequacy of force fields and parameters for long timescale simulations for RNA is relatively untested as compared to protein MD [83] . Nonetheless , it seems improbable that a strand exchange observed only in the presence of ligand , and leading to decreased RMSD relative to X-ray coordinates ( Figure 4 , Figure S3 ) , is solely a result of instabilities or imperfections of the force fields [84] . The two terminal base pairs of the P1 helix are transient in the original model 51 simulation with SAM , although the RMSD relative to the X-ray structure for each base pair remains lower than before the strand invasion . This observation may reflect instabilities in the force field , a genuine tendency towards “fraying” [85] , [86] , or a longer simulation may be required to reach a thermodynamically stable state . The success of this study in observing a strand migration event should motivate efforts to optimize and validate parameters and protocols for long timescale MD simulations for RNA .
The atomic models for the RNA construct described in Figures 2 and 3 were generated using MC-Sym [40] installed locally . The aptamer core ( highlighted in blue in the figure ) was modeled using its counterpart in the known structure of the yitJ SAM-I riboswitch ( PDB ID: 3NPB ) [60] . The other parts of the construct were built from the library of small fragment RNA structures , known as Nucleotide Cyclic Motifs ( NCMs ) [56] . An explicit triplet constraint was applied on the three nucleotides highlighted in purple ( A4 , U110 and A136 ) . In this way we sampled the 3D space in which these three nucleotides are proximal to each other . In these three nucleotides , the two As are competing for base pairing with a U . The scripts used to generate the models can be found in Appendix S2 . Different RMSD threshold values were tested to ensure exhaustive sampling in the local region bridging the partial P1 and the AT helix ( A109 , U110 and A111 ) . Models with small differences ( low pairwise RMSD ) in pseudo-dihedral angle of the A109-A111 region were filtered out . Energy minimizations ( max step is 2000 or gradient tolerance <1 . 0 ) were performed on the atomic structures of the models generated from MC-Sym runs using Nucleic Acid Builder ( NAB ) [87] . AMBER99bsc0 force field [88] and Generalized Born model [89] with an inverse Debye-Huckel length of 0 . 19 Å−1 [90] were used in the energy minimization procedure . 149 models were generated in this step . This energy minimization is mainly to rebuild the chain connectivity for models generated from MC-Sym without introducing the sampling effect of the force field . Thus , we used MC-Sym to sample the possible placement of the AT helix in the 3D structures and the geometry of the potential nucleation site of the P1 helix close to the SAM binding pocket . The modeling assumed that the folding of the aptamer core is similar to that in the crystal structure of the aptamer domain . After the energy minimization step , models with high van der Waals energy were filtered out . There are two reasons for high van der Waals energy: 1 ) steric clashes that cannot be released by energy minimization , 2 ) broken chain connectivity that cannot be bridged during energy minimization . Models were chosen following the three criteria listed in the results section under “Selection of starting models for MD simulation” . For the models in the presence of SAM , the ligand was placed in the binding pocket while maintaining most of the interactions ( except the contacts with the end base pair AU in the partial P1 helix ) observed in the crystal structure of the aptamer domain complex ( PDB: 3NPB ) . The simulations are run on Anton [36] . The equilibrated structures for Anton were prepared using local HPC clusters following the MD protocol as described in our previous study [31] ( also see “Details of MD simulations” in Text S1 ) . The trajectory was recorded for every 200 ps . The definition of hydrogen bond probability ( HBP ) of the hydrogen bond i at time t is similar to that in reference [91]: ( 1 ) where , and is defined as ( 2 ) Here is the distance between hydrogen and hydrogen bond acceptor , is the angle of hydrogen bond donor , hydrogen and hydrogen bond acceptor and scaling constant Å . In the reference state Å and rad . The list of hydrogen bonds monitored is listed in Table S3 .
|
Folding dynamics is crucial for RNA function . Riboswitches are a classic example . A typical riboswitch senses the cellular concentration of a small molecule . By refolding itself into a new structure , the riboswitch converts that information into changes in rates for synthesis of related metabolites . Understanding how the small molecule physically changes RNA structure can help us to target riboswitches , which occur mainly in bacteria , for drug design , or to engineer new riboswitches . This understanding has been blocked because 1 ) we cannot view intermediate stages experimentally and 2 ) simulations cannot reach the timescale for the structural conversion . Recent advances in RNA structure modeling enable us to model intermediate states . A new computer specialized for long timescale molecular dynamics ( MD ) simulations , called Anton , helps us to extend the simulation timescale . We modeled intermediate riboswitch structures , focusing on a reduced segment of the structure-switching region , in order to reduce the time required for a transition . We simulated an MD trajectory in which a small molecule converted the structure of this reduced switching region . Some steps in riboswitch structural transitions are therefore accessible to the newly extended MD timescale . Wider availability of resources like Anton can aid the advancement of riboswitch engineering and novel antibiotic design .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biochemistry",
"rna",
"structure",
"prediction",
"rna",
"structure",
"genomics",
"computational",
"chemistry",
"molecular",
"dynamics",
"nucleic",
"acids",
"protein",
"folding",
"gene",
"expression",
"molecular",
"genetics",
"rna",
"synthesis",
"biology",
"computational",
"biology",
"chemistry",
"biophysics",
"simulations",
"biophysics"
] |
2013
|
The Impact of a Ligand Binding on Strand Migration in the SAM-I Riboswitch
|
Microbial pathogens have developed efficient strategies to compromise host immune responses . Cryptococcus neoformans is a facultative intracellular pathogen , recognised as the most common cause of systemic fungal infections leading to severe meningoencephalitis , mainly in immunocompromised patients . This yeast is characterized by a polysaccharide capsule , which inhibits its phagocytosis . Whereas phagocytosis escape and macrophage intracellular survival have been intensively studied , extracellular survival of this yeast and restraint of host innate immune response are still poorly understood . In this study , we have investigated whether C . neoformans affected macrophage cell viability and whether NF-κB ( nuclear factor-κB ) , a key regulator of cell growth , apoptosis and inflammation , was involved . Using wild-type ( WT ) as well as mutant strains of C . neoformans for the pathogen side , and WT and mutant cell lines with altered NF-κB activity or signalling as well as primary macrophages for the host side , we show that C . neoformans manipulated NF-κB-mediated signalling in a unique way to regulate macrophage cell fate and viability . On the one hand , serotype A strains reduced macrophage proliferation in a capsule-independent fashion . This growth decrease , which required a critical dosage of NF-κB activity , was caused by cell cycle disruption and aneuploidy , relying on fungal-induced modification of expression of several cell cycle checkpoint regulators in S and G2/M phases . On the other hand , C . neoformans infection induced macrophage apoptosis in a capsule-dependent manner with a differential requirement of the classical and alternative NF-κB signalling pathways , the latter one being essential . Together , these findings shed new light on fungal strategies to subvert host response through uncoupling of NF-κB activity in pathogen-controlled apoptosis and impairment of cell cycle progression . They also provide the first demonstration of induction of aneuploidy by a fungal pathogen , which may have wider implications for human health as aneuploidy is proposed to promote tumourigenesis .
Cryptococcus neoformans is a facultative intracellular pathogen that is the most common cause of systemic fungal infections leading to meningoencephalitis in immunocompromised patients , and notably in people infected with HIV [1] , [2] . This saprophytic basidiomycete fungus is characterized by the presence of a polysaccharide capsule , composed of glucuronoxylomannan ( GXM ) , galactoxylomannan ( GalXM ) and mannoproteins . The capsule constitutes the main virulence factor of C . neoformans and inhibits its phagocytosis [3]–[5] . Infection by C . neoformans is thought to result from its inhalation as basidiospores and usually leads to asymptomatic pneumonia , followed by a latent phase that can last many years [6] . When immunodepression arises , reactivated yeasts disseminate into the bloodstream , reach the central nervous system and cause fatal meningoencephalitis if left untreated . In the pathogenesis of cryptococcosis , macrophages play a major defence role [7]–[11] . To evade the host immune system and macrophage-mediated killing in particular , C . neoformans has developed several stratagems . Among those , its phagocytosis by innate immune cells is inhibited through both capsule-dependent ( for review see [12] ) and capsule-independent mechanisms [13] , [14] . Once phagocytosed , C . neoformans has the ability to exit the macrophage through a mechanism that does not kill the host cell thereby avoiding inflammation [15] . In addition , C . neoformans can survive inside the phagolysosome and macrophage serves as a site for both fungal replication and reservoir during latency [16] , [17] . Remarkably , macrophages do not spontaneously phagocytose C . neoformans , as they do for other microbes and human pathogenic fungi . They require its opsonisation , a process involving complement or antibodies , production of which takes several days in the mouse [18] . Without opsonisation , and therefore during the early phase of infection prior to antibody generation and in tissues with low levels of complement , very little production of inflammatory cytokines is triggered by C . neoformans [19] , [20] . Programmed cell death ( PCD ) of immune cells may constitute another way to limit inflammation , as apoptotic inflammatory cells fail to release their proinflammatory and histotoxic substances . Indeed , purified capsule components , GXM and GalXM polysaccharides , have been shown to induce apoptosis of T cells [21] , [22] as well as macrophages [23] , [24] . However , the precise mechanisms involved in macrophage PCD are not fully elucidated and little is known about other strategies elicited by unopsonised C . neoformans to promote immune evasion , especially about a direct effect of this pathogen on viability and growth of the macrophage itself . NF-κB is a ubiquitous transcription factor with post-translationally regulated activity , which plays a pivotal role in inflammation , immunity , cell growth and apoptosis through the control of expression of major immunomodulatory , cell proliferation and death regulatory genes [25]–[27] . Five mammalian NF-κB subunits that can form homo- or hetero-dimeric combinations have been identified: p50 , p52 , p65/RelA , c-Rel and RelB . In most cells , NF-κB dimers are sequestered in the cytoplasm by interaction with inhibitory proteins , the IκBs , namely IκBα , IκBβ , IκBε , p105 and p100 . The nuclear translocation of NF-κB is regulated by two prevailing activation pathways [28] . The classical/canonical one , induced by engagement of cytokines as well as death and pattern recognition receptors , depends on the IKK complex , which is composed of two catalytic subunits IKKα/1 and IKKβ/2 and a regulatory subunit NEMO/IKKγ . Upon stimulation , the IKK complex triggers phosphorylation of two Ser residues within the N-terminus of the IκBs , leading to their ubiquitination and degradation by the proteasome thereby releasing NF-κB dimers to the nucleus . In macrophages , activation of classical NF-κB dimers is predominantly controlled by IKK2 [29] , [30] . The alternative pathway , induced by certain members of the TNF family such as lymphotoxin-β or CD40L [31] , [32] , is NEMO- and IKK2- independent , and specifically involves phosphorylation of IKK1 by NF-κB-Inducing Kinase ( NIK ) . This in turn triggers phosphorylation of p100 , the main RelB inhibitor , and its processing to p52 thereby freeing RelB/p52 heterodimers . These two pathways are complexly interconnected and activation of the alternative pathway may be regulated by the induction of p50/p65 , or Rel complexes [33] . In this study , we have explored whether C . neoformans like other microbes , such as bacteria or viruses , uses inhibition of cell viability to restrain host immune response and asked how NF-κB might control fungal-regulation of macrophage survival . We report that unopsonised C . neoformans elicited apoptosis of macrophages as well as repressed cell proliferation through disruption of macrophage cell cycle via the modification of expression of an array of cell cycle checkpoint regulators . We further show that the classical and alternative pathways of NF-κB activation were differently required for these processes . Collectively , these findings shed light on a new fungal strategy to evade host response through uncoupling of NF-κB activity in pathogen-induced apoptosis and cell cycle progression impairment . They also disclose for the first time fungal-induced aneuploidy .
To explore the effects of unopsonised C . neoformans on macrophage viability , we first verified that in our conditions of infection , phagocytosis of C . neoformans by J774 macrophage-like cells was negligible . Two serotype A strains were studied in parallel , the wild-type ( WT ) ( KN99α ) and an acapsular mutant ( cap59D , devoid of GXM the major capsule component [34] ) , expected to be more phagocytosed [35] . Indeed , a maximum of 4% of phagocytosis was observed for the WT strain with 1–3 yeasts ingested per cell after 48 h of infection . For the acapsular strain , phagocytosis although higher , did not exceed 15 . 7% , with in that case 1–20 yeasts internalized per cell ( Figure S1 ) . This indicated that in our experimental settings , unopsonised WT and even acapsular C . neoformans remained mostly extracellular . We next examined the effect of unopsonised C . neoformans infection on macrophage cell viability . We found a significant decrease in viability of J774 macrophage-like cells , measured by ATPmetry with unopsonised KN99α from 48 h post-infection ( p . i . ) onward ( Figure 1A ) . This decline of viability was dose-dependent ( Figure 1B ) but independent of the capsule as assessed by a similar effect of WT ( KN99α ) and capsule mutant strains of C . neoformans ( cap59D , lacking GXM; cas1D , missing O-acetylation of the capsule [36]; suppressors of uge1D , devoid of GalXM with a doubling time similar to that of WT at 37°C [36] ) ( Figure 1C and Figure S2 ) . However , a serotype D strain of C . neoformans led to a mild effect even at high M . O . I . ( 20 ) with an overall profile of viability parallel to that of mock-treated cells , and therefore completely different from that of serotype A-infected cells ( Figure 1D ) . This fungal-induced inhibition of viability required both viable yeasts and pathogen-cell contact as revealed by the analogous behaviour of heat-inactivated C . neoformans and mock-treated J774 cells , as well as the absence of effect of J774 cell infection by KN99α in transwells ( Figure S3 ) . Remarkably , fungal-induced reduction of viability also occurred in primary macrophagic cells such as bone marrow derived macrophages ( BMDM ) ( Figure 1E ) . Since cell viability reflects the balance between cell proliferation and death , we then investigated whether apoptosis was elicited by C . neoformans in macrophages . TUNEL+ cells were detected 48 h p . i . with the WT strain of C . neoformans in 11 . 6% of J774 macrophage-like cells and in 20 . 7% of BALB/c BMDM and persisted 72 h p . i . ( Figure 1F , G ) . Capsule mutant strains of C . neoformans ( cap59D lacking GXM , or uge1D devoid of GalXM ) did not trigger any apoptosis , indicating an essential function of these capsular components ( Figure 1H ) . Then , we assessed whether macrophage cell growth was inhibited by fungal infection . Indeed , significantly reduced numbers of mitotic J774 macrophage-like cells were detected by immunocytochemical analysis using two classical proliferative markers Ki-67 and phospho-histone H3 ( Ser10 ) from 48 h p . i . onward ( Figure 2A , B ) . Such fungal-induced diminution of proliferation also operated in BMDM ( Figure 2C ) and more generally in the spleen in vivo 3 d p . i . of mice challenged with C . neoformans in a model that mimics the systemic infection in humans [37] , as revealed by phospho-histone H3 immunohistochemistry ( Figure 2D ) . We then asked whether this decrease of proliferation might originate from a disruption of the host cell cycle by C . neoformans . Representative graphs of cell cycle distribution determined by propidium iodide staining and flow cytometry showed a clear shift of the peak corresponding to J774 cells ( Figure 3A , Figure S4A ) or BMDM ( Figure S4B ) in G0/G1 48 or 72 h p . i . . This shift , ascertained by invariance of a chick erythrocyte control standard peak ( Figure S5 ) , indicated an augmentation of DNA content , which varied in the range of 1 . 5 to 2 diploid DNA content . In addition , quantification of cells in the various phases of the cell cycle disclosed a reduction in proportion of cells both in S and G2/M upon infection ( Figure 3B ) , in accordance with the immunocytochemistry data shown above ( Figure 2A , B ) . Altogether , these results revealed that C . neoformans infection impaired macrophage cell cycle together with a modification of its DNA content . To further confirm an effect of this fungal pathogen on cell ploidy , metaphase preparations from untreated ( mock ) or 48 h-infected J774 cells or BMDM were analyzed ( Figure 3C , D ) . Overall , we noticed a strong decrease ( about 3- to 10-fold less ) in the total number of metaphases for both J774 or BMDM infected cells compared to mock-treated ones , consistently with the cell cycle defects unveiled above . When focusing on metaphasic chromosomes , we observed that infection by C . neoformans induced a variation in chromosome number in both J774 cells and BMDM . We next asked whether these effects on cell cycle and ploidy were also triggered by an acapsular strain of C . neoformans . Representative graphs of cell cycle distribution determined by propidium iodide staining and flow cytometry disclosed , as for WT C . neoformans , a clear shift of the peak corresponding to J774 cells in G0/G1 infected for 48 h by the acapsular strain ( cap59D ) ( Figure 4A ) . In addition , quantification of cells in the various phases of the cell cycle revealed a similar decrease of the number of infected cells in S and G2/M phases by both strains ( Figure 4B ) . Consistently , metaphase preparations showed a modification of chromosome numbers in J774 cells infected by the acapsular strain ( cap59D ) ( Figure 4C ) . Thus in macrophages , numerical changes in whole chromosomes ( aneuploidy ) were induced by a fungal pathogen . Given the seminal role of NF-κB in growth control and survival , we next investigated whether C . neoformans induced NF-κB activity . EMSA analysis disclosed that both WT ( KN99α ) and capsule mutant strains of C . neoformans led to a strong increase in NF-κB binding activity , represented by two major complexes , I and II , in J774 murine macrophage-like cells ( Figure 5A , B ) . Supershift experiments identified complex II as p50 homodimers . Pathogen-induced complex I consisted of p50-containing heterodimers including p50/p65 as well as specific p52-containing dimers ( Figure 5B ) . NIK is a serine-threonine kinase that is critical for the induction of the IKK1-dependent processing of p100 . Western blot analysis of total J774 protein extracts revealed that NIK expression levels were induced by WT C . neoformans ( KN99a ) , especially at 24 and 48 h p . i . , but only when proteasome was inhibited by MG132 treatment ( Figure 5C ) . This indicated that the steady-state levels of NIK were elevated upon infection of J774 cells , and suggested that , although NIK might be abundantly produced in J774 cells by C . neoformans infection , it was rapidly degraded by the proteasome . The levels of NIK expression correlated well with those of phosphorylated p100 ( Figure 5C ) . C . neoformans also enhanced p100 levels after 24 and 48 h of infection and , in accordance with the EMSA and NIK/Phospho p100 data , induced processing of the p100 to p52 , resulting in increased p52 levels from 24 h p . i . onward ( Figure 5D ) . Thus , both the alternative and the classical pathways of NF-κB activation were induced by C . neoformans in macrophages . We then asked whether this induction of NF-κB DNA binding activity was associated with an increase in NF-κB-dependent gene expression . In a NF-κB-reporter assay , C . neoformans infection of transfected J774 cells elicited after 4 h a small rise in NF-κB activation , which was amplified after 24 h ( Figure 5E ) . When challenging in vivo κB-lacZ reporter transgenic mice by C . neoformans in a model that mimicked systemic infection in humans , increase in β-galactosidase+ cells , including macrophages , was observed 3 d p . i . in the spleen ( Figure 5F–I ) and correlated with high fungal burden ( 2 . 7×103 CFU±1 . 0 ) . Taken together these results demonstrated C . neoformans-induced NF-κB-transactivation both in vitro and in vivo . To decipher the role of NF-κB in these processes , stable J774 clones with constitutively altered NF-κB activity or IKK1- or IKK2-dependent signalling ( Figure S6 ) were generated by overexpression of either the super-repressor/IκBα-AA ( SR ) or a kinase-dead mutant of IKK2 ( IKK2 DN ) or a kinase-active mutant of IKK2 ( IKK2 DA ) or a kinase-dead mutant of IKK1 ( IKK1 DN ) . To specifically evaluate the contribution of the alternative pathway of NF-κB activation , nfκb2−/− BMDM [38] , which lack both p100 and p52 , and their C57BL/6 wild-type controls , were also used . When classical NF-κB activity was inhibited by overexpression of the super-repressor ( SR ) , apoptosis occurred at 72 h p . i . later than in control WT J774 cells ( Figure 6A , B ) , as in the case of overexpression of a constitutive negative mutant of IKK2 ( IKK2 DN ) ( data not shown ) . In contrast , overexpression of a constitutive active mutant of IKK2 ( IKK2 DA ) , which led to constitutively enhanced activation of NF-κB ( Figure S6C ) , reversed the trend and resulted in advanced apoptosis at 24 h p . i . . Remarkably , no apoptosis was observed in the absence of the alternative pathway of NF-κB activation ( nfκb2−/− ) ( Figure 6B ) , nor when IKK1 DN stable J774 clones were used . This similar behaviour of nfκb2−/− BMDM and IKK1 DN stable J774 clones indicated a specific blockade of the alternative pathway in these stable clones , as inferred from their biochemical characterization showing constant p52 levels during fungal infection ( Figure S6C , D ) . These results suggested that the alternative activation pathway of NF-κB was essential for pathogen-induced PCD , whereas the classical pathway controlled its onset . To unravel the molecular mechanisms controlled by NF-κB , which could account for fungal-induced cell apoptosis , we screened by Western blotting total protein extracts from WT or mutant J774 cells for activation of various apoptosis effectors ( Figure S7 ) . In WT J774 cells , selective proteolytic cleavage of apoptosis initiator caspases ( caspase-8 and -9 ) , executioner caspase ( caspase-3 ) and poly ( ADP-Ribose ) polymerase ( PARP ) , both a substrate of caspase-3 and a caspase-independent apoptosis effector , was observed upon C . neoformans infection . FASL and TRAIL-R1/DR4 protein levels were also increased in WT J774 from 24 h p . i . onward ( Figure S8 ) . Inhibition of classical NF-κB dimers delayed production of the above-mentioned activated caspases and PARP upon infection , whereas overexpression of a constitutive active mutant of IKK2 ( IKK2 DA ) led to earlier and stronger fungal-induced proteolytic processings . ( Figure S7 ) . When the IKK1-dependent pathway ( IKK1 DN ) was constitutively repressed , fungal infection triggered less active forms of caspases to levels that were below the threshold of effective DNA fragmentation ( Figure S7 and Figure 6B ) . Altogether , these results indicated that fungal infection induced PCD in macrophage-like cells in an NF-κB-dependent manner through both the extrinsic ( ligand-receptor linked ) and intrinsic ( mitochondrion-mediated ) apoptosis activation pathways . We then investigated whether NF-κB regulated similarly fungal-induced inhibition of cell viability and alteration of cell cycle . In the absence of the alternative pathway of NF-κB activation ( nfκb2−/− ) , a significant decrease in viability of nfκb2−/− BMDM measured by ATPmetry was observed upon fungal infection , as in control C57BL/6 macrophages ( Figure 7A ) . In these primary C57BL/6 macrophages , a very low contribution of apoptosis to cell viability ( as fungal-induced apoptosis in WT BMDM reached 7% maximum of total cells ( Figure 6B ) ) explained the similar cell viability curves of WT and mutant BMDM . Altogether , these data indicated that the alternative pathway was not required for fungal-induced cell viability inhibition . Remarkably , when the level of classical NF-κB activity was modified in whatever way , infected cells displayed no significant difference in viability compared to mock-treated cells in contrast to what was observed in infected WT J774 cells ( Figure 7A ) . Consistently , no effect on cell cycle was seen either upon infection in any of these stable J774 clones with constitutively altered classical NF-κB activity ( Figure 7B ) . When the alternative pathway of NF-κB activation ( nfκb2−/− ) was abrogated , no modification of the number of mitotic cells compared to WT was detected at any time by phospho-histone H3 ( Ser10 ) immunostaining ( data not shown ) , confirming unaltered cell proliferation by C . neoformans in these mutant BMDMs devoid of p100 and p52 . Therefore , only the classical pathway of NF-κB activation was indispensable for fungal-triggered inhibition of cell growth and alteration of cell cycle . To decipher the molecular mechanisms by which C . neoformans repressed cell cycle progression , we purified by cell sorting mock treated ( M ) or 48 h-infected ( I ) J774 macrophages at various phases of the cell cycle and analyzed expression of an array of mitotic regulators by SDS PAGE and Western blotting ( Figure 7C ) . Fungal infection led to enrichment of p27KIP1 , a universal cyclin-dependent kinase inhibitor in G0/G1 and S phases . Conversely , cyclin-dependent kinases - such as Cdk2 ( which drives transition from G1 to S phase by interacting with cyclin-A and -E ) and Cdk1 ( responsible together with cyclin-B1 for cell cycle progression from G2 to mitosis ) as well as S-phase kinase-associated protein 2 ( Skp2 ) - were down-regulated in both S and G2/M phases . Similarly , cyclin-D1 and -E levels diminished in S and G2/M , and cyclin-A and -B1 levels in G2/M phase . Thus , concerted regulation of cell cycle effectors was orchestrated by C . neoformans infection and promoted cell cycle impairment at the S and G2/M phases . As DNA damage may cause chromosomal instability , we then asked whether the number of γ-H2AX foci , markers of DNA double-strand breaks , were modified by fungal infection . The absence of significant increase in γ-H2AX foci upon C . neoformans infection ( Figure S9 ) and lack of detection of p53 or phosphorylated CHK1 in J774 macrophage-like cells ( data not shown ) argued against fungal activation of the DNA damage signalling cascade . To search for the molecular mechanisms responsible for fungal-triggered aneuploidy , we next looked for putative modification levels upon C . neoformans infection of proteins that have been demonstrated to drive chromosome missegregation and instability when overexpressed or down-regulated , such as Mad2 a protein essential for spindle assembly during mitosis [39] , [40] . Indeed , fungal infection elicited a strong decrease in total Mad2 levels from 48 h p . i . onward ( Figure 8A ) and more specifically in the S and G2/M phases of cell cycle ( Figure 8B ) . This suggested that fungal-induced aneuploidy was mediated at least partly by Mad2 down-regulation levels . The observation upon infection of a normal cell cycle and DNA content in J774 stable clones with impaired NF-κB activity or signalling ( Figure 6B ) suggested that these processes were regulated by NF-κB . Consistently , reduced levels of cyclin-D1 upon C . neoformans infection were observed in WT J774 cells only but not in these stable J774 clones ( data not shown ) . We then asked whether modifications of Mad2 levels upon fungal infection depended on NF-κB activation . Western blot of total protein extracts from stable J774 clones with impaired NF-κB activity or IKK2-dependent signalling revealed that fungal-induced changes in Mad2 levels disappeared when classical NF-κB activity or signalling was inhibited or increased ( Figure 8C ) . Therefore , fungal-induced aneuploidy was most likely the consequence of NF-κB-controlled Mad2 down-regulation . As for WT ( KN99α ) C . neoformans , the acapsular strain ( cap59D ) , which also induced cell cycle alteration and aneuploidy ( Figure 4B , C ) , led upon infection to decreased levels of MAD2 , cyclin-D1 and Skp2 ( Figure 8D ) , in contrast to serotype D strain ( JEC21 ) , which had no impact on the levels of these proteins as expected . To determine whether fungal-induced effects on cell cycle and aneuploidy applied to tissue macrophages in vivo , we next isolated alveolar macrophages from κB-lacZ reporter transgenic mice mock-treated or infected for 3 d with WT ( KN99α ) C . neoformans . Western blot analysis of total protein extracts revealed that although cyclin-D1 levels were unaffected , both MAD2 and Skp2 levels were diminished in alveolar macrophages upon in vivo infection ( Figure 8E ) . Thus even in tissue macrophages in vivo , down-regulation of MAD2 and a cell cycle control protein such as Skp2 , both regulated by NF-κB , were triggered by a fungal pathogen .
Regulation of host cell survival by pathogens has emerged as a way to control progression of innate immune responses upon infection . In this study , we report that C . neoformans directly affected two host functions essential for macrophage viability , apoptosis and cell cycle , in an NF-κB-dependent manner and describe for the first time induction of aneuploidy by a fungal pathogen . We show that in macrophages C . neoformans induced both the classical and alternative NF-κB activation pathways . Analyses of mice harbouring a knock-in of a IKK1-kinase dead mutant at the ikka locus [29] or mice with myeloid-specific IKK2 knock-out have shed light on a peculiar function of IKK1-dependent signalling as well as on myeloid-specific IKK2-dependent signalling [41] in the suppression of M1 macrophage activation . Notably , increase in nuclear p50/p50 dimers , previously described to direct macrophages towards a M2 anti-inflammatory phenotype [42] , was also elicited by this fungal pathogen in macrophage-like cells J774 ( Figure 4B lanes 1 , 2 ) . Hence , rise in nuclear pools of p50 homodimers , p50- and p52-heterodimers may in part explain how NF-κB contributes to the immune tolerance triggered by unopsonised C . neoformans . Importantly , analysis of infected κB-lacZ reporter transgenic mice in experimental conditions that mimicked the systemic infection in humans indicated that such activation of NF-κB by C . neoformans occurs in vivo . Fungal infection compromised macrophage viability in two ways . First , as other microbes [43] C . neoformans induced macrophage apoptosis . Programmed cell death of inflammatory cells is one of the physiological mechanisms that contributes to the resolution of inflammation [44] since their apoptosis decreases tissue damage and limits the inflammatory response . Use of WT as well as mutant strains of C . neoformans disclosed an essential PCD-promoting role of the capsule , in accordance with previous reports obtained with purified C . neoformans capsular polysaccharides in T cells or macrophages [21]–[24] , [45] . Fungal-induced PCD might thus result from direct interaction of the fungus with the macrophage , although recognition of unopsonised C . neoformans by macrophages is usually poor , or from isolated capsule , as capsule shedding commonly occurs during cryptococcal infections [8] . Analysis of stable clones with suppressed NF-κB activity or impaired IKK1- or IKK2-dependent signalling , or primary BMDM without alternative NF-κB activation revealed that fungal-induced apoptosis required differentially these activation pathways upon infection . The alternative pathway and IKK1-dependent signalling were essential , while the classical pathway of NF-κB activation and IKK2-dependent signalling controlled its onset through activation of caspase-8 , -3 , -9 and PARP cleavage . NF-κB can behave as a cell death promoting or protecting factor depending on the nature of stimulus and cell type involved , which together determine cell fate [46] . As such , IKK1-dependent signalling is essential for group B Streptococcus-induced apoptosis of macrophages , as it is here for a fungal pathogen [29] . Proapoptotic activity is thought to proceed through some of NF-κB transcriptional targets , including FAS and its ligand or TRAIL receptors and TRAIL [47] . Consistently , upregulation of FASL and TRAIL-R1/DR4 receptor was detected in J774 cells 24 h p . i . . Targeting host cell multiplication through modulation of the cell cycle is another survival scheme commonly followed by bacteria and viruses [48] , [49] . We report here that C . neoformans significantly affected macrophage viability , proliferation and cell cycle progression . It is noteworthy to point out that in our experimental settings for the various cell types used ( macrophagic-like J774 cells , BALB/c or C57BL/6 BMDM ) , macrophage viability mainly reflected cell growth , as fungal-induced apoptosis remained low ( around or below 10% ) , except for BALB/c BMDM where it reached up to 27% 72 h p . i . . GXM- , O-acetylation- or GalXM-deficient strains behaved as WT serotype A C . neoformans strain . These results indicated that inhibition of macrophage viability occurred in a capsule-independent fashion . They differ from the inhibition of human T cell proliferation reported by purified GalXM , as the latter might display induction of apoptosis , an effect induced by the high concentrations of polysaccharide used in these experiments that are not reached in our experimental settings when using the whole yeast [21] . The fact that serotype D C . neoformans strain had no significant impact on cell viability at M . O . I . 5 ( data not shown ) and led to a mild effect even at high M . O . I . ( 20 ) , with a viability curve which paralleled that of mock-treated cells , suggests that here C . neoformans-induced inhibition of viability at high M . O . I . most probably reflected an increase of apoptosis . Inhibition of proliferation per se is therefore a likely specific feature of C . neoformans var . grubii ( serotype A ) strain , which diverged 18 million years ago from C . neoformans var . neoformans ( serotype D ) strain [50] . It further demonstrates that this effect , which we have shown to require both cell contact and viable yeasts , is not due to a general stress response , but is genuinely induced by this fungus . Moreover we have established that analogous cell cycle alterations as well as mitotic regulators modifications were produced by WT and acapsular serotype A strains . Collectively these findings suggest that , as regards the pathogen side , the mechanism involved in the impairment of cell cycle and aneuploidy in macrophages is likely complex and will probably involve several fungal components . Consistent with our in vitro findings , inhibition of cell growth triggered by C . neoformans occurred in vivo in the whole spleen of κB-lacZ transgenics but surprisingly was not restricted to macrophages . This general cell growth inhibition by C . neoformans in vivo might thus result from other and yet undetermined mechanisms . C . neoformans compromised macrophage cell growth through perturbation of the cell cycle . Remarkably , phagocytosis of C . neoformans by macrophages was reported to transiently stimulate progression from G1 to S in both macrophage-like cells and primary BMDM with a concomitant increase of cyclin-D1 expression levels [51] , [52] . However , when inside the macrophages , phagocytosed live yeasts suppressed BMDM growth by decreasing cyclin-D1 expression [52] . Thus according to its presence in the milieu as an extracellular or intracellular pathogen and to its survival strategy , C . neoformans modulates macrophage cell cycle for its own profit . Combined regulation of various cyclins and their corresponding kinases or kinase inhibitor was orchestrated by C . neoformans infection and promoted cell cycle impairment at the S and G2/M phases in J774 cells with a requirement for NF-κB distinct to that of PCD . Here , the alternative pathway of NF-κB activation was dispensable , whereas the classical pathway and the IKK2-dependent signalling were essential . Reduction or rise of classical NF-κB dimers similarly prevented fungal-induced inhibition of cell viability and alteration of the cell cycle . A critical NF-κB dosage seems therefore to be required for fungal-triggered perturbation of cell cycle and viability . In another context , survival of lymphocyte progenitors was also shown to rely upon a limited range of NF-κB activity [53] . Several cell cycle checkpoint effectors have been identified as NF-κB target genes , including cyclin-D1 and Skp2 [54] , [55] . NF-κB regulation of the cell cycle is complex . It has been shown to control the G1/S transition of the cell cycle and to be required also for G2/M progression [56] . P52 or p50 heterodimers with RelB or c-Rel were associated with decreased expression of cyclin-D1 and Skp2 in S and G2 phases . Enforced expression of c-Rel causes growth arrest at the G1/S transition [57] and p65 down-regulates cyclin-E gene expression [58] . The induction of p52- and p65-heterodimers ( Figure 4B ) by fungal infection could therefore explain in part the down-regulation of cyclin-D1 , -E and Skp2 . Interestingly , when analysing the behaviour of tissue macrophages upon in vivo infection by C . neoformans , only reduction of Skp2 levels were observed in alveolar macrophages . The absence of cyclin-D1 variation is most likely due to the limited proliferative potential of these cells . Anyway , it remains that the significant decrease of Skp2 levels in tissue macrophages , both an NF-κB target and an important cell cycle regulator , argues strongly in favour of the in vivo relevance of this cytotoxicity mechanism . Modification of host cell ploidy through perturbation of cell cycle checkpoints is often triggered by viruses , including HIV and HCV [59] , [60] , and bacteria , such as H . pylori or E . faecalis [61] . Our results display the first report of induction of aneuploidy by a fungal pathogen . All observations have been made on both macrophagic-like cell line and primary cells ( BMDM ) indicating that our data cannot be attributed to specificities of cancer or immortalized cells , which are prone to become aneuploid . Moreover , the absence of a significant increase in γ-H2AX foci upon C . neoformans infection of macrophages argues against fungal activation of the DNA damage signalling cascade . It also suggests that mammalian DNA integrity is not directly affected by C . neoformans infection . Decreased levels of the spindle assembly checkpoint protein Mad2 were observed in whole cell extracts or cell-sorted infected macrophages at the S and G2/M phases . Genetic analyses of both human cancer cells and murine primary embryonic fibroblasts harbouring only one allele of Mad2 have correlated Mad2 haplo-insufficiency to defective mitotic checkpoint , elevated rate of chromosome missegregation and aneuploidy [39] . Therefore , it is likely that fungal-induced aneuploidy may result from Mad2 down-regulation . Moreover , this aneuploidy depends on NF-κB activation since fungal-induced modification of DNA content and changes in Mad2 levels disappeared in stable J774 clones with modulated NF-κB activity or signalling . Importantly , analyses of alveolar macrophages from mock-treated or C . neoformans-infected mice revealed a significant decrease of MAD2 levels upon infection , suggesting that aneuploidy may also be triggered in vivo by this fungus . Together , these findings provide novel insight into our understanding of the mechanisms whereby a fungal pathogen hijacks and shapes the host immune response to its own benefit through in part uncoupling of NF-κB activity in apoptosis , and cell cycle impairment and aneuploidy . These findings may have also wider implications as studies with mouse models of chromosome instability [62] have shown that aneuploidy may directly contribute to tumour formation . More specifically Mad2 haplo-insufficiency leads to high frequency of lung carcinoma [39] and carcinogenesis is enhanced when p53 is absent [63] . In certain contexts and when oncogenic or tumour-suppressor loci are mutated , fungal infection might therefore potentially participate via aneuploidy induction to tumourigenesis .
This study was carried out in strict accordance with the French and European regulations on care and protection of the Laboratory Animals ( EC Directive 86/609 , French Law 2001-486 issued on June 6 , 2001 ) . Animals were housed in the Institut Pasteur animal facilities accredited by the French Ministry of Agriculture to perform experiments on live mice ( accreditations # A 75 15-27 and B 75 15-05 issued on November 12 , 2004 and May 22 , 2008 respectively ) . The protocol was approved by the veterinary staff of the Institut Pasteur animal facility and was performed in compliance with the NIH Animal Welfare Insurance #A5476-01 issued on 02/07/2007 . All efforts were made to minimize suffering during animal handling and experimentation . Murine macrophage cells , construction of J774 stable clones with NF-κB-modulated activity or signalling , transgenic mice and infection conditions are described in Protocol S1 . Wild-type and mutant C . neoformans var . grubii ( serotype A ) strains , C . neoformans var . neoformans ( serotype D ) strain and growth conditions are described in Protocol S1 . Reichert DIC imaging , apoptosis and viability assays were done as described in Protocol S1 . Cultured cells or spleen tissue sections were processed for immunocytochemistry as described in Protocol S1 . Details of cell cycle analysis by flow cytometry , karyotype obtention and sorting of J774 cells at the various phases of the cell cycle are provided in Protocol S1 . EMSA , NF-κB-luciferase assay and immunoblot analysis were performed as described in Protocol S1 . Statistical analysis was done as described in Protocol S1 .
|
Cryptococcus neoformans , the only encapsulated pathogenic yeast , is responsible for severe opportunistic meningoencephalitis mostly in immunocompromised patients . It is a facultative intracellular pathogen and , as such , has the ability to survive intra- and extracellularly . Whereas interactions of C . neoformans with macrophages , especially its phagocytosis escape and intracellular survival , have been intensively studied , little is known about other schemes allowing extracellular survival of this yeast and restraint of host innate immune response . Here , we report that Cryptococcus neoformans compromised macrophage viability in two ways . Firstly , fungal infection elicited a strong decrease in macrophage proliferation in a capsule-independent fashion . This inhibition was subsequent to fungal-induced cell cycle disruption and chromosome aberrations ( aneuploidy ) , a phenomenon commonly triggered by bacteria or viruses but for the first time described for a fungus . Secondly , this pathogen promoted apoptosis in a capsule-dependent manner . Our findings unravel a new process by which a fungal pathogen dampens the immune response using uncoupled activity of NF-κB ( a key regulator of cell growth , apoptosis and inflammation ) in fungal-induced apoptosis and inhibition of cell proliferation . This may have larger implications for human health , as animal models suggest that aneuploidy promotes tumourigenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"mycology",
"molecular",
"cell",
"biology",
"cell",
"biology",
"chromosome",
"biology",
"immunity",
"biology",
"microbiology"
] |
2012
|
Fungal-Induced Cell Cycle Impairment, Chromosome Instability and Apoptosis via Differential Activation of NF-κB
|
Cereal storage proteins are major nitrogen sources for humans and livestock . Prolamins are the most abundant storage protein in most cereals . They are deposited into protein bodies ( PBs ) in seed endosperm . The inner structure and the storage mechanism for prolamin PBs is poorly understood . Maize opaque10 ( o10 ) is a classic opaque endosperm mutant with misshapen PBs . Through positional cloning , we found that O10 encodes a novel cereal-specific PB protein . Its middle domain contains a seven-repeat sequence that is responsible for its dimerization . Its C terminus contains a transmembrane motif that is required for its ER localization and PB deposition . A cellular fractionation assay indicated that O10 is initially synthesized in the cytoplasm and then anchored to the ER and eventually deposited in the PB . O10 can interact with 19-kD and 22-kD α-zeins and 16-kD and 50-kD γ-zeins through its N-terminal domain . An immunolocalization assay indicated that O10 co-localizes with 16-kD γ-zein and 22-kD α-zein in PBs , forming a ring-shaped structure at the interface between the α-zein-rich core and the γ-zein-rich peripheral region . The loss of O10 function disrupts this ring-shaped distribution of 22-kD and 16-kD zeins , resulting in misshapen PBs . These results showed that O10 , as a newly evolved PB protein , is essential for the ring-shaped distribution of 22-kD and 16-kD zeins and controls PB morphology in maize endosperm .
Cereal crops are major source of the world’s food supply [1 , 2] . Many storage proteins in cereal grains are valuable nitrogen sources for humans and livestock [3] . Prolamins are major storage proteins that occur only in cereal grain endosperm [4] . Studies of the protein storage mechanism are critical for improving cereal grain nutritional quality . However , the exact storage mechanisms of prolamins remain poorly understood . Maize ( Zea mays ) is the first staple cereal crop in the world . Maize kernels primarily contain 70–75% starch , 8–10% protein , and 4–5% lipids [5] . More than 60% of proteins are prolamins and are also known as zeins . Zeins are divided into four major classes , namely α-zeins ( 19 and 22 kD ) , β-zein ( 15 kD ) , γ-zeins ( 16 , 27 , and 50 kD ) , and δ-zeins ( 10 and 18 kD ) according to their primary structure and different solubilities . α-Zeins are encoded by a multigene family , and other zeins are encoded by single-copy genes [6–8] . Zeins are synthesized in endosperm cells [9] by membrane-bound polyribosomes , transported into the lumen of the rough endoplasmic reticulum ( RER ) and then aggregated into the formation of protein body ( PB ) [10 , 11] . The exact process of zein deposition into PB remains unclear . However , the immunolocalization of different zeins showed that they have a specific localization during the PB packing process [12] . Initially , only β- and γ-zein proteins accumulate in the PBs , and then α- and δ-zeins penetrate into the matrix formed by β- and γ-zeins and expand the PBs into larger spherical structures . The β- and γ-zeins then become a thin layer that surrounds the PBs [12 , 13] . The 19-kD α-zein is located throughout the PB core , but the 22-kD α-zein is located only in a discrete ring at the interface between the α-zein-rich core and the 27-kD γ-zein-rich peripheral region [14] . However , the mechanism that underlies distinct zein distribution patterns remains poorly understood . Studies of opaque endosperm mutants shed light on the development of PBs . The first class of opaque mutants reduced the zein content , resulting in small , unexpanded PBs , such as opaque2 ( o2 ) , o6 , o7 , and mto140 [15–18] . The second class of opaque mutants , such as floury2 ( fl2 ) , fl4 , De-B30 and Mc , has phenotypes that are derived from defective zeins [19–22] . The third class of opaque mutants , such as floury1 ( fl1 ) and o1 , does not cause qualitative or quantitative changes in the zeins . fl1 encodes a PB membrane protein that affects the distribution of 22-kD α-zein in the PBs . In fl1 , the 22-kD α-zein is dispersed more randomly throughout the PBs [14] . o1 encodes a plant-specific myosin protein that is responsible for ER morphology and motility , therefore forming smaller but more PBs in the endosperm [23] . These studies indicate that the perfect packing of zeins in PBs depends on not only zeins but also other proteins , but thus far , only a few have been studied . In this study , we cloned and studied opaque10 ( o10 ) , a new opaque mutant that affects zein assembly and PB morphology . O10 encodes a new protein that is only present in cereal plants . O10 has unique sequence features that are responsible for its unique sub-cellular distribution and accumulation . O10 directly interacts with zein proteins and is required for the ring-shaped distribution of 22-kD and 16-kD zeins in PBs in particular . The cloning of O10 expanded our understanding of the mechanism of proper zeins distribution within PBs and its relations to PB shape and endosperm opacity .
o10 is a putative ethylmethane sulfonate ( EMS ) -induced mutant that was generated by M . G . Neuffer and named by Oliver Nelson [24] . The o10-N1356 mutant from the Maize Genetics Cooperation Stock Center was crossed into the W22 genetic background . On the F2 ears , the progenies exhibit a 3:1 segregation of wild-type ( +/+ and o10/+ ) ( vitreous ) and opaque ( o10/o10 ) kernel phenotypes , indicating that o10 corresponded to a single recessive gene mutation . The o10 kernels exhibited a typical opaque endosperm phenotype when viewed on a light box . The transverse section of the mature kernels showed a much thinner layer of vitreous endosperm than that of the wild-type ( Fig 1A ) . A scanning electron microscopy ( SEM ) examination of the mature kernel endosperm showed that the starch granules in the peripheral region of the mutant kernels were loosely packed ( S1A Fig ) . To examine if there was a corresponding change in the storage components of the mutant kernels , we examined the major biochemical components in the mutant and wild-type kernels . A quantitative analysis showed that there was no distinct change in the zein , non-zein , and total proteins in opaque kernel endosperm compared the wild-type ( S1D Fig ) . The results of the SDS-PAGE also indicated that there was no obvious difference in the major zein components between the mutant and wild-type endosperm at either the mature or developing stages ( S1C Fig ) . We also analyzed the total starch and lipid contents but did not find obvious differences between the o10 and wild-type kernels ( S1E and S1F Fig ) . The endosperm resin section at 12 d after pollination ( DAP ) and 18 DAP showed that the starch granule development was not affected in o10 ( S1B Fig ) . Transmission electron microscopy ( TEM ) revealed that PBs were spherical in the wild-type endosperm , whereas many PBs were irregularly shaped in the o10 endosperm ( Fig 1B ) . Based on these results , O10 might function in PB formation . We used a map-based approach to clone O10 . After characterizing a population of approximately 12 , 000 opaque mutant kernels and 1 , 500 wild-type kernels from the F2 population , the o10 gene was placed between the molecular markers SSR217401-a and Indel557-1 , which is a region that encompasses a physical distance of 110 kb . This region was covered by 2 BAC clones ( AC204586 and AC217401 ) . There are five candidate genes within the interval ( Gene 1: AC217401 . 3_FG004 , Gene 2: GRMZM2G510280 , Gene 3: GRMZM2G346263 , Gene 4: GRMZM2G510235 and Gene 5: GRMZM2G044557 ) ( Fig 1C ) . Sequence comparison of the five candidate genes between wild-type and mutant alleles revealed a typical EMS-induced G/A transition at the 3’ end of intron 6 in Gene 3 ( Fig 1D ) . This point mutation resulted in the retention of a 95-bp intron ( intron 6 ) , causing a frame-shift and a premature stop codon in the mature transcript . To confirm that Gene 3 is the O10 gene , functional complementation was performed using a transformed wild-type candidate gene allele . Phenotyping and genotyping analysis indicated the transgenic kernels carrying the transformed wild-type Gene 3 sequence functionally complemented the lost function of o10 and rescued the mutant phenotype ( S2A–S2C Fig ) . We also made an RNAi construct to target the Gene 3 sequence for maize transformation . Phenotyping and genotyping analysis indicated that all the transgenic kernels exhibited an opaque phenotype ( S2D–S2F Fig ) . These results indicated that Gene 3 is the O10 gene . O10 encodes a protein with 1 , 059 amino acids , but in o10 , it encodes 1 , 009 amino acids , with the first 963 amino acids being identical to those of the wild-type protein ( Fig 1E ) . In using the antibody ( anti-O10 ) that was raised against the N-terminal ( 1–541 aa ) region of this protein , a protein with an apparent molecular weight of approximately 220 kD was detected in all the wild-type and o10 mature kernels; however , the protein contents of the o10 kernels dramatically decreased compared to those of the wild-type . In using the antibody ( anti-O10-C ) that was raised against the C-terminal ( 1020–1059 aa ) region of the protein , the protein was detected in the wild-type mature kernels but not in mutant mature kernels ( Fig 1F ) . NCBI BLAST searches using the full-length O10 revealed that O10 is a novel protein that is only present in cereal species ( Fig 2A ) . The N-terminal portion ( 1–268 aa ) of O10 has a homologous sequence in different plants . Most of these plants are dicotyledons , such as Glycine max and Nicotiana tomentosiformis , but surprisingly not Arabidopsis ( Fig 2A ) . In the middle portion , O10 has a repeating region containing seven unknown repeats ( 269–786 aa ) . The homologous sequences of these repeats are only present in cereal species . However , the repeat number in different cereals is highly variable . In maize , the repeat number is seven ( Zea mays-1 to Zea mays-7 ) , and in sorghum , the number is two ( Sorghum bicolor-1 and Sorghum bicolor-2 ) . In other cereal species , only one repeat is found ( S3A Fig ) . O10 has a predicted transmembrane domain in its C terminus ( 1002–1024 aa ) . The transmembrane domain in the C terminus is highly conserved among cereal species . It is also sporadically present in plants ( Nelumbo nucifera ) , insects ( Bactrocera cucurbitae ) and even in bacteria ( Chryseobacterium palustre ) ( S3B Fig ) . A real-time quantitative PCR analysis was used to examine O10 gene expression in different tissues . The results indicated that O10 is most significantly expressed in immature kernels; however , lower levels of expression were also detected in other tested tissues ( Fig 2B ) . An immunoblot analysis with an O10-specific antibody ( anti-O10 ) showed that the O10 protein significantly accumulated in the kernel but was barely detected in the tassel , ear , silk , leaf and sheath ( Fig 2C ) . In developing kernels , O10 expression was clearly detected at 7 DAP and sharply increased and peaked at approximately 11 DAP . The expression then gradually decreased until 35 DAP , with a second peak at approximately 23 DAP ( Fig 2D ) . The O10 protein was only detectable at 11 DAP by an O10-specific antibody ( anti-O10 ) and then gradually increased during kernel development and continuously accumulated until the late development stage ( 35 DAP ) ( Fig 2E ) . In mature kernels , the O10 protein also highly accumulated . This result indicated that the O10 protein continuously accumulated during kernel development . In using the antibodies ( anti-O10 and anti-O10-C ) , the apparent size of the detected protein was approximately twice that of the predicted one ( 114 . 58 kD , http://www . expasy . org/tools/pi_tool . html ) . Gel filtration chromatography also indicated that O10 was primarily detected in fraction 4 , which is larger than 158 kD ( fraction 5 ) ( S4A Fig ) . The full-length ORF of O10 expressed in E . coli also showed a similar apparent size ( Fig 3A and 3B ) . We speculated that the O10 protein likely presented as a dimer; however , dimerization likely did not occur through a di-sulfide bond between cysteines because strong denaturing treatment did not disrupt the dimer ( S4B Fig ) . The full-length of O10 ORF was further divided into three fragments and expressed in E . coli ( Fig 3A ) . Only the induced protein of the second fragment containing the seven-repeat portion still displayed an increased apparent size , which was approximately twice the predicted size ( Fig 3B ) . Therefore , the seven-repeat sequence in the middle portion of O10 was responsible for forming a potential dimer . Previous studies indicated that the resistance of the dimer to denaturant is common to several proteins that form hydrophobic intermolecular interactions [25 , 26] . Thus , we analyzed the hydrophobicity of the seven-repeat domain of O10 and found that every repeat had a hydrophobic segment . We postulated that the dimer of O10 might be formed by the intermolecular hydrophobic forces of the seven-repeat domain ( S5 Fig ) . Gene cloning and sequence analysis indicated that the o10 mutation led to the loss of the C-terminal O10 transmembrane domain , which might have a sub-cellular membrane-localization function . The ORFs corresponding to wild-type O10 and mutated o10 were fused to the C terminus of YFP to produce the YFP-O10 and YFP-o10 constructs , respectively . These constructs were used to agroinfiltrate N . benthamiana leaves . A confocal microscope observation showed that YFP-O10 fluorescence was primarily presented by some unknown spherical vesicles ( Fig 4A ) . We co-expressed YFP-O10 with an ER marker ( mCherry-HDEL ) in N . benthamiana leaves , and the results showed that these unknown spherical vesicles could co-localize with the ER around the nucleus ( Fig 4B ) . However , these spherical vesicles were not observed in the YFP-o10 agroinfiltrated cells ( Fig 4C ) . YFP-o10 could co-localize with ER throughout the cells ( Fig 4D ) . These results indicated that the transmembrane domain was responsible for the localization of O10 in N . benthamiana cells . To explore the effects of the o10 mutation in maize kernels , we first compared the gene expression of O10 in the wild-type and mutant by real-time quantitative PCR . The wild-type and mutant kernels at three developmental stages ( 15 DAP , 18 DAP and 21 DAP ) were analyzed . The results indicated that at the transcript level , O10 expression exhibited no obvious differences between the wild-type and mutant ( S6 Fig ) . We then used an immunoblot with an O10-specific antibody ( anti-O10 ) to examine the O10 protein levels in the wild-type and mutant . Compared to the O10 in the wild-type , the o10 in the mutant kernels had similar protein accumulation levels at 15 DAP and 18 DAP . However , o10 dramatically decreased at 21 DAP in the mutant ( Fig 4E ) . A similarly low level of o10 was also found in the mature kernels of the mutant ( Fig 1F ) . Therefore , the mutation in o10 does not affect its gene transcription but greatly affects its protein accumulation during kernel development . We then used a cellular fractionation assay to detect the O10 distribution in the subcellular fractions during kernel development . The total proteins extracted from developing kernels of wild-type and o10 at 15 DAP , 18 DAP and 21 DAP were then separated into soluble and total membrane fractions , respectively . An immunoblot analysis was performed with the O10-specific antibody ( anti-O10 ) . Both O10 and o10 had significant protein levels in the soluble ( cytoplasmic ) fractions , and no difference was found between wild-type and mutant kernels at 15 DAP , 18 DAP and 21 DAP ( Fig 4F ) . However , in terms of total membrane fractions , o10 dramatically decreased in the mutant kernels in compared to the wild-type ( Fig 4G ) . This result indicated that the mutation in o10 greatly affected the partitioning and accumulation of o10 in the total membrane fraction . We further isolated the ER and PB fractions from the total protein of 21-DAP wild-type and o10 kernels by discontinuous Suc gradient centrifugation . An immunoblot analysis indicated that there was an extremely high level of O10 accumulation in the wild-type within the PB fraction . However , in o10 , there was a low level of o10 accumulation in both the ER and PB fractions ( Fig 4H ) . These results indicated that in endosperm , O10 was predominantly deposited into PBs , and the mutation in o10 greatly affected the deposition of o10 into PBs . The cellular fractionation assay found O10 to be predominantly deposited into PB fractions . Because PBs are surrounded by the ER , the O10 protein may be attached to the ER or may be located inside the PBs . We then purified the PBs of 21-DAP wild-type kernels and used Triton X-100 to peel off the ER surrounding the PB core and then analyzed the peel fraction ( the surrounding ER ) and the zein-rich interior PB core by immunoblot . Like 22-kD α-zein , the immunoblot results using anti-O10 antibody indicated that O10 primarily accumulated inside the PBs , with a much lower amount in the surrounding ER fraction ( Fig 5A ) . This indicated that O10 is finally deposited inside the PBs . The immunoblot using an anti-O10-C antibody also detected a similar distribution of O10 in the PBs and surrounding ER ( Fig 5A ) . This indicated that the TMD was still present in the O10 protein when it was deposited into the PB . To confirm the interior localization of O10 in PBs , an immunolocalization analysis was performed with an anti-O10 antibody ( anti-O10 ) on ultrathin sections of 21-DAP wild-type kernels . The results indicated that O10 is indeed localized inside PBs . Moreover , we also observed a featured distribution of O10 in PBs at the interface between the 19-kD α-zein-rich region ( the inner darker region ) and the γ-zein-rich region ( the peripheral lighter region ) ( Fig 5B ) . The specialized localization of O10 in PBs suggested its potential function in association with zeins . A protein-protein interaction assay then was performed between O10 and different zeins with a yeast two-hybrid system . A full-length ORF ( O10-full ) and its three fragments , including the N terminus of O10 ( O10-N , 1–268 aa ) , the middle region comprising the repeats ( O10-M , 269–786 aa ) and the C-terminal region ( O10-C , 787–1000 aa ) , were inserted into pGBKT7 as baits . The O10-full , O10-M and O10-C baits all exhibited self-activation in yeast and were not suitable for further analysis . Only O10-N passed the self-activation test and was used for the interaction assay . The coding sequences 19-kD α-zein ( AF546188 . 1 ) , 22-kD α-zein ( GRMZM2G044625 ) , 15-kD β-zein ( GRMZM2G086294 ) , 16-kD γ-zein ( GRMZM2G060429 ) , 27-kD γ-zein ( GRMZM2G138727 ) , 50-kD γ-zein ( GRMZM2G138689 ) , and 10-kD δ-zein without the signal peptide were inserted into pGADT7 as the preys . Yeast strain AH109 was co-transformed with pGBKT7-bait and pGADT7-preys . Matings containing both the bait and prey were spotted on SD/-Leu/-Trp medium and SD/-Ade/-His/-Leu/-Trp medium as a series of dilutions . All of the matings had similar numbers of vigorously growing colonies on SD/-Leu/-Trp media . However , only the matings between the O10-N and 19-kD α-zein , 22-kD α-zein , 16-kD γ-zein and 50-kD γ-zein prey constructs could grow on SD/-Ade/-His/-Leu/-Trp medium , while the other mating combinations could not grow . This result implied that the N-terminal portion of O10 probably had protein-protein interactions with 19-kD α-zein , 22-kD α-zein , 16-kD γ-zein and 50-kD γ-zein ( Fig 5C ) . To verify the interaction between O10 and different zeins , we performed a luciferase complementation image ( LCI ) assay . The N terminus of O10 ( O10-N ) was fused to the N-terminal domain of luciferase ( NLUC ) , and the full-length of 19-kD α-zein , 22-kD α-zein , 15-kD β-zein , 16-kD γ-zein , 27-kD γ-zein , 50-kD γ-zein , and 10-kD δ-zein without the signal peptide were separately fused to the C-terminal domain of luciferase ( CLUC ) . The results showed that the co-transfection of O10-N-NLUC with 19-kD α-zein-CLUC , 22-kD α-zein-CLUC , 16-kD γ-zein-CLUC or 50-kD γ-zein-CLUC could produce strong luciferase activity ( Fig 5D ) , indicating that O10 can interact with these types of zeins in tobacco cells . These results confirmed the interaction of O10 with 19-kD α-zein , 22-kD α-zein , 16-kD γ-zein and 50-kD γ-zein . Considering that O10 can interact with 19-kD and 22-kD α-zeins and 16-kD and 50-kD γ-zeins , an immunolocalization analysis was performed for these zeins , as well as for the 27-kD γ-zein . In 21-DAP wild-type endosperms , the 19-kD α-zein , 22-kD α-zein , 27-kD γ-zein and 50-kD γ-zein all showed distinct distributions in the PBs , in agreement with previous studies [14 , 27] . The 16-kD γ-zein had a similar location as that of the 22-kD α-zein , which was at the interface between the β- and γ-zein-rich peripheral layer and the α-zein-rich inner region . In 21-DAP o10 endosperm , the distribution of 19-kD , 27-kD and 50-kD zeins in PBs was the same as that in the wild-type . However , the 16-kD γ-zein and 22-kD α-zein changed in terms of their distinct distribution . The 16-kD γ-zein was diffused into the β- and γ-zein-rich peripheral layer , and they were no longer restricted to the interface layer . The 22-kD α-zein was also no longer restricted to the interface layer , but they diffused into the α-zein-rich core zone of PBs ( Fig 6A ) . We merged ten immunolocalization pictures of 16-kD γ-zein , 22-kD α-zein and O10 into the same PB ( Fig 6B ) . A real-time quantitative PCR analysis indicated that the expressions of 19-kD α-zein , 22-kD α-zein , 15-kD β-zein , 16-kD γ-zein , 27-kD γ-zein , 50-kD γ-zein , and 10-kD δ-zein at 18 DAP were similar between the wild-type and o10 . At 24 DAP , most of the zeins still had similar expression levels; however , the 16-kD γ-zein had a significantly increased transcript level compared to that of the wild-type ( S7 Fig ) . A more sensitive HPLC profiling analysis confirmed that the 16-kD γ-zein was the only significantly increased zein in mature o10 kernels compared to the wild-type ( S8 Fig ) . Therefore , o10 not only changed the distribution of 16-kD γ-zein in PBs but also increased its accumulation in PBs . Previous studies have shown that O1 and FL1 are two non-zein proteins that are directly related to PB development [14 , 23] . To test whether O10 , O1 and Fl1 had direct interactions , a yeast two-hybrid assay of O10-N ( 1 aa-268 aa ) with an O1-tail ( 877 aa-1520 aa ) and full-length FL1 was performed . The results indicated that the N terminus of O10 could not interact with either O1-tail or FL1 in yeast ( Fig 7A ) . The interaction between O10 and O1 or FL1 was further analyzed by an LCI assay . The full-length of O10 was fused to the C-terminal domain of luciferase ( CLUC ) , and the C terminus of O1 ( 877–1520 aa ) or FL1 ( 150–302 aa ) was fused to the N-terminal domain of luciferase ( NLUC ) . The results showed that the co-transfection of FL1-C-NLUC with O10-CLUC could produce strong luciferase activity ( Fig 7B ) , while the co-transfection of O1-C-NLUC with O10-CLUC could not produce luciferase activity . These results indicated that O10 can interact with the C terminus of FL1 in tobacco cells . A previous study indicated that O1 interacts with an hsp70-interacting protein ( HIP , GRMZM2G023275 ) [23] . We further tested whether HIP interacted with O10 by a yeast two-hybrid assay . The result indicated that O10-N might interact with HIP ( Fig 7A ) . To verify the interaction between O10-N and HIP , we performed an LCI assay . The results showed that the co-transfection of O10-N-NLUC with HIP-CLUC could produce strong luciferase activity ( Fig 7B ) . These results indicated that O10 may interact with HIP and FL1 . We also analyzed the accumulation of O10 and some types of zeins in other opaque mutants , including o1 , fl1 , fl4 , o2 and o7 . The immunoblot results showed that in o1 , fl4 and fl1 , there was no distinct change in the zein proteins compared to the wild-type , and the O10 content was decreased in o1 , fl4 but not in fl1 . In o2 , the contents of α-zeins and 50-kD γ-zein dramatically decreased; the O10 content did not change . In o7 , the content of all zeins decreased , especially that of 16-kD γ-zein and 19-kD α-zein , and the O10 content dramatically decreased ( Fig 7C ) .
Newly created genes are common in many species . Distinct molecular processes contribute to the formation of new genes through processes such as gene duplication , gene fusion , or the alteration of the existing gene structure , or genes can arise from previously non-coding DNA [28] . For example , Jingwei ( Jgw ) is a recently evolved gene in Drosophila that encodes a protein containing two domains from two different proteins . Its chimeric structure is derived from exon shuffling between the Adh and Ymp genes , and it thereby acquired new biochemical functions [29 , 30] . The high rate of this chimeric gene origination is common in plants . In rice , 380 chimeric genes are recruited from previously existing genes [31] . A BLAST sequence analysis indicated that O10 is a cereal-specific protein . The N terminus of O10 has homologs in most plants ( Fig 2 ) . However , the TMD of the C terminus is rarely present in other plants , but it does exist in some insects and bacteria ( S3 Fig ) . The middle portion of O10 with repeated regions only exists in the cereal family ( Fig 2 ) . Apparently , O10 is generated by the fusion of at least three distinct domains with independent origins . These sections were merged into one protein in cereals , most likely by similar gene fusion or shuffling mechanisms . Our results showed that the newly generated O10 protein has new biological functions that arise through a combination of different functional domains with different origins . O10’s new function is important for the development of PB in cereals; hence , such a random sequence fusion event might be functionally selected and inherited over the course of evolution . Furthermore , O10 is still a fast-evolving gene . This statement is evidenced by the highly variable repeat number of the middle portion among O10 homologous proteins in different cereals ( S3 Fig ) . Because O10 has an important function in PBs and is still subjected to fast evolution , we predict that the PBs in cereal endosperm might have a relatively recent origin and are still undergoing rapid evolution . Maize endosperm PBs consist mostly of zeins but also non-zein proteins . PBs can be induced in Nicotiana benthamiana leaves by transient transformations with Zera that are fused to a fluorescent marker protein ( DsRed ) . The proteome analysis of these induced PBs revealed 195 additional proteins in addition to Zera-DsRed , including a broad range of proteins [32] . It is reasonable to predict that maize PBs would also have many non-zein proteins that participate in the formation of PB . Three non-zein proteins were characterized in maize PB , namely BIP ( ER lumen binding protein ) , O1 and FL1 . BIP is an ER chaperone and is associated with the folding and assembly of zeins in PB [33] . O1 encodes a plant-specific myosin protein that targets ER and PB . O1 is responsible for ER morphology and motility and affects the formation of PBs on the ER [23] . FL1 is an ER membrane-localized protein and can interact with 19-kD α-zein and 22-kD α-zein . This protein controls the proper distribution of 22-kD α-zein in PBs [14] . In this study , O10 was identified as a new non-zein protein in maize PB . Our results indicated that O10 was initially synthesized in the cytoplasm and was ultimately deposited inside the PBs . To be deposited inside the PBs , O10 should undergo at least three processes as follows: transport into the ER lumen , retention in the ER lumen , and deposition into PB . A sequence analysis indicated that O10 has no predicted signal peptide; thus , there may be other factors to aid O10 in entering the ER lumen across membranes . We found that the N terminus of O10 can interact with HIP ( Fig 7 ) . HIP is a cytosolic protein that interacts with the ATPase domain of HSP70 [34 , 35] . Genetic and biochemical evidence supports a role for cytosolic HSP70s in the post-translational translocation of precursor proteins into the ER and mitochondria [36 , 37] . Therefore , chaperone HSP70 might take part in the transportation of O10 into the ER lumen . Many mechanisms can lead to protein retention in the ER lumen . The HDEL or KDEL tetrapeptides at the C terminus were identified as a retention signal in the ER for ER luminal proteins , such as in BIP [38 , 39] . The double-lysine motif ( KKXX or KXKXX ) in the C terminus of type I integral ER membrane proteins , such as Wbp1 in yeast , was identified as another ER retention signal [40] . The TMD was also shown to take part in protein retention in the ER lumen [41–43] . The o10 mutation caused the loss of the TMD in the C terminus of O10 ( Fig 1 ) . The accumulation of o10 in the ER dramatically decreased when the TMD was absent ( Fig 4 ) . These results showed that the TMD in the C terminus of O10 plays an important role in O10 retention in the ER lumen . The suppressed ER function might directly affect the targeting and transportation of O10 to PB . In fact , in opaque mutants with disrupted ER functions , such as o1 and fl4 , the accumulation of O10 also significantly decreased ( Fig 7 ) . An interaction with zeins would be consistent with the deposition of O10 inside the PBs . Among O10-interacting zeins , the 16-kD γ-zein appeared to be more important for O10 aggregation . There was a correlation between the accumulation of O10 and 16-kD γ-zein in other opaque mutants . In fl1 , the contents of different zeins were barely affected , and the O10 content did not change . In o2 , the contents of α-zeins and 50-kD γ-zein dramatically decreased; however , the O10 content did not change , as for the 16-kD γ-zein . In o7 , all of the zeins were generally down-regulated , especially 16-kD γ-zein and 19-kD α-zein , and the O10 content dramatically decreased . Coincidentally , the expression and accumulation of the 16-kD γ-zein gene increased in o10 , but the other zeins did not change ( S7 and S8 Figs ) . The increased expression and accumulation of 16-kD γ-zein might be feedback regulation related to the failure of O10 deposition inside PBs together with 16-kD γ-zein . We proposed a possible model for the cellular route of O10 ( Fig 8 ) . O10 is initially synthesized in the cytoplasm and , perhaps with the aid of HIP and HSP70 , is transported into the ER lumen . Its TMD assists in its retention in the ER , and eventually , the deposition of O10 inside PBs probably depends on the interaction with 16-kD γ-zein . Mutated o10 protein without the TMD may still be transported into the ER lumen with the aid of HIP because the O10 N terminus still can interact with HIP . Indeed , the results of subcellular localization in N . benthamiana leaves indicated that o10 co-localized with ER ( Fig 4D ) . However , the dramatic decrease of o10 in the ER fraction suggested that the loss of TMD greatly affected the proper retention of o10 in the ER lumen . The o10 in ER might be degraded afterwards . Previous studies found that the distribution of different zeins in PB showed restricted patterns . In general , the β- and γ-zeins were located in the periphery of PBs , while the α- and δ-zeins were located in the core of PB [12] . The mechanism for the special distribution of zeins is poorly understood . It was believed that the intrinsic hydrophilic-hydrophobic properties of different zeins and the direct interactions between them determined their proper distribution in PBs [44 , 45] . Further analysis revealed the distinct distributions of different α-zeins in PB . The 19-kD α-zein is located throughout the core of PB . However , the 22-kD α-zein has a special localization at the interface region between the α-zein-rich core and the γ-zein-rich periphery of PB [14] . In this study , we found that different γ-zeins also had distinct distributions . The 27-kD γ-zein and the 50-kD γ-zein were located throughout the peripheral region of PB . However , the 16-kD γ-zein , similar to the 22-kD α-zein , was also located at the interface between the core and the periphery of PBs ( Fig 6 ) . The specialized distribution of 22-kD α-zein and 16-kD γ-zein could not be fully explained by either their hydrophilic/hydrophobic properties or their interactions with other zeins . O10 could interact with both 22-kD α-zein and 16-kD γ-zein and co-localized with them ( Fig 5 ) . A loss of function of O10 would disrupt the ring-shaped distribution of 22-kD α-zein and 16-kD γ-zein ( Fig 6 ) . This suggested that O10 may be functionally responsible for the ring-shaped distribution of 22-kD α-zein and 16-kD γ-zein . Therefore , these three proteins form a newly defined structural zone of maize PBs , i . e . , the ring-shaped interface region . The ring-shaped interface was important for the morphology of PBs . The disruption of this interface region in o10 may have led to the formation of misshapen PBs There is further evidence that the zein component of this interface region , namely the 22-kD α-zein and 16-kD γ-zein , are important for PB morphology . For example , an RNAi experiment in which the 22-kD α-zein alone was reduced caused misshapen PBs [46] . However , the reduction of both 19- and 22-kD α-zeins only severely restricted PB expansion but did not affect PB morphology [47] . These data suggested that the 22-kD α-zein might have a structural role in PB morphology . A frame-shift mutation of 16-kD γ-zein in Mc1 led to angular PBs , suggesting that the 16-kD γ-zein might be functionally related to PB morphology [22 , 33] . The formation of this interface layer consisting of both α- and γ-zeins that was attached by non-zein protein O10 might function as an insulator for the hydrophilic peripheral region and the hydrophobic inner core of PB , thereby stabilizing the overall PB structure ( Fig 8 ) . Two opaque mutants , o1 and fl1 , share common features with o10 . All three mutants have very minor changes in zein contents [14 , 23] . o1 and o10 endosperm both display misshapen PBs . Although O1 and O10 do not directly interact , they share a common potential interacting protein , namely HIP [23] . The accumulation of O10 significantly decreased in o1 , which provided an explanation for the misshapen PBs in o1 endosperm . FL1 could directly interact with α-zeins , and in fl1 , the ring-shaped distribution of 22-kD α-zein was affected [14] . Since FL1 localizes in the PB membrane and 22-kD α-zein is deposited inside PBs away from the PB membrane , therefore previous researches have speculated that maybe some unidentified factors are involved in 22-kD α-zein targeting in addition to FL1 [14] . In this study , we found that FL1 could directly interact with O10 in tobacco cells ( Fig 7 ) and the distribution of 22-kD α-zein was also affected in o10 . This suggested a functional link between O10 and FL1 in determining the ring-shaped distribution of 22-kD α-zein . According to the above analysis , we speculated that O10 might firstly interact with 16-kD γ-zein in ER lumen . FL1 perhaps promotes 22-kD α-zein’s interaction with O10 near the PB membrane when O10 and 16-kD γ-zein are entering the PB ( Fig 8 ) . Then the complex consisting of O10 , 16-kD γ-zein and 22-kD α-zein is targeted to the interface region between the α-zein-rich core and the γ-zein-rich periphery of PB , by some unknown mechanism . The functional relationship among O1 , O10 and FL1 requires further investigation .
The maize o10-N1356 stock was obtained from the Maize Genetics Cooperation stock center . The mutant was crossed into a W22 genetic background to produce the F2 populations . Kernels of the F2 ears exhibited a 3:1 segregation of wild-type kernels ( o10/+ or +/+ ) and homozygous mutant kernels ( o10/o10 ) were used for analysis . The root , stem , third leave , silk , tassel , and ear tissues were collected from at least three W22 plants at the V12 stage . The seeds were used for genetic transformation as in [17] . All of the plants were cultivated in a field at the Shanghai University campus in Shanghai , China . For the protein measurements , the endosperm of o10 and wild-type mature kernels was separated from the embryo and pericarp by dissection after soaking the kernels in water . The samples were dried to constant weight , pulverized with a mortar and pestle in liquid N2 , and then measured according to a previously described protocol [17] . All of the measurements were replicated at least three times . For the lipid measurements , fifty mature kernels of the wild-type or o10 were collected from well-filled , mature ears . The kernels were pulverized with a mortar and pestle in liquid N2 , and 100 mg of dried flour was used for lipid extraction and then measured according to a previously described protocol [17] . All of the measurements were replicated at least three times . For the starch measurements , fifty mature kernels of the wild-type and o10 were ground in liquid N2 . The resulting powders were dried to a constant weight . Finally , the total starch was measured using an amyloglucosidase/α-amylase starch assay kit ( Megazyme ) . The protocol referenced the method in [20] . All of the measurements were replicated at least three times . For scanning electron microscopy , the o10 and wild-type mature kernels from the same F2 ear were prepared according to the previously described method [20] . The samples were observed with a scanning electron microscope ( S3400N; Hitachi ) . For transmission electron microscopy , immature kernels ( 21 DAP ) of o10 and the wild-type were prepared as previously described [17] . The samples were observed with a Hitachi H7600 transmission electron microscope . A population of 12 , 000 homozygous opaque kernels and 1500 wild-type kernels from F2 ears was used for gene mapping . Molecular markers that were distributed throughout maize ( Zea mays ) chromosome 1 were used for preliminary mapping . Molecular markers for fine mapping ( S1 Table ) were developed to localize the o10 locus to a 110-kb region . The corresponding DNA fragments were amplified from o10 and wild-type plants using KOD Plus DNA polymerase ( Toyobo ) and sequenced using a MegaBACE 4500 DNA analysis system ( Amersham Biosciences ) . Two O10 antibodies were produced in this study: anti-O10 and anti-O10-C . For anti-O10 antibody production , the cDNA sequence encoding the N terminus of O10 ( 1 aa–541 aa ) was inserted into pGEX-4T-1 ( Amersham Biosciences ) at the EcoRI- and BamHI-digested sites . The primers were 5’-AAAAGAATTCATGGGCATGAGCTTGCACGCCGCGCG-3’ and 5’-CGCGGATCCTCATGATCCCACTGATTCAGCTAGAG-3’ . The GST-tagged O10 fusion protein was purified using the ÄKTA protein purification system ( GE Healthcare ) with a GSTrap FF column . For anti-O10-C , the C-terminal peptides of O10 ( 1020 aa–1059 aa ) were synthesized at Shanghai ImmunoGen Biological Technology . Both antibodies were prepared by Shanghai ImmunoGen Biological Technology in rabbits according to standard protocols . For 16-kD γ-zein , 19-kD α-zein , 22-kD α-zein , 27-kD γ-zein and 50-kD γ-zein antibody production , regions of low similarity of 16-kD γ-zein , 19-kD α-zein , 22-kD α-zein , 27-kD γ-zein and 50-kD γ-zein were selected according to a previous study [48] . The cDNAs responsible for the selected polypeptides were cloned into pGEX-4T-1 ( Amersham Biosciences ) , and glutathione S-transferase-tagged fusion protein was purified with the ÄKTA purification system ( GE Healthcare ) using a GSTrap FF column . Antibodies were prepared by Shanghai ImmunoGen Biological Technology in rabbits according to standard protocols . For functional complementation transgene tests , a 3180-bp ORF sequence of O10 with two restriction enzyme sites for BamHI and SalI was cloned by PCR using the following primers: 5’-CGCGGATCCATGGGCATGAGCTTGCACGCCGCGCG-3’ and 5’-CCGCTCGAGTCAGGGACGTTTTCTCTGCCCA-3’ . The resulting fragment was cloned into a pHB vector , which carries a bar resistance marker and a CaMV 35S promoter . The CaMV 35S promoter was substituted by a 2 kb upstream DNA sequence from the start codon of O10 with two restriction enzyme sites for EcoRI and BamHI for identical O10 expression . The amplification primers were 5’-CCGGAATTCTCTGCTGCTGATGTCTTGT-3’ and 5’-GCGGATCCGGCCGTGTGCCGTAGTTT-3’ . The construct was transferred into Agrobacterium tumefaciens ( GV3101 ) . Agrobacterium-mediated maize transformation was performed according to known protocols [49] . Five independent transgenic lines were generated . The transgenic lines were all backcrossed to o10/o10 plants by two successive generations to obtain kernels with a homozygous o10 locus . The SSR217401-a marker that was linked tightly with the o10 locus was used to identify homozygous o10 kernels . To generate the RNAi construct , we used the fragment containing 384 bp from the cDNA of O10 ( 30 bp-414 bp ) according to the sequence analysis on http://rnaidesigner . thermofisher . com/rnaiexpress/ . The forward-oriented fragment was amplified with the following pair of primers: 5’-CTTCTCGAGGCACGAGGATCTGAGCTG-3’ and 5’-CTTCCATGGTTTTCCACTACTTCTACCGA-3’ with restriction enzyme sites for XhoI/NcoI . The reverse fragment was amplified with the primers 5’-CTTTCTAGAGCACGAGGATCTGAGCTG-3’ and 5’-CTTGGATCCTTTTCCACTACTTCTACCGA-3’ with restriction enzyme sites for XbaI/BamHI . The construct was transferred into Agrobacterium tumefaciens ( GV3101 ) . Agrobacterium-mediated maize transformation was performed according to published protocols [49] . Six independent transgenic lines were generated . The transgenic lines were all backcrossed with W22 for 3 generations . For RNA extraction and Real-Time PCR analysis , the protocol referenced the method in [23] . The primers that were used for Real-Time PCR analysis are listed in S1 Table . The total kernel protein ( 21 DAP ) was extracted by grinding in liquid nitrogen and suspending in extraction buffer ( 50 mM Tris-Cl , 2 . 5 mM EDTA , 150 mM NaCl , 0 . 2% NP-40 , 20% glycerol , 1 mM PMSF and 1% plant cocktail [Sigma Aldrich] ) on ice for 20 min . The lysate was then centrifuged at 12 , 000 rpm for 5 min , the pellet was discarded , and the lysate was centrifuged one more time . A Superdex 200 10/300 GL Column ( GE Healthcare ) was first equilibrated in protein extraction buffer with an ÄKTA purifier system ( GE Healthcare ) until the UV baseline was smooth . A 500-μl volume of maize kernel lysate was injected into the system , and the flow speed was adjusted to 0 . 5 ml per minute . After the first volume ( 6 ml ) flowed through , consecutive fractions of 500 μl each were then collected . The protein complex was then concentrated by acetone and analyzed by immunoblotting . The ORF sequences of O10 and o10 were amplified by PCR using KOD plus polymerase ( Toyobo ) with the primers 5’-CACCATGGGCATGAGCTTGCACGCCGCGCG-3’ and 5’-TCAGGGACGTTTTCTCTGCCCAA-3’ . Amplified fragments were subcloned into pENTR/D-TOPO with the Gateway TOPO cloning kit ( Invitrogen ) and sequenced . The right entry clone was introduced into a pB7WGY2 plant expression vector through an LR reaction of the Gateway system ( Invitrogen ) [50] . The well-established fluorescent protein marker mCherry-HDEL was used to label the ER [51] . The expression vectors were transformed into Agrobacterium tumefaciens strain GV3101 . The agro-infiltration procedure was performed as previously described [52] . The subcellular fractions of endosperm cell were prepared as previously described [23 , 53] . To peel off the ER surrounding PB , the separated PB fraction was resuspended in Buffer B ( 10 mM Tris-HCl , pH 8 . 5 , 10 mM KCl , 5 mM MgCl2 , 1 mM DTT , and proteinase inhibitor cocktail ) , added to 5% Triton X-100 [Sigma-Aldrich] , shaken at 4°C for 4 h , and then centrifuged at 10 , 000 g and 4°C for 20 min . Then , the pellets were resuspended in an equal volume of Buffer B . Immunoblot analyses were performed as previously described [14] . The purified anti-O10 , anti-O10-C and 22-kD α-zein antibodies were used at 1/500 , while the anti-tubulin antibody ( Sigma-Aldrich ) and anti-BIP antibody ( Santa Cruz Biotechnology ) were used at 1/1000 . All of the O10 bait constructs were made in pGBKT7 . For the O10-N bait , the cDNA sequence of the N terminus of O10 ( O10-N , 1 aa-268 aa ) was fused downstream from the GAL4 BD domain in pGBKT7 at the EcoRI and BamHI restriction sites . The primers were 5’-AAAAGAATTCATGGGCATGAGCTTGCACGCCGCGCG-3’ and 5’-CGCGGATCCTCATTTGCTTTCTGCTTCATCAGA-3’ . The clones were sequenced to ensure that an in-frame fusion with the GAL4 DNA binding domain had been created and that there were no mutations that would cause amino acid substitutions . For the pGADT7-preys , all of the zein-coding regions were amplified with genomic DNA ( zein genes do not have introns ) . The primers were designed to exclude the 5’ region encoding the signal peptide and to incorporate an EcoRI site at the 5’ end and a BamHI at the 3’ end for cloning into pGADT7 , with the exception of the 3’ primer for the 16-kD γ-zein , which was designed with an XhoI site , and the 5’ primer for the 50-kD γ-zein , which was designed with a NedI site . The coding regions of the O1-Tail ( 877–1520 aa ) , FL1 and HIP were amplified with maize endosperm cDNA that carried restriction sites for EcoRI/PstI , EcoRI/BamHI and SfiI/BamHI , respectively . The primers that were used for amplification are listed in S1 Table . The zeins , O1-Tail , FL1 , and HIP were cloned and inserted in-frame into pGADT7 . Yeast strain AH109 was co-transferred with pGBKT7-O10-N and pGADT7-Preys . The interaction between T-antigen and Human P53 was used as a positive control . The interaction between pGBKT7-O10-N and pGADT7-empty was used as a negative control . The putative positive clones were further spotted with a dilution series onto SD/-Leu/-Trp medium and SD/-Ade/-His/-Leu/-Trp medium . To confirm the interaction between O10 and different types of zeins or HIP , the cDNA sequence of the N terminus of O10 ( O10-N , 1–268 aa ) was cloned into JW771 ( NLUC ) , the full-length ORF of HIP , and all of the zein coding regions without the 5’ region encoding the signal peptide were cloned into JW772 ( CLUC ) , yielding O10-N-NLUC and ZEIN/HIP-CLUC constructs for the LCI assay . To analyze the interaction between O10 and O1 or FL1 , the full-length ORF of O10 was cloned into JW772 ( CLUC ) , and the cDNA sequences of C terminus of O1 ( 877 aa-1520 aa ) and FL1 ( 150 aa-302 aa ) were cloned into JW771 ( NLUC ) , yielding O10-CLUC and O1-C/FL1-C-NLUC constructs for the LCI assay . The constructs were transformed into Agrobacterium tumefaciens strain GV3101 . The agro-infiltration procedure was performed as previously described [50] , and the resulting luciferase signals were captured using the Tanon-5200 image system . These experiments were repeated at least three times with similar results . Immunolabeling was performed as previously described [20] . The purified anti-O10 antibody was used at 1/200; the 22-kD α-zein and 50-kD γ-zein antibodies were used at 1/1000; and the 16-kD γ-zein , 19-kD α-zein and 27-kD γ-zein antibodies were used at 1/5000 . To measure the zeins , 10 wild-type and o10 kernels were ground into powder in liquid N2 , and 150 mg of dry powder was dissolved in 1 ml of extract buffer ( 70% ethanol , 5% β-mercaptoethanol , and 0 . 5 sodium acetate ( w/v ) ) . The samples were shaken at room temperature for at least 4 h and centrifuged at 500 g for 20 min at room temperature . The supernatants were diluted 5 times and measured by HPLC . The relative concentrations of the zeins were computed using their peak areas . All of the measurements were repeated three times . Sequence data from this article can be found in the GenBank/EMBL data libraries under the following accession numbers: 10-kD δ-zein , AF371266; 15-kD β-zein , M12147 ( GRMZM2G086294 ) ; 16-kD γ-zein , AF371262 ( GRMZM2G060429 ) ; 18-kD δ-zein , AF371265 ( GRMZM2G100018 ) ; 19-kD α-zein , M12146 ( AF546188 . 1 ) ; 22-kD α-zein , NM_001112529 ( GRMZM2G044625 ) ; 27-kD γ-zein , AF371261 ( GRMZM2G138727 ) ; 50-kD γ-zein , BT062750 ( GRMZM2G138689 ) ; O1 , GRMZM2G449909; FL1 , NM_001112594 ( GRMZM2G094532 ) ; O10 , XM_008667043 . 1 ( GRMZM2G346263 ) ; Zm HIP , GRMZM2G023275; and Zm UBQ , BT018032 .
|
Through the positional cloning of the maize classic endosperm mutant opaque10 ( o10 ) , we identified a novel protein critical for PB morphology . O10 is a fast-evolving cereal-specific gene with recent origin . A thorough characterization of its three functional domains revealed its important functions for storage protein deposition and organization in PBs . O10 determines a ring-shaped layer in PBs through direct interaction with two major storage proteins ( 22-kD and 16-kD zeins ) . This newly characterized PB layer maintains a stable spherical morphology for PB . This study advanced our understanding of PB structure and function . Furthermore , this study demonstrated the origin of a new functional gene and the functional evolution of a storage organelle that is highly valuable to humans .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"anatomy",
"luciferase",
"chemical",
"compounds",
"molecular",
"probe",
"techniques",
"enzymes",
"immunoblotting",
"enzymology",
"cloning",
"carbohydrates",
"organic",
"compounds",
"endosperm",
"cereal",
"crops",
"plant",
"science",
"model",
"organisms",
"crops",
"molecular",
"biology",
"techniques",
"plants",
"starches",
"research",
"and",
"analysis",
"methods",
"nutrient",
"and",
"storage",
"proteins",
"grasses",
"crop",
"science",
"immunoblot",
"analysis",
"proteins",
"maize",
"oxidoreductases",
"chemistry",
"molecular",
"biology",
"agriculture",
"biochemistry",
"plant",
"and",
"algal",
"models",
"organic",
"chemistry",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"organisms",
"fruit",
"and",
"seed",
"anatomy"
] |
2016
|
Maize opaque10 Encodes a Cereal-Specific Protein That Is Essential for the Proper Distribution of Zeins in Endosperm Protein Bodies
|
Somatic mutations contribute to the development of age-associated disease . In earlier work , we found that , at high frequency , aging Saccharomyces cerevisiae diploid cells produce daughters without mitochondrial DNA , leading to loss of respiration competence and increased loss of heterozygosity ( LOH ) in the nuclear genome . Here we used the recently developed Mother Enrichment Program to ask whether aging cells that maintain the ability to produce respiration-competent daughters also experience increased genomic instability . We discovered that this population exhibits a distinct genomic instability phenotype that primarily affects the repeated ribosomal RNA gene array ( rDNA array ) . As diploid cells passed their median replicative life span , recombination rates between rDNA arrays on homologous chromosomes progressively increased , resulting in mutational events that generated LOH at >300 contiguous open reading frames on the right arm of chromosome XII . We show that , while these recombination events were dependent on the replication fork block protein Fob1 , the aging process that underlies this phenotype is Fob1-independent . Furthermore , we provide evidence that this aging process is not driven by mechanisms that modulate rDNA recombination in young cells , including loss of cohesion within the rDNA array or loss of Sir2 function . Instead , we suggest that the age-associated increase in rDNA recombination is a response to increasing DNA replication stress generated in aging cells .
One of the greatest risk factors associated with carcinogenesis is age . Cancer risk increases exponentially toward the end of life in humans and other mammalian species [1] . Somatic genetic changes contribute significantly to the development of most tumors . However , the rate at which spontaneous mutations arise in normal adult cells has been hypothesized to be too low to generate all the genetic changes necessary to produce tumors at the observed rates [2]–[3] . Consequently , Loeb et al . developed the mutator hypothesis , which postulates an increased mutation rate in precancerous cells [3] . A variety of mechanisms could lead to such an increase , but a favored model is that sporadic mutations in , or epigenetic silencing of , genes responsible for maintaining genome integrity lead to increased rates of mutation . Once acquired , this mutator phenotype may serve as the driving force toward carcinogenesis as individuals age . When genomes of tumors are examined , loss of heterozygosity ( LOH ) is observed as a common mechanism in which the sole functional allele of a tumor suppressor gene is inactivated by somatic mutation [4]–[5] . LOH can be generated by many different mutational events , including point mutations , small deletions or inversions , mitotic recombination or chromosome loss . Recent advances in high-resolution single-nucleotide polymorphism arrays ( SNP arrays ) have revealed that a surprising number of tumors contain long tracts of homozygosity that are not accompanied by a change in gene copy number . This type of LOH arises from somatic mitotic recombination events and is referred to as partial ( or acquired ) uniparental disomy ( UPD ) [6] . Importantly , UPD can alter the genotype for hundreds of genes following a single event , thereby amplifying its potential to contribute to cancer development [7] . Within any genome , there are regions that exhibit higher rates of mitotic recombination than the genomic average as the result of their proximity to hotspots or common fragile sites [8] . These regions typically represent slow-replicating sequences , and agents that generate DNA replication stress often reveal their fragility . These regions are also often associated with non-histone protein complexes that inhibit DNA replication fork progression . This generates DNA replication stress that may lead to an increased frequency of DNA damage due to replication fork collapse [9]–[10] . While conservative repair of this DNA damage would have no genetic consequence , the higher incidence of damage results in a greater chance that an alternative repair pathway will be utilized that does confer a genotypic change . Previously , we found evidence for a mutator phenotype associated with advancing replicative age in a common lab strain of the budding yeast , Saccharomyces cerevisiae . Pedigree analysis revealed that old cells begin to produce offspring that have dramatically higher incidences of genomic instability , which is manifest as an apparent ∼100-fold increase in LOH on at least two different chromosomes [11] . Virtually all the LOH occurred via mitotic recombination and gave rise to UPD genotypes . Subsequent analysis revealed that these LOH events were a consequence of loss of mitochondrial DNA in daughter cells , which led to a transient “crisis” state characterized by cell cycle arrest and high mortality [12] . Cells that survived this crisis showed a high frequency of LOH events in their nuclear genome . While yeast cells lacking mitochondrial DNA cannot perform oxidative phosphorylation ( respiration ) , they remain viable by relying on aerobic glycolysis ( fermentation ) . In contrast , most eukaryotic cells retain , and even require , respiration . Thus we were interested in determining whether S . cerevisiae cells that retain respiratory competence throughout their replicative life span ( RLS ) also exhibit a mutator phenotype . The frequency at which functional mitochondria are successfully segregated during cell division varies widely between strains of S . cerevisiae; alleles in over 100 genes can affect mitochondrial DNA transmission frequency [13]–[15] . Therefore , in this study we sought to use a strain in which respiration competence was faithfully transmitted with increasing replicative age . We recently developed the Mother Enrichment Program ( MEP ) , a genetic program to facilitate replicative aging studies in yeast [16] . The MEP provides an efficient and inducible selection against newborn daughter cells . When the MEP is active , mother cells continue to divide and age normally , while the division of newborn daughter cells is arrested . The MEP provides the opportunity to follow a cohort of mother cells in liquid culture throughout their entire RLS without any requirement for removing progeny cells . Once cells have reached a desired age , the MEP can be switched off and aged mothers will resume production of viable daughters , allowing for colony-based phenotypic analysis . Compared to pedigree analysis , which is done by single-cell micromanipulation , the MEP can dramatically improve the sensitivity of the LOH assay by increasing the sample number of aged cells by over three orders of magnitude . Here we report our finding that replicative age is accompanied by a progressive decline in rDNA array stability , leading to higher incidence of LOH affecting the right arm of chromosome XII .
In order to assess whether MEP strains could be useful for analyzing LOH rates in aging yeast cells , we first examined whether old mother cells produced daughters that retained respiration competence . Specifically , we examined the inheritance of respiration competence in daughter cells of a MEP strain ( UCC5185 ) using pedigree analysis . We found that ∼65% of the mother cells produced daughters with respiration competence through their entire RLS ( Median of 36 generations , Figure S1A ) . This is in striking contrast to the strain we originally examined for age-associated LOH ( UCC809 ) , where only ∼5% of the mother cells produced respiration-competent daughters throughout their life span ( Figure S1A and [11]–[12] ) . Furthermore , for those UCC5185 mother cells that did produce respiration-incompetent daughters , the median age at which this occurred was significantly later in the mother's life span ( 23 generations versus 10 generations in UCC809 ) ( Figure S1B ) . Together , these results demonstrate that the majority of UCC5185 cells are capable of producing respiration-competent daughters throughout their life span , thus providing the basis for our analysis of LOH in such a population . LOH events in diploid cells can be visualized using colony color phenotypes and , when combined with half-sector analysis , the rate of LOH can be determined [17]–[20] . We used two heterozygous markers that alter colony color when lost: Loss of MET15 function , resulting in a black colony sector , or loss of ADE2 function , resulting in a red colony sector . LOH events that occur within the first cell division after plating will generate sectors that form one half of a colony and can be used as a direct measure of mutation rate . Replicative life span is measured in terms of the number of mitotic divisions an individual cell completes before senescence . To measure LOH rates as mother cells divide and age , logarithmically growing MEP cells were inoculated into rich media containing estradiol and aged in liquid culture . Estradiol activates the MEP , allowing mother cells to divide and age normally while daughter cells are rendered incapable of cell division . Throughout the 95-hour aging period , samples were harvested and washed to remove estradiol , then plated to solid medium for half-sector analysis . Once the MEP is inactivated by removal of estradiol , surviving mother cells can produce viable daughter cells and form colonies . Because aged mother cells represent the only cells in the aging liquid culture capable of forming colonies ( Figure 1A and [16] ) , no fractionation of cell populations was required to isolate aged mothers for the analysis . The vast majority of colonies at each time point grew robustly and were respiration-competent . Respiration-competent colonies were easily distinguished from those incapable of respiration , which show a severe growth defect on glucose media and thus could be excluded from our analysis . We determined the LOH rates in young cells at loci on chromosome IV and chromosome XII simultaneously . By scoring ∼20 , 000 colonies per time point , we measured an LOH rate of 6 . 69×10−4 events/cell division at MET15 on chromosome XII and a rate of 1 . 25×10−4 events/cell division at an intergenic region on chromosome IV ( Figure 1A and 1B ) . The LOH rate at MET15 was significantly higher than the chromosome IV locus ( p = 0 . 0059 , Fisher's exact test using contingency tables ) , suggesting that MET15 LOH rate is affected by proximity to a “hotspot” . As mother cells were aged in liquid culture , a robust and significant increase in LOH rates associated with increasing replicative age was observed at the MET15 locus ( Figure 1A ) . This increase was significant at 45 hours ( p<0 . 0001 , Fisher's exact test using contingency tables ) , a time point where ∼25% of the original mother population retained viability , and continued to increase as the viability of the population declined below 10% . Thus , the increase in LOH rate at MET15 affects cells that have exceeded the median life span potential of the population . In contrast , while there was a trend towards increasing LOH rates with increasing replicative age for the first 72 hours at the chromosome IV locus , it did not reach the level of significance at any age ( p>0 . 05 , Fisher's exact test using contingency tables ) ( Figure 1B ) . These results suggested that this age-associated increase in LOH was locus-specific , affecting MET15 on chromosome XII . The MET15 locus lies ∼250 kbp distal to the rDNA array , a series of ∼150–200 tandemly repeated copies of the genes encoding the structural RNA components of the ribosome [21] . To determine if the rDNA array represented a “hotspot” that was responsible for elevated LOH rates at MET15 , we constructed a diploid strain that carried heterozygous markers immediately adjacent to each end of the rDNA array and at MET15 ( Figure 1C ) . In aged populations , we isolated half-sectored colonies based on LOH at MET15 and scored each sector for the presence of the rDNA-linked markers on chromosome XII . We found that 90% of MET15 LOH events were linked to LOH events at the TRP1 marker distal to the rDNA array , but not the KANMX marker proximal to the array . This indicates that most MET15 LOH events originate within the rDNA array , with homozygosity extending ∼250 kbp from the rDNA array to the MET15 locus , and presumably the remaining ∼625 kbp to the telomere ( Figure 1D ) . This LOH pattern is indicative of mitotic recombination events between homologous chromosomes . The fraction of mitotic recombination events originating within the rDNA array was significantly higher than expected based on the sequence distance represented by the rDNA array ( based on a 95% confidence interval , binomial distribution ) , consistent with previous examinations of LOH at MET15 [11] , [17] . Thus , MET15 LOH events are primarily serving as a read out for an age-associated increase in mitotic recombination that affects the rDNA array . LOH events in S . cerevisiae generating half sectors can be further classified as either reciprocal ( LOH occurs in both cells ) or nonreciprocal ( one cell undergoes LOH while the other remains heterozygous ) ( see Figure 2A and reviewed in [5] ) . This classification has facilitated a better mechanistic understanding of LOH events . For instance , in a previous examination of mutations that increase LOH rates in young cells , we found that the ratio of reciprocal to non-reciprocal LOH events reflected the type of defect [17]: Mutations that affect specific DNA repair pathways bias LOH events away from the wild type ratio . By contrast , mutations that increase LOH rates but do not alter the ratio were consistent with a general increase in DNA damage that did not alter normal repair pathways . To examine reciprocal and non-reciprocal events in aging cells , we used the diploid MEP strain UCC5185 carrying ADE2 and MET15 in opposition at the MET15 locus . By examining the color of both half sectors , it is possible to distinguish reciprocal from non-reciprocal events ( Figure 2A ) . Consistent with the first MEP strain examined above , we observed a significant increase in total LOH events at MET15 in populations of aging UCC5185 cells ( Figure 2B ) . By 45 hours , when populations have passed their median viability , median LOH rates were increased 2-fold compared to the rate observed in the young population ( Figure 2B ) ( p<0 . 0001 , Fisher's exact test based on contingency tables ) . Between 70 and 95 hours , when population viability had fallen below ∼10% , median LOH rates were increased by four to eight-fold ( p<0 . 0001 , Fisher's exact test based on contingency tables ) . We found both reciprocal and non-reciprocal events increased with similar kinetics in aging populations ( Figure 2C ) , indicating that cells maintain a stable ratio of these events throughout the aging process . This result suggests that aging cells may experience an increased frequency of DNA damage within the rDNA array , rather than an age-associated defect in a particular repair pathway . In haploid cells , double stranded breaks ( DSBs ) within the rDNA are often initiated by the DNA replication fork-blocking ( RFB ) activity of Fob1 [22]–[23] . Replication forks traveling opposite to the direction of transcription of the 35S rRNA are blocked by the specific interaction of Fob1 with sequences within the non-transcribed region 1 ( NTS1 ) [24] . This source of DNA replication stress can cause fork collapse to generate DSBs , which can be repaired by homologous recombination to yield a variety of products [25]–[26] . We tested whether Fob1 was required for age-associated LOH at the MET15 locus . In young cells , fob1Δ led to a reduction in the MET15 LOH rate of ∼2-fold ( Figure 3 ) , but this rate was still significantly higher than the LOH rate on chromosome IV , suggesting that other mechanisms must contribute to the hotspot activity affecting MET15 . When we measured MET15 LOH in aging fob1Δ cells we found the age-associated increase in rate of LOH was completely suppressed ( Figure 3 ) . This result indicated that Fob1 activity is required for this aging phenotype and suggests that Fob1-mediated DSBs are a critical intermediate for these LOH events . We offer two simple models to explain the requirement for FOB1 in the age-associated increase in LOH at MET15 . Fob1 activity may be required to generate an aging factor or process that leads to increased LOH . Alternatively , an aging process may occur independent of Fob1 , while Fob1 remains necessary for this process to manifest as LOH events in old cells . To distinguish between these modes of action , we generated a diploid MEP strain in which FOB1 is expressed from a tetracycline-repressible promoter [27] . When the TET-FOB1 strain is aged over a 70-hour time course in the absence of repressor , cells exhibit a normal increase in LOH ( Figure 4; Fob1 ON ) . In contrast , when the TET-FOB1 strain is aged over a 70-hour time course in the presence of doxycycline ( a tetracycline analog ) , it behaves as a fob1 null allele and age-associated LOH is suppressed ( Figure 4; Fob1 Off ) . Critically , after the TET-FOB1 strain is aged over a 65-hour time course in the presence of doxycycline , removal of the repressor results in a rapid ∼5-fold increase in LOH levels within 5 hours ( Figure 4; Fob1 Off → On ) . This result suggests that the process responsible for increased LOH rates in aging cells occurs independently of Fob1 , but that Fob1 activity is required in old cells for this aging process to generate LOH events . We hypothesize that the aging process may involve attenuation of an activity that normally keeps the rate of homologous recombination relatively low within the rDNA array . By extension , elimination ( e . g . by gene mutation ) of such an activity in young cells should phenocopy the age-associated increase in LOH . Our characterization of the age-associated LOH phenotype has delineated two criteria such candidate activities must possess: Deletion of a candidate activity must show a FOB1-dependent increase in LOH at the MET15 locus , and the ratio of reciprocal-to-nonreciprocal LOH events should be the same as in wild type cells . One mechanism by which LOH could increase in old cells is through an age-associated defect in sister chromatid cohesion , which normally restricts template selection during DSB repair to favor sister chromatids [28]–[31] . Such a defect would allow more recombination to occur between homologs during repair of Fob1-mediated DSBs , thus giving rise to increased LOH . The effect of disrupting sister chromatid cohesion was examined by deleting the genes encoding the cohibin complex proteins Csm1 and Lrs4 . The cohibins physically interact with both Fob1 and cohesins , anchoring cohesin rings within the rDNA array and inhibiting both unequal sister chromatid exchange ( USCE ) in haploid cells and homologous recombination in diploids [17] , [31] . Consistent with these reports , we found that csm1Δ or lrs4Δ alleles resulted in high rates of MET15 LOH in young cells ( Figure 5 ) . But importantly , when csm1Δ or lrs4Δ were combined with fob1Δ , rDNA recombination rates in young cells remained high ( Figure 5 ) . Thus , loss of cohesion leads to Fob1-independent rDNA recombination , which strongly suggests that this is unlikely to be the mechanism underlying the increased rDNA recombination observed in aging cells . The protein deacetylase Sir2 has been implicated in the yeast aging process and life span determination [32] . Furthermore , loss of Sir2 function leads to high rates of USCE in haploid cells , resulting in dramatic copy-number instability within the rDNA array [33]–[35] , and increased rates of homologous recombination between rDNA arrays in diploid cells [11] . Therefore , we examined the potential role of Sir2 in age-associated LOH in detail . It was recently shown that Sir2 protein levels decline in aging haploid cells [36] . To determine if Sir2 levels are reduced in aging diploid cells , we purified aged mother cells from liquid MEP cultures and prepared total protein extracts . When the MEP is activated , the time a mother cell spends in culture corresponds to her replicative lifetime [16] . Hence we can age-match different strains by aging these MEP cultures for equivalent periods of time ( the strains analyzed here divide at the same approximate rate; data not shown ) . To confirm age matching between strains , purified 26-hour populations were stained with calcofluor white to count bud scars ( see Figure 6 legend ) . Western blotting with an anti-Sir2 antibody confirmed that Sir2 protein levels were dramatically reduced in aging diploid cells when normalized to total protein ( Figure 6A ) . Sir2 levels in aging populations had fallen 10-fold by 26 hours , and were further reduced by 50 hours . This same decline was also observed with a polyclonal antiserum raised to a different region of the Sir2 protein ( data not show ) . However , this behavior was not a universal feature of proteins in aging cells: Western blotting against the vacuolar protein Vma2 ( Figure 6A ) and the kinase Pkc1 ( data not shown ) both showed no decline in aging cells . Therefore , Sir2 protein levels specifically decline in aging cells , however this decline precedes any significant change in LOH rates by 20 hours , or approximately 12 generations . The observation that Sir2 protein levels decline precipitously 20 hours before rDNA recombination rates increase significantly led us to further examine the relationship between Sir2 depletion and age-associated LOH . We tested the effect of SIR2 gene copy number on rDNA recombination using diploid strains homozygous for a tandem repeat of SIR2 under control of its native promoter ( SIR2OE ) . Previous studies indicated that this modest level of over-expression is sufficient to increase silencing of RNA Polymerase II-dependent transcription within the rDNA array [34] , [37] and extend RLS [32] , while higher expression of Sir2 is toxic [38] . Sir2 over-expression in young ( log-phase ) diploid cells was verified by Western blot ( Figure 6A ) . The decline in Sir2 protein levels was partially suppressed in the SIR2OE strain: Sir2 levels in SIR2OE populations were elevated at 26 hours compared to wild type , but declined precipitously to wild type levels by 50 hours ( Figure 6A ) . Next , we examined LOH rates at MET15 in the SIR2OE strain to determine if increased Sir2 levels suppressed the age-associated change in LOH rates . Although MET15 LOH rates in young SIR2OE cells were significantly lower than the wild type strain ( p<0 . 0001 , Fisher's exact test based on contingency tables ) , we observed no difference in the timing or magnitude of the age-associated increase in MET15 LOH rates in SIR2OE cells compared to wild type ( Figure 6B ) . This result offers further support that Sir2 protein levels are not the determining factor regulating rDNA recombination in aging cells . To confirm this result , we attempted to stabilize Sir2 protein levels by a second method . In a previous study , Dang , et al . found that deletion of SAS2 , the gene encoding the catalytic histone acetyltransferase subunit of the SAS complex that acts antagonistically to Sir2 for modification of histone H4 at lysine 16 , stabilized Sir2 levels in aging cells [36] . When we constructed a sas2 deletion strain in the MEP background and examined Sir2 protein levels , we saw no evidence of Sir2 stabilization in this mutant background ( Figure 6C ) . Finally , we asked whether deletion of fob1 , which does suppress age-associated rDNA recombination , might also stabilize Sir2 protein levels in aging cells . Again , we observed no stabilization of Sir2 protein levels ( Figure 6C ) . Thus , we found no evidence of a correlation between Sir2 protein levels and LOH rates in aging populations . Following the same logic behind our examination of cohesion , we characterized the pattern of LOH events observed in young cells when sir2 was deleted . We found that elevated MET15 LOH rates in young sir2Δ cells were dependent on FOB1 ( Figure 7A ) . In this regard it shows similarity to the age-associated LOH phenotype . However , the ratios of reciprocal to non-reciprocal events in sir2Δ cells was significantly shifted to favor non-reciprocal events ( Figure 7B ) , indicating that loss of Sir2 function does not accurately phenocopy the age-associated LOH phenotype . Lastly , we examined LOH rates in aging sir2Δ cells to determine if there is an age-associated increase in the absence of Sir2 . Consistent with our earlier results [11] , there was a very high level of LOH in young cells . Nevertheless , after 70 hours of replicative aging , LOH rates at MET15 increased significantly ( p = 0 . 0003 , Fisher's exact test based on contingency tables; Figure 7C ) . It is worth noting that sir2Δ cells have a short RLS [32]; thus at 70 hours only the longest-lived 1% of the total population are represented . When taken together , these data - the lack of correlation between Sir2 protein levels and rDNA recombination rates in aging cells , the difference in reciprocal/non-reciprocal LOH events between aging wild type and sir2Δ cells , and the Sir2-independent increase in LOH in aging cells - indicate that declining Sir2 levels are insufficient to explain the age-associated rDNA recombination phenotype . Intrachromosomal recombination between the rDNA repeats can produce extrachromosomal rDNA circles ( ERCs ) that have been postulated to induce senescence in aging mothers by an undefined mechanism [39] . Each repeat contains an origin of DNA replication , allowing ERCs to replicate independently , but lacks a centromeric sequence to ensure equal partitioning of ERCs between mother and daughter cells during mitosis [39]–[40] . Consequently , ERCs preferentially accumulate within aging mother cells . It has been suggested recently that ERCs can influence rDNA array stability in haploid cells [41] . Given these findings , we explored whether ERC levels play a role in age-associated LOH in diploid cells . In order to examine this issue , we found it necessary to further develop methods employed to examine ERC levels with age . Previous analyses of ERC levels have been technically limited to the examination of relatively young cells ( ∼10 generations ) [42] , whereas we observe age-associated increases in LOH after the median life span of our diploid strains ( 36 generations for UCC5185 ) [16] . To better understand the kinetics of ERC accumulation , we purified aged mother cells from MEP cultures and quantified ERC levels by Southern blotting . To improve ERC quantitation , we removed the linear rDNA array by digestion with RecBCD exonuclease , a highly processive exonuclease that does not degrade intact or nicked circular DNA [43] ( Figure S2 ) . Monomeric , dimeric , and multimeric ERC species were identified based on their migration rates [44] ( Figure 8A ) and quantified by densitometry ( Figure 8B ) . First , we compared ERC levels in wild type and fob1Δ cells . A modest increase in all three species of ERCs was evident in wild type cells after aging replicatively for 26 hours , a relatively young age when the population maintains ∼90% viability ( Figure 1A and [16] ) . As the population approaches its median RLS by 50 hours , ERC levels continued to increase dramatically . Similar to previous reports , in relatively young cells , ERC levels in a fob1Δ diploid were low compared to wild type cells ( Figure 8B and [42] ) , but their accumulation was not completely prevented . After 26 hours of replicative aging , total ERC levels in the fob1Δ strain had increased ∼5-fold above young wild type levels , but were still reduced ∼50% compared to the age-matched wild type population . Interestingly , as the fob1Δ population continued to age , ERC accumulation continued: Total ERC levels in fob1Δ cells were still ∼50% lower than wild type cells after aging replicatively for 50 hours , but by this age ERC levels in the fob1Δ strain had increased ∼17-fold above young wild type levels . These results indicate that ERC accumulation is an aging process that is not strictly dependent on Fob1 activity . The delayed accumulation of ERCs could reflect a reduction in the rate of formation of ERCs from the rDNA array in the fob1Δ strain , while amplification of ERCs by replication is unaffected [39] . While the fob1Δ strain permitted us to determine the utility of our ERC measurements with age , it was not useful for examining the effect of ERC accumulation on LOH because , as shown above , Fob1 is required for the age-associated LOH events . For this , we examined a genetic background that could reduce ERC accumulation independently of Fob1 . Deletion of BUD6 has been reported to disrupt the asymmetric segregation of ERCs , resulting in extended life span and a prediction of reduced ERC accumulation in mother cells [45] . Consistent with this prediction , we found that deletion of bud6Δ reduced the level of ERCs throughout the life span of a population . The bud6Δ cells showed a 50% reduction in ERCs after 50 hours of replicative aging: A level that was achieved in wild type populations aged for only 26 hours ( Figure 8B ) . Remarkably , we saw no change in the age-associated increase in LOH rate at MET15 in the bud6Δ strain ( Figure 8C ) . This indicated that in the presence of Fob1 activity , a 50% reduction in ERC accumulation was not sufficient to suppress age-associated LOH .
Previously we discovered an age-associated , genome-wide increase in LOH in yeast that results from the loss of mtDNA and respiratory function in the progeny of aging cells [11]–[12] . Here we used our recently developed MEP [16] to examine LOH in cells that retain respiratory function throughout their life span . Although the MEP strains are in the same S288C strain background as strains used in previous studies , they display a greater capacity to maintain mitochondrial function both in logarithmically growing cultures and during aging [11] , [46] . This greater stability of mitochondrial function in the MEP strain is similar to that found for most natural isolates of S . cerevisiae [14] . The faithful maintenance of functional mitochondria depends on over 100 genetic loci [15] and we have not determined the precise genetic basis for the phenotypic differences we observe . Using the MEP , we have discovered a distinct age-associated LOH phenotype in cells that retain respiration competence . This new LOH phenotype is distinguished by an apparent specificity for the rDNA array and dependence on the replication fork-block protein Fob1 . As cells pass their median life span , they experience a significant increase in homologous recombination within the rDNA array which leads to LOH along a ∼875 kbp span from the rDNA to the telomere of the right arm of chromosome XII . These genomic alterations are mechanistically equivalent to events that generate partial uniparental disomy in mammalian cells , which has recently been found to occur at high frequency in many human cancers [6] , [47]–[48] . We observed a significant increase in both reciprocal and non-reciprocal recombination events , which likely report two different routes of DSB repair . Reciprocal LOH events likely result from strand invasion leading to Holliday junction formation and resolution with crossing over [49] , while the non-reciprocal events we observed are consistent with break-induced replication ( BIR ) initiated within the rDNA and propagated to the telomere of chromosome XII [17] , [49] . These recombination events also differ from the age-associated genomic instability described in cells that have lost respiration competence [11] , where LOH events were almost uniformly non-reciprocal . The phenotypic differences between the earlier studies and those reported here support the conclusion that a different mechanism leads to age-associated LOH events in cells that maintain respiration competence . Because most eukaryotic cells cannot tolerate the loss of mtDNA , it is likely that the findings we report here about genomic instability may be relevant to the aging process in other organisms . Taking advantage of the requirement for Fob1 , we investigated potential mechanisms leading to the age-associated LOH phenotype . Disruption of cohesion , via the deletion of the cohibins LRS4 and CSM1 , was eliminated as a potential mechanism because it leads to Fob1-independent LOH events . While SIR2 deletion generated Fob1-dependent LOH events in young cells , the LOH events observed in this strain were significantly biased towards a non-reciprocal pathway compared to the LOH events generated in young or old wild type cells . Additionally , when sir2Δ cells were aged they also displayed an age-associated increase in LOH in the longest-lived fraction of the population . Both the difference in repair bias and presence of an age-associated increase in LOH independent of SIR2 suggests that loss of Sir2 function in aging cells is likely not the driver of age-associated LOH . Aging is typically accompanied by an exponentially increasing hazard rate of death – i . e . the probability that an individual within a population will die within a given time interval . Previous biochemical analyses of aging cells focused on cells aged for only ∼10 generations due to technical limitations [50] . At this age , a diploid cell population in the common laboratory S288C strain is still relatively young: It retains >95% viability and has yet to experience the dramatic increase in hazard rate of death that affects cells near their median RLS [16] , [51] . Using the MEP , we developed an effective purification method to isolate age-matched populations near their median RLS which allowed us to make new observations that change our understanding of age-associated processes . Earlier reports concluded that fob1Δ effectively suppressed the accumulation of ERCs [32] , [42] . By examining populations at their median life span we found that ERC levels in fob1Δ cells actually increase significantly as cells approach their median RLS , although their accumulation was reduced/delayed compared to wild type cells . The extension of our ability to observe age-associated genetic and biochemical changes with the MEP allows us to begin to develop an understanding of the order of events that affect an aging population . It was recently reported that Sir2 protein levels decline in aging haploid cells after only 7-9 generations , a relatively early point in their life span [36] . Our findings are consistent with this earlier work: Sir2 protein levels in aging diploid cells show a precipitous decline by 24 hours . However , this occurs at a time in which LOH rates have not yet increased and ∼95% of cells remain viable . Furthermore , when Sir2 was over expressed , no suppression of age-associated LOH rates was observed . Taking all these results together , we interpret these data to indicate that declining Sir2 protein levels do not correlate with the increase in age-associated LOH and offer further evidence that Sir2 is unlikely to be responsible for the age-associated increase in LOH . One potential caveat to our interpretation is that if there is significant cell-to-cell heterogeneity in levels of Sir2 over-expression , then LOH may occur in a subpopulation of cells with lower Sir2 levels . Why might reduced Sir2 protein levels in aging cells fail to alter rDNA recombination rates in a fashion similar to young cells ? One possible explanation is a phenotypic lag of the increased rDNA recombination that lasts for >10 generations after the reduction of Sir2 protein . For instance , if Sir2 is completely absent in most cells , then one of it substrates ( e . g . K16 of histone H4 ) may not immediately become acetylated [52] . In fact , newly synthesized histones start out in an unacetylated state , and how a heritable switch between acetylated and unacetylated states of histones occurs remains mysterious [53] . Another possibility is that a very low level of Sir2 protein may be sufficient to suppress rDNA recombination . Relevant to this idea , different regions of silent chromatin can compete for recruitment of Sir2 [37] , [54] . If the rDNA array is dominant in such a competition , the effect of declining Sir2 levels could be masked . If ERC accumulation is the aging process that leads to age-associated LOH , their influence on LOH must be Fob1-dependent . Indeed , it has previously been shown that plasmids containing the RFB sites from an rDNA repeat can integrate into the chromosomal rDNA array in a Fob1-dependent manner [55] . Thus , ERCs may be capable of initiating recombination events in a Fob1-dependent manner that somehow increases the frequency of homologous recombination in diploid cells . Alternatively , accumulating ERCs could simply titrate away factors that affect the frequency or stability of stalled DNA replication forks and thus increase their conversion to DSBs within the rDNA array [26] . In order to determine whether ERCs play a role in age-associated LOH , it is necessary to eliminate ERC formation and/or accumulation in aging cells . Using the bud6Δ strain we reduced ERC levels at a median RLS by approximately 50% , down to a level that was equivalent to a younger wild type population ( 26 hours of replicative aging; Figure 8B ) . It is significant that at this younger age , wild type cells show no increase in LOH rate at the MET15 locus . Despite this reduction in ERC levels , the bud6Δ strain showed no suppression of age-associated LOH with age . While we interpret this as evidence that ERC accumulation does not drive age-associated LOH at MET15 , it remains formally possible that this reduced level of ERCs is still above a threshold required to increase LOH , or that a subpopulation of cells contain higher ERC levels that drive age-associated LOH events . A clearer understanding of the contribution of ERCs to age-associated LOH will require more refined control of ERC initiation and accumulation . Previously , life span extension by fob1Δ was interpreted as a result of reduced ERC accumulation to a degree at which RLS becomes limited by an alternative mechanism [32] , [42] . However , we found that ERCs accumulate to high levels even in fob1Δ cells ( Figure 8 ) . This result could be interpreted in two ways that support opposing models for how life span is limited in fob1Δ cells . First , the delayed accumulation of ERCs early in life could extend , but still limit , RLS in fob1Δ cells . Alternatively , the life span-limiting function of ERCs may be Fob1-dependent , suggesting that RLS in fob1Δ cells is limited by an alternative mechanism . Recently it has been argued that rDNA stability , rather than ERC accumulation , limits life span [41] . However , because life span potential is ‘re-set’ in daughter cells [56] , we cannot conclude that an LOH event ( which would be heritable ) directly limits RLS: Instead , a reversible aging process that impacts rDNA stability , and can generate LOH events at some frequency , may limit RLS . Curiously , we find that the highest rates of age-associated LOH are observed at the latest time points , where the population consists of the longest-lived survivors ( <10% of the population remains viable ) , and when the hazard rate of death has declined from its peak near the median life span [16] . This lack of correlation between LOH rate and hazard rate of death suggests that rDNA recombination is not limiting for life span of most cells . Nevertheless , the longest-lived individuals in the population no longer experience the same hazard rate of death , and thus their life span may be limited by a different mechanism ( e . g . rDNA recombination ) than the average population . If pathways that have been previously identified to modulate rDNA recombination do not adequately account for the age-associated increase in LOH rates , can this phenotype be a response to a change in a more general biological process that is most readily manifested at the rDNA ? Since DSBs within the rDNA array normally arise through the interaction between DNA replication forks and Fob1 [57] , we speculate that DSB rates in the rDNA are modulated by DNA replication stress . This model is supported by several lines of evidence: A hypomorphic allele of the essential DNA replication helicase encoded by DNA2 results in an increased frequency of DSBs within the rDNA , which can be suppressed by deletion of FOB1 [57]–[58] . Similarly , mutations in DNA polymerase α and δ subunits can also lead to increased DSBs within the rDNA array [26] , [59] . Deletion of RRM3 , a helicase that functions to remove non-histone protein barriers from DNA , also affects rDNA recombination [60]–[61] . Further evidence comes from a screen for deletion mutants that increase LOH in young cells [17] , which classified mutations based on locus specificity , magnitude and ratio of reciprocal/non-reciprocal events . The group of deletion mutants that showed a bias toward increasing LOH at MET15 on chromosome XII ( and presumably originate in the rDNA ) included five genes implicated in the regulation of nucleotide pools , which can be a source of DNA replication stress [62] . While the effects of DNA replication stress on the rDNA may be dependent on Fob1 , the mode of action may not necessarily be specific to the rDNA . Ivessa , et al . identified genomic regions prone to DNA replication fork pausing in an rrm3Δ mutant which included centromeric regions , tRNAs and sub-telomeric sequences in addition to the rDNA array [9] . Similarly , RRM3 was originally identified in a genetic screen for mutations that induce gene duplication at the tandemly repeated CUP1 locus [60] , which suggests that other regions of the genome that combine replication fork blocks with repetitive sequence elements could also generate age-associated LOH events . While repetitive elements are uncommon in the S . cerevisiae genome , the ubiquitous nature of these features in mammalian genomes suggests great potential for age-associated genomic instability generated by a similar mechanism .
Genotypes of all yeast strains used in this study are provided in Table S1 , oligonucleotide sequences are provided in Table S2 , and plasmids are listed in Table S3 . The two-chromosome LOH reporter strain UCC8918 was generated by mating UCC8917 x UCC5181 . UCC8917 was derived from UCC5179 by one-step integration of a PCR fragment carrying ADE2 into an intergenic region of chromosome IV ( coordinates 1 , 515 , 634–1 , 515 , 738 ) , which was generated using oligos MarthaN/H2L and MarthaN/H2R with pRS402 [63] as a template . The diploid strain with multiple heterozygous markers on chromosome XII ( UCC8915 ) was generated by mating UCC8914 x UCC5179 . UCC8914 was derived from UCC8913 by one-step integration of a PCR fragment carrying TRP1 into an intergenic region distal to the right end of the rDNA array ( coordinates 486062-486189 ) , which was generated using oligos RDNRF and RDNRR with pRS304 [64] as a template . Integration was verified by PCR from genomic DNA using oligos RDNRconF and RDNRconR . UCC8913 was derived from UCC5181 by one-step integration of a PCR fragment carrying KANMX into an intergenic region proximal to the left end of the rDNA array ( coordinates 450191-450372 ) , which was generated using oligos RDNLF and RDNLR with pRS400 [63] as a template . Integration was verified by PCR from genomic DNA using oligos RDNLconF and RDNLconR . MEP deletion strains for bud6Δ , csm1Δ , lrs4Δ , sir2Δ and sas2Δ were generated by one-step gene disruption and verified by PCR using the following oligonucleotides and DNA templates: BUD6: deletion- Bud6delF/bud6delR , template- pRS306 , confirmation- Bud6delcheck/Bud6delchkDN . CSM1: deletion- CSM1A/CSM1D , template- UCC7629-1 genomic DNA , confirmation- CSM1A/CSM1B . LRS4: deletion- LRS4A/LRS4D , template- UCC7598-1 genomic DNA , confirmation- LRS4A/LRS4B . SIR2: deletion- SIR2KO1/SIR2KO2 , template- pRS400 [63] , confirmation- 5′SIR2/3′SIR2 . SAS2: deletion- SAS2KOF/SAS2KOR , template- pRS400 [63] , confirmation- SAS2conF/SAS2conR . The sir2Δ fob1Δ double mutant was constructed by transforming UCC8832 with plasmid pRS314-SIR2 [65] to complement the mating defect of sir2Δ , followed by mating to UCC524 and sporulation . Because both deletions are marked with KANMX , PCR was used to identify double mutant spore products to generate UCC8839 and UCC8840 . These haploids were mated to generate UCC8844 , which was subsequently cured of the pRS314-SIR2 plasmid . Standard mating and sporulation , followed by PCR to identify double mutant spore products , was used to generate csm1Δ fob1Δ and lrs4Δ fob1Δ double mutant strains . To generate SIR2OE strains , a genomic clone of SIR2 along with ∼500 bp of upstream sequence and ∼250 bp of downstream sequence was subcloned by PCR amplification with SIR2ecoriF and SIR2ecoriR primers , digested with EcoRI , and ligated into pRS306 cut with EcoRI to create the integration plasmid pRS306-SIR2 . The plasmid was cut with BglII and transformed into UCC5179 and UCC5181 to generate strains UCC8908 and UCC8909 . Correct integration was confirmed by PCR . These haploids were mated together to generate the diploid strain UCC8910 . The SIR2hemi diploid was generated by transforming UCC8832 with pRS314-SIR2 , mating to UCC5179 , and subsequently curing the diploid of the plasmid . The TET-OP2 promoter was inserted upstream of the endogenous FOB1 locus by one-step integration of a PCR fragment generated using primers FOB1tetF and FOB1tetR with pKAN-TETO2 as a template . The TETR'-SSN6 cassette was inserted by one-step integration into the met15Δ0 locus by digesting plasmid pLMI-tetR'S with PacI . The LEU2 marker was subsequently exchanged for ADE2 or MET15 using the primers LEU2swapF and LEU2swapR with the appropriate plasmid templates ( pRS402 and pRS401 , respectively [63] ) . The VP16 activation domain from the tTA activator was replaced with activation domain A of GCN4 by generating a GCN4A PCR product using GCN4salAF and GCN4ascAR primers with plasmid pGCN4 ( a gift from S . Hahn ) as a template . The PCR product was digested with SalI and AscI and subcloned into pUI-tTA-ADH1term-URA3 digested with the same enzymes to create pUI-tTA-GCN4A-ADH1term-URA3 . The tTA-GCN4A-ADH1term-URA3 cassette was integrated into a neutral site on Chromosome I ( coordinates 17030-17205 ) by one-step integration of a PCR product using the primers URA3-tTA-intCHRIF and URA3-tTA-intCHRIR . Subsequent strain construction to generate haploid and diploid MEP strains with the appropriate genotypes was performed by standard methods . Pedigree analysis was performed as previously described [11] . Briefly , daughter cells born to individual mothers were micromanipulated to new positions on a YEPD plate and allowed to form colonies . Colonies were scored for mitochondrial function by replica plating to YEP+glycerol . Diploid cells were grown to saturation overnight in YC media lacking adenine and methionine . Cells were used to inoculate YEPD cultures , which were grown to log phase by incubation with shaking , 30°C for 3 hours . Cells were counted and used to inoculate 25 ml YEPD +1 µM estradiol cultures at 2×104 cells/ml and incubated with shaking , 30°C for 95 hours . At indicated times , samples were harvested , washed , and plated to lead nitrate media . Sample volumes were adjusted appropriately to maintain a colony density of ∼500–1000 colonies/150 mm plate . Plates were incubated at 30°C for 3 days and colonies were counted using a Geldoc XR+ imaging system ( Biorad ) . Plates were further incubated at room temperature for 2–3 days for color development before scoring for half sectors . Reciprocal and non-reciprocal colonies were counted separately , and rates of total half sectors were calculated as ( 2*reciprocal + non-reciprocal/total colonies ) [17] . Lysates were prepared using NaOH lysis followed by TCA precipitation [66] . TCA pellets were resuspended in SUME buffer ( 1% SDS , 8 M Urea , 10 mM MOPS , pH 6 . 8 , 10 mM EDTA ) and total protein concentration was determined using the BCA protein assay ( Thermo Scientific ) . 10 µg total protein per lane was run on 10% tris-glycine polyacrylamide gels ( PAGEr gold , Lonza ) and transferred to Immobilon P membrane ( Millipore ) using a semi-dry transfer apparatus . Western blot analysis was performed by standard methods and developed with Supersignal West Pico ( Thermo Scientific ) . Antibodies: Goat α-Sir2 ( yN-19; Santa Cruz Biotechnology ) , mouse α-Sir2 ( sc-25753; Santa Cruz Biotechnology ) mouse α-Vma2 ( Invitrogen ) , mouse α-Pkc1 ( Invitrogen ) , and HRP-conjugated secondary antibodies ( Jackson Immunoresearch Laboratories ) . Quantitation of Western blots was performed by densitometry using ImageJ comparing exposures that fell within a linear range of detection . Log phase cultures of cells were harvested and labeled with NHC-Biotin as previously described [16] . Labeled populations were transferred to YEPD and incubated for 30°C , 2 hours to ensure that most labeled cells were mothers ( have completed at least one cell division ) . Cells were counted and used to inoculate 1 . 5 liter cultures of YEPD +1 µM estradiol +100 µg/ml ampicillin at a density of 2×104 cells/ml . Cultures were incubated at 30°C , shaking at 100 rpm for the indicated times before purification . Cells were harvested by centrifugation , resuspended at 6×108 cells/ml in RNAlater ( Ambion ) and fixed at room temperature , 45 minutes . Cells were harvested by centrifugation and resuspended at 2×108 cells/ml in PBS +2 mM EDTA and incubated with 1/20 volume streptavidin beads ( Miltenyi Biotec ) , 4°C , 30 minutes . Cells were harvested by centrifugation and resuspended in 4 ml 40 mM Tris HCl pH 7 . 4 and layered onto Percoll Plus gradients ( GE Healthcare ) . Gradients were spun at 4°C , 20 minutes at 2000 RPM in a GS-6R tabletop centrifuge ( Beckman ) . A brown , flocculent layer of cell debris was removed from the top of the gradients with a pipette , and the remainder of the gradient was pooled with 40 ml PBS +2 mM EDTA . Cells were harvested by centrifugation , resuspended in 45 mL PBS +2 mM EDTA and purified on an Automacs Pro separator system ( Miltenyi Biotek ) . Genomic DNA was isolated from purified aged populations by standard methods . 1 µg of genomic DNA was digested overnight with BamHI + RecBCD ( a gift from G . Smith , FHCRC ) and separated by gel electrophoresis ( 0 . 8% agarose , 2 V/cm for 36 hours ) . DNA was visualized by staining with ethidium bromide and transferred to nitrocellulose membranes by standard methods . Membranes were hybridized with a 32P labeled double-stranded probe specific to the rDNA generated with oligos RDN5S-2 and RDN25S-2 using plasmid pDL05 as a template , visualized on a Typhoon phosphorimager ( GE Health Sciences ) , and quantified using ImageJ . For normalization of DNA samples , 1 µg of genomic DNA was digested with BamHI , separated by gel electrophoresis ( 0 . 8% agarose , 10 V/cm , 3 hours ) , transferred and hybridized with a probe specific to NPR2 generated with oligos 5_NPR2 and 3_NPR2 . Because over expression of Fob1 also increases rDNA recombination rates [67] , we generated a weaker version of the tet-activator by replacing the VP16 activation domain with activation domain A of Gcn4 [27] , [68] . By carefully titrating Fob1 expression levels using a single copy of both TETOP2-Fob1 and the tTA-GCN4A activator in diploid cells , we could express FOB1 at normal levels in the absence of doxycycline , while effectively suppressing expression in the presence of 20 µg/ml doxycycline ( Figure 5A ) . Diploid cells were grown to saturation overnight in YC media lacking adenine and methionine . Cells were used to inoculate YEPD +/- 20 µg/ml doxycycline cultures , which were incubated with shaking , 30°C for 5 hours . Cells were counted and used to inoculate 40 ml YEPD +1 µM estradiol +/- 20 µg/ml doxycycline cultures at 2×104 cells/ml and incubated with shaking , 30°C for 95 hours . Samples were harvested and plated as described above for the LOH assay . To shift FOB1 expression from repression to activation , 15 ml of the + doxycycline culture was harvested at the 65 hour time point , washed , and resuspended in 15 ml fresh YEPD +1 µM estradiol .
|
There is a striking correlation between age and the onset of many diseases , such as cancer , suggesting that the aging process itself can contribute to their development . Cancer is a genetic disease caused by the accumulation of a series of deleterious somatic mutations leading to unchecked proliferation . In humans , it is well established that normal mutation rates are not sufficient to account for the sharp increase in cancer rates in aging populations , suggesting a change in mutation rate is a necessary component of cancer development . Here , we find that the aging process in the budding yeast Saccharomyces cerevisiae leads to an increased rate of homologous recombination within a repetitive DNA sequence element , the ribosomal rDNA array . While these mutational events are initiated primarily at this single locus , they are propagated to the end of the chromosome and thus affect hundreds of genes . These results suggest that the aging process itself could contribute to increasing mutation rates and perhaps to the onset of age-associated disease .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"dna",
"metabolism",
"cancer",
"genetics",
"genetic",
"mutation",
"chromosome",
"structure",
"and",
"function",
"developmental",
"biology",
"organism",
"development",
"mutation",
"types",
"dna",
"chromosome",
"biology",
"biology",
"molecular",
"biology",
"aging",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"dna",
"repair",
"dna",
"recombination",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] |
2011
|
Replicative Age Induces Mitotic Recombination in the Ribosomal RNA Gene Cluster of Saccharomyces cerevisiae
|
Phlebotomine sand flies are insects that are highly relevant in medicine , particularly as the sole proven vectors of leishmaniasis . Accurate identification of sand fly species is an essential prerequisite for eco-epidemiological studies aiming to better understand the disease . Traditional morphological identification is painstaking and time-consuming , and molecular methods for extensive screening remain expensive . Recent studies have shown that matrix-assisted laser desorption and ionization time-of-flight mass spectrometry ( MALDI-TOF MS ) is a promising tool for rapid and cost-effective identification of arthropod vectors , including sand flies . The aim of this study was to validate the use of MALDI-TOF MS for the identification of Northern Amazonian sand flies . We constituted a MALDI-TOF MS reference database comprising 29 species of sand flies that were field-collected in French Guiana , which are expected to cover many of the more common species of the Northern Amazonian region , including known vectors of leishmaniasis . Carrying out a blind test , all the sand flies tested ( n = 157 ) with a log ( score ) threshold greater than 1 . 7 were correctly identified at the species level . We confirmed that MALDI-TOF MS protein profiling is a useful tool for the study of sand flies , including neotropical species , known for their great diversity . An application that includes the spectra generated here will be available to the scientific community in the near future via an online platform .
Phlebotomine sand flies ( Diptera: Psychodidae: Phlebotominae ) are insects of great medical relevance because they are the most frequent vectors of leishmaniasis [1 , 2] and also transmit various other human pathogens including bacteria and viruses [3] . Leishmaniases are a range of diseases caused by flagellated protozoans of the genus Leishmania ( Kinetoplastida: Trypanosomatidae ) , transmitted through the bite of an infected female sand fly [2] . The epidemiology of leishmaniasis is complex , due to the wide diversity of Leishmania and the sand fly species involved . In the Americas , 56 sand fly species are known to be potential vectors of 15 Leishmania species [2] . In French Guiana , Leishmania ( Viannia ) guyanensis is the most prevalent Leishmania species in humans and is mainly responsible for localized cutaneous leishmaniasis [4] . Other Leishmania species such as L . braziliensis can be more clinically debilitating , since they can cause mucocutaneous ( nose , mouth , and throat commitment ) or diffuse cutaneous leishmaniasis , requiring specific medical management [4 , 5] . On the Guiana Shield , the main vector of L . guyanensis is Nyssomyia umbratilis [6 , 7] . However , sand fly species transmitting medically important Leishmania species are partially identified or not yet identified , such as those related to the L . braziliensis local transmission cycle [2 , 7] . Additionally , vector ecology can evolve with environmental changes such as deforestation and urbanization [1 , 2 , 7] . Human activities in deforested areas may result in epidemics , as seen by the reported outbreaks of cutaneous leishmaniasis in French Guiana [5] and Argentina [8] , for instance . Therefore , the identification of sand fly species associated with transmission hotspots together with a better description of sand fly communities are essential to improving the understanding of leishmaniasis epidemiology [1 , 2 , 8] . To date , sand fly ecology studies have been limited due to the complexity of species identification , which requires the meticulous and time-consuming labor of entomological taxonomy experts [9] . Molecular identification has been proposed as an alternative method and molecular reference databases have been made available [10 , 11] . However , cost analysis for huge series of samples shows that extensive screening remains expensive . The recent advent of protein profiling using matrix-assisted laser desorption ionization time-of-flight mass spectrometry ( MALDI-TOF MS ) is about to revolutionize medical entomology [12] . Several studies have shown that MALDI-TOF MS is suitable for sand fly identification , and in-house databases have been constructed [13–17] . Only four species of sand flies from South America have been tested to date and implemented in an in-house MALDI-TOF MS database [13] . A disadvantage of in-house databases is the restricted access to local utilization . As a solution , the implementation of a centralized online platform has been suggested for sand fly identification [12 , 13] . The aim of the study was to create a MALDI-TOF MS reference database of French Guiana sand flies . Sand flies were captured from different geographic locations in French Guiana , an Amazonian territory along the Northern South American Atlantic coast . A mass spectra library ( MSL ) was implemented , based on molecular identification of field-collected sand flies . The reliability of the MSL was then evaluated using a blind test . This MSL will be included in an online platform dedicated to phlebotomine sand fly identification that is currently being developed at Sorbonne University .
To constitute and validate a MSL , collected sand flies were identified by DNA sequencing and divided into two different panels , a construction panel and a validation panel ( Table 1 and Fig 1 ) . The separation of the sampling into construction and validation panels was sequential . The first 206 sand flies analysed were attributed to the construction panel because it gave a sufficient number of species ( n = 20 ) and individuals by species ( from one to ten ) to build a MSL . The remaining 199 specimens , different from the samples of the construction panel , which had not been analysed before , were attributed to the validation panel . The construction panel comprised 260 sand flies collected from four study sites located in a pristine forest ( at the Nouragues reserve ) , in secondary forest ( Saint-Georges and Regina ) and in logging forest ( Counami ) . The validation panel comprised 199 sand flies , including 93 individuals that were collected at the same study sites and periods as those of the construction panel and 106 that were collected at four additional study sites and study periods . Additional sites were three peri-urban sites of secondary and edge forest ( around Cayenne ) and one pristine forest site ( along the Approuague River in Saut Grand Machicou ) . Between February 2017 and July 2017 , male and female sand flies were collected from different types of forested habitats in French Guiana ( Table 1 , Fig 2 ) . Sand flies were captured using Centers for Disease Control ( CDC ) Miniature light traps with Incandescent light ( [John W . Hock Company , Gainesville , FL , USA , CDC] ) , set between 6 pm and 6 am . Back at the laboratory , the sand flies were rapidly killed by freezing at −20°C and dissected immediately thereafter into four parts ( head , thorax with legs and wings , abdomen and genitalia ) , at room temperature . After dissection , the thoraxes with legs and wings were stored dry-frozen at −80°C , and the other body parts were stored in 70% ethanol at −80°C . Thoraxes with legs and wings were put aside for MALDI-TOF MS analysis and abdomens were put aside for molecular identification . Head and genitalia were kept for morphology in a future study ( not analysed at the time of this study ) , in case of discordant or uninterpretable identification results . Before analysis , the period of dry-frozen storage at −80°C varied between 10 days and 7 months , with a mean time of 4 months . Thoraxes with wings and legs were rinsed in ethanol 70% for 10 min in a 1 . 5-mL microcentrifuge tube . Tubes were centrifuged at ( 13 , 000 rpm , 10 min ) and supernatant was discarded . After a second centrifugation ( 13 , 000 rpm , 2 min ) , the remaining ethanol solution was then eliminated using a micropipette and left to evaporate . Proteic extraction consisted in adding 10 μL of 70% formic acid . After a manual homogenization with a micropipette , the homogenate was incubated for 5 min . Then 10 μL of 100% acetonitrile was added and left to incubate for 5 min . The homogenate was centrifuged ( 13 , 000 rpm , 2 min ) and 1 μL of the supernatant of each sample containing the protein extract was deposited onto a steel target plate ( Bruker Daltonics , Wissembourg , France ) . Once dried , the deposits were covered with a 1-μL alpha-cyano-4-hydroxycinnamic acid ( HCCA ) matrix prepared in 50% acetonitrile and 2 . 5% trifluoroacetic acid . To ensure the reproducibility of the results , a total of ten replicates were spotted for each isolate to be included in the MSL and a total of four replicates for each isolate of the panel to be tested . Mass spectra were acquired with a Microflex LT ( Bruker France SAS ) using the default acquisition parameters . The spectra were acquired in linear mode in the ion-positive mode at a laser frequency of 60 Hz and mass range of 2–20 kDa . The data was automatically acquired using AutoXecute in FlexControl v3 . 4 software ( Bruker France SAS ) , and exported into Maldi Biotyper v4 . 1 software for data processing with the default parameters and spectra analysis . All specimens from the validation panel , with a valid MALDI-TOF MS-based identification and mass spectrum quality , were secondarily implemented in the reference database . The species of the blind test that were not previously represented in the MSL or that were represented at a number of specimens less than 10 , were spotted in ten replicates until reaching ten references per species , with the same method as applied for the MSL construction .
The intra-clade genetic distance calculated on the 208 reference sequences , before and after the addition of our sequences to the analysis , was less than 3% for all the species . As expected , because of insufficient resolution , the 16S rDNA sequencing identification failed to discriminate morphologically closely related species Trichopygomyia trichopyga / Trichopygomyia depaquiti as well as three species of the genus Nyssomyia: N . umbratilis , N . yuilli and N . antunesi ( S2 Table ) . Ten individuals did not cluster with any species in the available reference sequences . MSL was composed of 20 species of field-collected sand flies . Between one and ten specimens of each MSL species were included , corresponding to 89 specimens . The validation panel was composed of 199 specimens , including from one to 45 specimens of 24 different sand fly species . The details of the sand fly species in each panel are found in Table 2 . The spectra had good resolution and intensity , with a large mass/charge interval , ranging from about 2 to 10 kDa ( Fig 3A , 3B and 3C ) . The spectra were highly homogenous and reproducible when obtained from different protein extract deposits of a single specimen ( Fig 3A ) . Variability of mass spectraprotein profiles was observed between different specimens of a single species , including when comparing spectra obtained from specimens of the same sex ( Fig 3B ) . When comparing spectra within a complex of species or between different species , the heterogeneity of mass spectraprotein profiles was observed ( Fig 3C ) . In the heat map grid of the CCI matrix values ( Fig 4 ) , the coloured squares of the central diagonal reflected the degree of reproducibility of each specimen’s mass spectra when compared to itself . Hot colours reflected high reproducibility of each specimen’s mass spectra . Around the central diagonal , spectra from various specimens of the same species were compared; hot colours showed a high level of intraspecific reproducibility of mass spectra ( diagonally ) , distributed in a cluster of species ( square ) . A mosaic of hot colours inside a cluster of identical species was indicative of heterogeneity of mass spectraprotein profiles between specimens and reflected intraspecific diversity . Outside of the diagonal , when the spectra of different species were compared , colder colours revealed lower CCI values with very little between-species MSP correlation , confirming the high intra-species specificity of the mass spectra . The matrix of CCI values of sand flies MSL is available in the supplementary data ( S3 Table ) . Cluster analysis of the dendrogram ( Fig 5 ) showed that each specimen belonging to the same species , either males and females , collected from various sites in French Guiana , clustered on the same branch . This result attests to the intra-species specificity of MALDI-TOF MS sand fly protein profiles and of the consistency with molecular identification . In concordance with molecular results , T . trichopyga and T . depaquiti were grouped together in a unique cluster of mass spectra . For the Nyssomyia genus , specimens were separated on two different branches of the dendrogram , whereas three were clustered in a monophyletic group of the maximum likelihood tree by molecular analysis of the 16S rDNA ( S1 Fig ) . According to the distribution of LS values , the interpretable identification result was defined as the best match of four spots with a LS value ≥1 . 7 ( Fig 6 ) . Of all the sand flies tested by MALDI-TOF MS during the blind test , 79% ( 157/199 ) gave interpretable MALDI-TOF MS-based identification results . A total of 37 samples corresponded to species molecularly identified that were missing in the MSL . When a corresponding reference spectrum was available in the MSL , 97% ( 157/162 ) of the MALDI-TOF MS-based identification results were interpretable . For specimens that did not have a corresponding reference spectrum in the MSL , 100% ( 37/37 ) of the identifications were not interpretable , because the best match of the four spots did not reach the threshold of 1 . 7 ( mean LS value = 1 . 36±0 . 095 ) . With the threshold value ≥1 . 7 , no misidentification was observed . Of the 157 sand flies with interpretable identification results , 100% were correctly classified at the species level with the best match LS value ranging from 1 . 7 to 2 . 6 ( mean LS value = 2 . 23±0 . 19 ) . Overall sensitivity was 79% when considering all the sand flies tested and 97% when considering only species with a corresponding reference in the MSL . Specificity was 100% . Five sand flies tested with a corresponding reference in the MSL had a LS value <1 . 7 and could not be identified . Three of them , a P . hirsutus ( LS value = 1 . 5 ) , a T . trichopyga/T . depaquiti ( LS value = 1 . 6 ) and a T . ininii ( LS value = 1 . 65 ) had a MALDI-TOF MS identification result concordant with molecular identification . Two of them , one P . amazonensis ( LS value = 1 . 63 ) and one E . infraspinosa ( LS value = 1 . 2 ) had a MALDI-TOF MS identification discordant with the molecular identification ( compared with P . claustrei and Nyssomyia sp . , respectively ) . This imply that a threshold lowered below 1 . 7 would have decreased specificity and increased the risk in giving wrong identification result . Complete dataset of sand flies identification results obtained by DNA sequencing and by MALDI-TOF MS is available in supplementary data ( S2 Table ) . The MSL was implemented with the mass spectra of nine additional sand fly species . Mass spectra of additional specimens of species already present in the MSL were also implemented to increase the diversity of mass spectra . The resulting reference database was made up of 282 specimens and 29 sand fly species ( Table 2 ) .
We confirm that MALDI-TOF MS protein profiling is well adapted to the identification of sand fly species , including in neotropical areas , known for its great diversity of sand fly species . MALDI-TOF MS can be a useful tool for rapid , inexpensive and accurate identification of sand flies but , like molecular methods , better accessibility to reference libraries for the scientific community would extend its utility . In the near future , this Northern Amazonian sand fly spectral database will be included in an online platform dedicated to phlebotomine sand fly identification , as already applied with success for identification of fungi and Leishmania of medical interest [26 , 37] . Recent studies have shown that MALDI-TOF MS was also accurate for the detection of Rickettsia spp . [41] and Borrelia crocidurae [42] in ticks , Plasmodium spp . in Anopheles mosquitoes [43] and Bartonella spp . in fleas [44] , by generating distinct mass spectra protein profiles between infected and uninfected arthropods . The possibility of identifying sand flies to the species level as well as the infection status by Leishmania parasites using MALDI-TOF MS would offer a significant opportunity for sand fly eco-epidemiology studies .
|
Phlebotomine sand flies are small insects , mostly known for their role in the transmission of Leishmania parasites to humans and other mammals . In French Guiana , the main clinical form of the disease manifests as cutaneous lesions also called American cutaneous leishmaniasis . The transmission of Leishmania from wild mammals to humans depends on the species of sand fly involved in the transmission . To better understand the mechanism of disease transmission , it is essential to accurately identify sand flies , including both vector and non-vector species . Until now , sand flies have mainly been identified using morphological and molecular methods . Recent studies have shown that a new tool based on protein profiling compiled in a library of spectra may be useful for the identification of arthropod vectors . This tool has the advantage of being less time-consuming , less expensive and does not require technical skills . The aim of this study was to assess the usefulness and accuracy of this new tool in identifying Northern Amazonian sand flies .
|
[
"Abstract",
"Introduction",
"Material",
"and",
"methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"tropical",
"diseases",
"sand",
"flies",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"organisms",
"protozoans",
"leishmania",
"neglected",
"tropical",
"diseases",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"infectious",
"diseases",
"zoonoses",
"mass",
"spectra",
"bioinformatics",
"ecosystems",
"biological",
"databases",
"chemistry",
"mass",
"spectrometry",
"protozoan",
"infections",
"disease",
"vectors",
"matrix-assisted",
"laser",
"desorption",
"ionization",
"time-of-flight",
"mass",
"spectrometry",
"analytical",
"chemistry",
"dna",
"sequence",
"analysis",
"sequence",
"databases",
"eukaryota",
"ecology",
"forests",
"database",
"and",
"informatics",
"methods",
"leishmaniasis",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"physical",
"sciences",
"spectrum",
"analysis",
"techniques",
"terrestrial",
"environments"
] |
2019
|
Identification of French Guiana sand flies using MALDI-TOF mass spectrometry with a new mass spectra library
|
Hookworm disease is a major global health problem and principal among a number of soil-transmitted helminthiases ( STHs ) for the chronic disability inflicted that impacts both personal and societal productivity . Mass drug administration most often employs single-dose therapy with just two drugs of the same chemical class to which resistance is a growing concern . New chemical entities with the appropriate single-dose efficacy are needed . Using various life-cycle stages of the hookworm Ancylostoma ceylanicum in vitro and a hamster model of infection , we report the potent , dose-dependent cidal activities of the peptidyl cysteine protease inhibitors ( CPIs ) K11002 ( 4-mopholino-carbonyl-phenylalanyl-homophenylalanyl- vinyl sulfone phenyl ) and K11777 ( N-methylpiperazine-phenylalanyl-homophenylalanyl-vinylsulfone phenyl ) . The latter is in late pre-clinical testing for submission as an Investigational New Drug ( IND ) with the US Federal Drug Administration as an anti-chagasic . In vitro , K11002 killed hookworm eggs but was without activity against first-stage larvae . The reverse was true for K11777 with a larvicidal potency equal to that of the current anti-hookworm drug , albendazole ( ABZ ) . Both CPIs produced morbidity in ex vivo adult hookworms with the activity of K11777 again being at least the equivalent of ABZ . Combinations of either CPI with ABZ enhanced morbidity compared to single compounds . Strikingly , oral treatment of infected hamsters with 100 mg/kg K11777 b . i . d . ( i . e . , a total daily dose of 200 mg/kg ) for one day cured infection: a single 100 mg/kg treatment removed >90% of worms . Treatment also reversed the otherwise fatal decrease in blood hemoglobin levels and body weights of hosts . Consistent with its mechanism of action , K11777 decreased by >95% the resident CP activity in parasites harvested from hamsters 8 h post-treatment with a single 100 mg/kg oral dose . A new , oral single-dose anthelmintic that is active in an animal model of hookworm infection and that possesses a distinct mechanism of action from current anthelmintics is discovered . The data highlight both the possibility of repurposing the anti-chagasic K11777 as a treatment for hookworm infection and the opportunity to further develop CPIs as a novel anthelmintic class to target hookworms and , possibly , other helminths .
One of a number of soil-transmitted helminthiases ( STHs ) that is deeply rooted in poverty , hookworm disease afflicts as much as 10% of the world's population in sub-Saharan Africa , South America , and South and South-East Asia [1] , [2] , [3] . The principal etiological agents in humans are the nematodes Necator americanus ( causing necatoriasis ) and Ancylostoma duodenale ( ancylostomiasis ) , although Ancylostoma ceylanicum is found in certain locales [4] , [5] . Hookworm zoonoses of minor medical importance also occur but these usually manifest with the restricted dermatitis condition of ‘cutaneous larva migrans’ , ( e . g . , [6] and references therein ) . Hookworm infection has been described as ‘silent and insidious’ [7] and to ‘drain out the vitality’ [8] of those afflicted due to chronic wasting and lethargy that has been often misconstrued as laziness [9] . Indeed , the notion of ‘draining’ is apt as most pathology arises from adult worms that attach to and feed on intestinal mucosa and blood [10] . Of the two main parasites , A . duodenale is the more voracious and fecund , sucking 0 . 1–0 . 2 mL blood and producing 28 , 000 eggs per day ( [11] and references therein ) . The disease is most strikingly manifested in the under-nourished , not least in directly causing or exacerbating existing iron-deficient anemia that can slow physical and cognitive development in children [8] , [12] , [13] , adversely affect fetal weight and growth , and contribute to premature birth and maternal mortality [14] , [15] . Treatment and control of STHs employs periodic de-worming , particularly of school children , using a small number of well-established drugs [3] , [16] , [17] , [18] . Of the six drugs stated in the World Health Organization's 17th Essential Medicines List for intestinal helminthiases , namely , albendazole ( ABZ ) , mebendazole ( MBZ ) , pyrantel pamoate , praziquantel , levamisole and niclosamide; http://www . who . int/medicines/publications/essentialmedicines/en/ ) , the first two benzimidazoles are most commonly employed to treat hookworm infection . Of these , ABZ is the more effective as a single oral dose drug [19] , [20] , [21] , [22] . This ease of administration , efficacy and safety record makes ABZ ideal for integration into mass drug administration ( MDA ) programs that aim to deliver packages of essential medicines for various tropical diseases ( [17] and references therein ) . Although not necessarily indicating genuine drug resistance [23] , decreased efficacy against hookworm infection has been reported for MBZ [22] , [24] , [25] making ABZ all the more important for control of this global disease . Experience in the animal health sector indicates that resistance to one member of the benzimidazole class usually extends to other members of the same class [16] , [26] . In this regard , the recent reports of less-than-expected cure rates with ABZ are disquieting [27] , [28] ( also a reviewed in [29] and [23] ) , and in one case , cannot be interpreted as either poor drug quality or a lack of patient compliance [27] . Thus , there is continued impetus not just to identify other anthelmintics ( e . g . , [30] , [31] for two notable examples ) but to eventually introduce vaccine therapy ( reviewed in [32] , [33] ) . Here , we report the striking therapeutic effect of the Clan CA ( MEROPS nomenclature [34] ) cysteine protease inhibitor ( CPI ) , K11777 ( N-methyl-piperazine-phenylalanyl-homophenylalanyl-vinylsulfone phenyl ) , in a well-established animal model of hookworm infection [31] , [35] , [36] , [37] , [38] that employs the golden Syrian hamster and A . ceylanicum . Originating in an industrial drug development program to treat osteoporosis [39] , attention to K11777's cidal activity in animal models of eukaryotic parasitism first arose with the cure of acute infection in mice by the etiological agent of Chagas' disease , Trypanosoma cruzi [40] . Subsequent studies have described its therapeutic benefit in various animal models for this [41] and other protozoal infections [42] , [43] , [44] , and for the flatworm parasite , Schistosoma mansoni [45] . In every case , the demonstrated molecular target of anti-parasite action has been Clan CA ( cathepsin ) proteases . K11777 is currently progressing through Investigational New Drug ( IND ) -enabling studies required by the U . S . Food and Drug Administration [46] with the goal of entering clinical trials as early as 2013 for treatment of Chagas' disease . The implications of the present discovery of cure of hookworm infection in an animal model and the possible new disease indication for this drug candidate and class are discussed .
The life cycle of A . ceylanicum is maintained in Golden Syrian hamsters ( Mesocricetus auratus; Harlan Sprague Dawley , Somerville , NJ ) as described [35] , [36] , [37] . Soluble hookworm protein extracts ( HEX ) were prepared [47] and protein concentrations determined using bicinchoninic acid ( BCA ) reagents ( Pierce Biotechnology , Rockford , IL ) . The maintenance and care of experimental animals complied with the National Institutes of Health guidelines for the humane use of laboratory animals and were approved by the Yale University Animal Care and Use Committee . The CPIs , K11002 and K11777 , were originally part of a series of compounds provided by James Palmer of Khepri Pharmaceuticals , South San Francisco , CA [39] . Both inhibitors were in sufficient quantities to perform the comparative experiments described below . A . ceylanicum eggs were recovered from infected hamster feces as described [48] and placed at a density of 100 eggs/well in a 96-well plate containing ABZ ( Sigma ) , K11002 or K11777 serially diluted in water . The total number of eggs and hatched larvae in each well were counted by light microscopy 24 h later . The total number of viable larvae , assessed based on motility , were also counted in each well . Data are expressed as the percentage of hatched larvae and the percentage of viable larvae in experimental wells relative to DMSO controls . To measure the effects of these CPIs on worm viability , A . ceylanicum adult hookworms were employed in a worm killing assay ( WKA ) as described [31] , [49] . Male and female worms were recovered from the small intestines of infected hamsters , washed three times in RPMI 1640 containing 20 U/20 µg/ml penicillin/streptomycin , 10 µg/ml Fungizone ( Invitrogen ) and placed at a density of 10 worms/well ( 2 wells/treatment ) into 24-well plates containing K11002 and K11777 ( 1–100 µM ) diluted in the same medium supplemented with 50% fetal calf serum . Control wells were treated with ABZ ( 50 µM ) or equivalent volumes of DMSO vehicle alone . To assess worm morbidity , individual worms were scored using a 5-point morbidity scale ( Figure S1 ) at 120 h post-treatment ( HPT ) . Groups of hamsters ( n = 6 ) were infected with 75 or 100 third stage A . ceylanicum larvae by oral gavage and followed for 24 days post-infection ( DPI ) to monitor blood hemoglobin levels and weight gain [37] . As indicated in the relevant figure legends , treatment regimens ( commencing 17 DPI ) targeted adult worms with K11777 ( prepared fresh in 200 µl water , q . d . or b . i . d . for up to two days orally or 3 days intra-peritoneally ( i . p . ) : ABZ ( prepared fresh in 200 µl water ) was given orally once or for 3 days . K11777 was chosen over K11002 for these experiments given its better solubility in intestinal fluids and oral bioavailability ( [50] , and references therein ) . On day 24 , hamsters were sacrificed and their intestinal worm burdens counted . To measure the effect of compounds on parasite CP activity , one hamster from each of the once orally-treated K11777 , ABZ and vehicle groups was sacrificed 8 h post-treatment . Worms recovered from hamsters 8 h post-treatment with a single oral dose of 100 mg/kg K11777 , 10 mg/kg ABZ or vehicle , were washed three times in RPMI 1640 and frozen on dry ice prior to assay . Worms were thawed in 100 µl assay buffer ( 0 . 05 M sodium acetate , pH 5 . 5 ) and homogenized using Kontes RNase-free disposable pellet pestles and microtubes for 10 min at room temperature ( r . t . ) . Homogenates were centrifuged at 5 , 000 g for 10 min and the supernatants removed for analysis . Supernatants ( 1–2 . 5 µl ) were mixed with 100 µl assay buffer containing 2 mM DTT in a black 96-well microtiter plate and left to stand at r . t . for 10 min . Then , 100 µL of assay buffer containing 2 mM DTT and 20 µM of the dipeptidyl fluorogenic substrate , benzyloxy carbonyl-phenylalanyl-arginyl-7-amido-4-methylcoumarin ( Z-Phe-Arg-AMC ) [51] was added with mixing . Linear rates of hydrolysis were followed in a Molecular Devices FlexStation for 10 min . In order to determine the contribution of CP activity to the total activity being measured , K11777 at 1 µM ( as a 1 µl aliquot in DMSO ) was added to the incubation prior to addition of substrate . Protein concentrations of supernatants were measured using the micro-Bradford assay ( BioRad ) . Worm morbidity and worm burden data were analyzed using one-way ANOVA and Tukey's Multiple Comparison Test . Hamster weight and hemoglobin data were analyzed using one-way ANOVA and Bartlett's test for equal variances . Paired t-tests were used to compare cysteine protease activities in extracts of worms recovered from hamsters . The GraphPad Prism software application ( version 5 . 01 ) was employed to derive statistics .
A . ceylanicum eggs were employed in an EHA to measure the effect of K11002 and K11777 on egg hatching and larval viability ( Figure 1 ) . We observed that even at the highest concentration tested ( 100 µM ) K11002 did not inhibit egg hatching relative to DMSO controls ( Figure 1A ) . However , K11002 reduced larval viability in a dose dependent manner ( Figure 1D , EC50 = 1 . 1 µM ) . After 24 h of K11002 treatment at concentrations ranging from 0 . 4–100 µM , newly hatched larvae were motionless and displayed wrinkled cuticles indicating death . In contrast , K11777 was a potent dose-dependent inhibitor of A . ceylanicum egg hatching ( Figure 1B , EC50 = 2 . 89 µM ) but did not impact larval viability ( Figure 1D ) . Larvae that had hatched in wells containing K11777 were active , displaying sinusoidal movement patterns and intact cuticles . The ovicidal activity of K11777 was statistically the same as that measured for ABZ ( Figure 1C , EC50 = 1 . 12 µM ) , which also did not interfere with larval viability ( not shown ) . When tested in an in vitro WKA , K11002 and K11777 displayed different potencies in terms of inducing worm morbidity ( Figure 2 ) . For example , at 50 µM , K11002 caused significantly less morbidity ( P<0 . 001 ) compared to ABZ ( Figure 2A , EC50 = 4 . 5±0 . 51 vs . 2 . 8±1 . 01 , respectively ) . In contrast , the morbidity caused by K11777 at 50 µM ( EC50 = 4 . 05±0 . 76 ) or 25 µM ( EC50 = 3 . 8±1 . 2 ) was not significantly different ( P>0 . 05; Figure 2B ) from that of ABZ at 50 µM suggesting that the CPI is as or more potent than the currently employed drug . Finally , at 72 h , combination treatment with ABZ and K11002 ( each at a 25 µM ) increased worm morbidity relative to treatment with either compound alone ( Figure 3 , P<0 . 01 ) . This was also true for combining 25 µM each of ABZ and K11777 at 72 and 96 h post-treatment ( P<0 . 001 and P<0 . 01 , respectively ) . We next tested the ability of K11777 ( the more soluble and bioavailable of the two CPIs in vivo [50] ) given orally ( Figure 4 ) or i . p . ( Figure S2 ) to positively influence animal weight gain , blood hemoglobin levels and intestinal worm burdens as compared to infected vehicle controls . The i . p . experiments using 50 mg/kg K11777 b . i . d . ×3 ( i . e . , total daily doses of 100 mg/kg ) did not significantly improve weight gain ( Figure S2A ) , but significantly improved blood hemoglobin levels ( Figure S2B; P<0 . 05 and P<0 . 001 at 17 and 24 DPI , respectively ) and cured infection ( Figure S2C ) . K11777 , given orally at 100 mg/kg b . i . d . ×1 and ×2 , did not significantly improve weight gain ( Figure 4A ) , but significantly improved blood hemoglobin levels ( P<0 . 001 at 21 and 24 DPI; Figure 4B ) and cured infection ( Figure 4C ) . Finally , a single oral dose of 100 mg/kg K11777 did not significantly improve weight gain ( Figure 4A ) , but significantly improved blood hemoglobin levels ( P<0 . 001 at 21 and 24 DPI; Figure 4B ) and decreased intestinal worm burdens by 90 . 1% ( P<0 . 001; Figure 4C ) . For comparison , a single oral dose of ABZ ( 10 mg/kg ) did not significantly improve hemoglobin levels or hamster weights , but significantly decreased intestinal worm burdens by 100% ( P<0 . 001 , Figure 4C ) . A standard assay using a dipeptidyl fluorogenic substrate for CP activity was employed to understand whether administration of K11777 8 h prior to worm recovery decreases the specific CP activity ( i . e . , as a function of protein concentration ) relative to those activities measured post-exposure to vehicle or ABZ . Thus , soluble extracts of worms exposed to K11777 contained just 5% and 23 . 5% of the CP activity measured in extracts of worms exposed to vehicle ( P = 0 . 0024 ) and ABZ ( P = 0 . 0074 ) , respectively ( Figure 5 ) . Extracts from ABZ-exposed worms contained 23% of the CP activity measured in worms exposed to vehicle ( P = 0 . 0054 ) . It is possible that the latter finding is due to the systemic degradation of cellular architecture and biochemistry inflicted by ABZ rather being due to direct inhibition of CP activity . In all extracts , activity could be abolished with K11777 at 1 µM ( not shown ) indicating that only CP activity was being measured .
The present data indicate that the CPIs K11777 and K11002 impair the survival of multiple life-stages of the human and animal hookworm parasite A . ceylanicum in a dose dependent manner in vitro . For K11777 , EC50 values for killing of eggs and morbidity in adult worms were equivalent to or better than those recorded for the current therapy of hookworm disease , ABZ . In addition , for both CPIs and at some of the time-points tested , there was a significant enhancement of morbidity in the presence of ABZ over compounds used alone . Most importantly , using an animal model of hookworm infection , we demonstrate the striking therapeutic benefit of K11777 administered under different regimens . Not least , a single oral dose of 100 mg/kg removed greater than 90% of adult worms whereas b . i . d . treatment provided cure . Consistent with this potent worm-kill , the trajectories of blood hemoglobin levels and animal body weight that would otherwise result in death of the host were universally reversed upon therapy . Finally , worm-kill after a single oral dose of K11777 was associated with a dramatic ( 95% ) loss of parasite CP activity , consistent with the compound's known mechanism of action ( see below ) . K11777 inhibits Clan CA cysteine proteases [39] including mammalian cathepsins B , L and S ( reviewed in [52] ) , orthologs of which have been successfully targeted to effect therapy in various animal models of parasitic infection ( see Introduction ) . Hookworms and other hematophagous nematodes devote considerable transcriptional effort into expressing a family of cathepsin B-like proteases that are associated with the parasite gut ( esophagus and cecum ) [53] , [54] , [55] and that contribute to degrading blood proteins to absorbable nutrients [56] , [57] , [58] , [59] , [60] , [61] . It is likely that these proteases are targeted by K11777 as evidenced here by the almost total inhibition of CP activity in A . ceylanicum exposed in vivo to the inhibitor relative to either ABZ- or vehicle-treated controls . The expression of hookworm proteases at the host-parasite interface , i . e . , the gut , has encouraged their investigation as vaccine targets ( reviewed in [57] , [62] ) . Modest reductions in worm burdens have been recorded with hookworm cathepsins B as vaccine candidates in hamsters ( N . americanus; up to a 29% decrease in challenge burdens [63] ) and dogs ( A . caninum; 18% [64] ) . Importantly , however , the latter study also showed that parasite fecundity , measured as eggs per gram of feces , was significantly decreased by 61% . In addition , the challenge worms recovered were smaller than controls and specific antibody that bound to the intestinal brush border interfered with the activity of the target cathepsin B [64] . Thus , the immunological evidence demonstration that hookworm viability can be negatively impacted through the targeting of gut cysteine proteases is powerfully underscored here with a chemical inhibitor . Unlike the situation for most of the previous organisms tested whereby relatively long-course dosing ( days or weeks depending on parasite target , e . g . , [40] , [45] ) , was needed to significantly ameliorate infection intensity and pathology , the present discovery of the hookworm parasite's sensitivity to K11777 , including after a single oral treatment , is remarkable and strongly encourages further investigation for two principal reasons . First , the acute response of hookworm infection to K11777 is commensurate with the short ( preferably single ) oral dosing regimen currently employed for the MDA of anthelmintics [2] , [17] , [65] . Second , given that K11777 is already being targeted for submission as an IND to treat Chagas' disease , the inhibitor could also be tested as an experimental drug for hookworm disease . To support the potential clinical use of K11777 in treatment of hookworm infection , it will be necessary to define oral efficacy in other animal models including N . americanus in hamsters and A . caninum or A . duodenale in dogs . Our results using i . p . dosing demonstrate that K11777 kills hookworm via the bloodstream . However , after oral dosing , the relative contribution of this access route and trans-cuticular migration to worm-killing is an open question , as it still is for some current anthelmintics [16] . Additional experimentation with a variety of CPIs that are more or less bioavailable will be required to answer this question - an answer that may influence the possible use of CPIs to treat other non-hematophagous gastro-intestinal nematodes . Pertinent to this discussion is the striking impact of CPI structure on the reciprocal activities of K11777 and ( the more hydrophobic ) K11002 in modulating larval viability . Whether or not K11777 as an entity goes forward as an investigative treatment for hookworm disease , the dramatic susceptibility of the parasite in vivo to CP inhibition encourages a broader investigation of structural analogs and other CPI scaffolds , perhaps similar to those currently under investigation for other parasitic diseases , including malaria and Human African Trypanosomiasis [66] , [67] , [68] . The search might also encompass CPIs under pre-clinical and clinical investigation for treatment of non-infectious diseases such as osteoporosis and auto-immune disorders [69] , [70] , [71] , [72] in order to take advantage of the extensive industrial experience in CPI design [71] , [73] . The benzimidazoles currently used to treat hookworm disease display a spectrum of activity against other , and often co-endemic , STH's . For example , a recent meta-analysis demonstrated that the standard single 400 mg dose of ABZ , in addition to producing a cure rate ( CR ) of 72% against hookworms , is effective against ascariasis ( CR 88% ) , and ( markedly less so ) against trichuriasis ( CR 28% ) [21] , relative efficacies that have been confirmed more recently [74] . For CPIs , it remains to be determined whether a spectrum of activity exists , nevertheless , the discovery of this new class of hookworm anthelmintic with a distinct mechanism of action from ABZ is potentially useful given the concerns regarding the future of benzimidazoles as effective drugs . In addition , as is now common in the animal health industry [16] , [75] and in the treatment of many infectious diseases of humans [76] , [77] , [78] , drug combinations must be considered if the efficacy of the few anthelmintic drugs available is to be protected . Thus , treatment of hookworm disease and other STHs might also benefit from drug combinations in which the contribution of CPIs could be substantial by improving either anthelmintic efficacy and/or spectrum of activity . In this regard , the improved anthelmintic activity measured in vitro when either CPI was combined with ABZ ( Figure 3 ) may be relevant , although further tests ( e . g . , employing different compound concentrations ) are required to fully understand whether additive effects might be involved . To conclude , we demonstrate a remarkable sensitivity of the hookworm parasite in an animal model of infection to the CPI , K11777 . The identification of a novel chemical class of anthelmintic is encouraging given the heavy reliance on benzimidazoles to treat human hookworm disease and the threat of emerging drug resistance .
|
In spite of the enormous prevalence of hookworm disease , just two drugs , albendazole and mebendazole , are most commonly employed for treatment and control , and both belong to the same benzimidazole chemical class . There exists , therefore , a pressing need to develop new , safe and inexpensive agents for the treatment of human nematode infections of global significance . We report the discovery of the striking efficacy of the cysteine protease inhibitor , K11777 , against hookworms both in vitro and in vivo and discuss the development of this class of compounds as novel anthelmintics for the clinical management of hookworm disease . K11777 is chemically distinct from all the current anthelmintics and , therefore , not likely to share resistance characteristics . We describe mechanism of action studies that demonstrate that cysteine protease activity in parasites recovered after in vivo treatment with K11777 is almost completely ( >95% ) abrogated . Lastly , we report that K11777 provides near cure ( >90% ) of hookworm infection in a single oral administration ( complete cure when given twice in one day ) . These results suggest that K11777 is on target to meet the current clinical practice and the logistics demanded for mass drug delivery of anthelmintics to humans ( i . e . , oral , single-dose treatment ) .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"biology",
"microbiology",
"parasitology"
] |
2012
|
Cure of Hookworm Infection with a Cysteine Protease Inhibitor
|
Little is known about contacts in the spliceosome between proteins and intron nucleotides surrounding the pre-mRNA branch-site and their dynamics during splicing . We investigated protein-pre-mRNA interactions by UV-induced crosslinking of purified yeast Bact spliceosomes formed on site-specifically labeled pre-mRNA , and analyzed their changes after conversion to catalytically-activated B* and step 1 C complexes , using a purified splicing system . Contacts between nucleotides upstream and downstream of the branch-site and the U2 SF3a/b proteins Prp9 , Prp11 , Hsh49 , Cus1 and Hsh155 were detected , demonstrating that these interactions are evolutionarily conserved . The RES proteins Pml1 and Bud13 were shown to contact the intron downstream of the branch-site . A comparison of the Bact crosslinking pattern versus that of B* and C complexes revealed that U2 and RES protein interactions with the intron are dynamic . Upon step 1 catalysis , Cwc25 contacts with the branch-site region , and enhanced crosslinks of Prp8 and Prp45 with nucleotides surrounding the branch-site were observed . Cwc25’s step 1 promoting activity was not dependent on its interaction with pre-mRNA , indicating it acts via protein-protein interactions . These studies provide important insights into the spliceosome's protein-pre-mRNA network and reveal novel RNP remodeling events during the catalytic activation of the spliceosome and step 1 of splicing .
The removal of introns from nuclear pre-mRNAs proceeds by way of two phosphoester transfer reactions and is catalyzed by the spliceosome , a large ribonucleoprotein ( RNP ) complex composed of the snRNPs U1 , U2 , U4/U6 and U5 and several proteins [1] . The spliceosome is a highly dynamic RNP machine that undergoes many changes in composition and conformation during its work cycle [2] . Initially , the U1 snRNP recognizes the 5’ splice site ( 5’ SS ) and U2 snRNP recognizes the branch-site ( BS ) of the pre-mRNA , resulting in the formation of the pre-spliceosome or A complex . The pre-formed U4/U6 . U5 tri-snRNP is then recruited , generating the B complex , which does not yet have an active site . Subsequent activation of the spliceosome ( leading to the Bact complex ) involves major rearrangement of the spliceosomal RNA–RNA interaction network . This rearrangement is catalyzed by the combined action of the RNA helicases Prp28 and Brr2 and leads to the displacement of the U1 and U4 snRNAs and the formation of new base-pair interactions between the U2 and U6 snRNAs and the 5’ SS [3] . Thus , a web of RNA–RNA interactions holds the 5’ SS and the BS together for step 1 catalysis [4] . The Bact complex , which contains U2 , U6 and U5 and ~40 proteins in the yeast S . cerevisiae [5] , is converted by the DEAH-box NTPase Prp2 , in co-operation with the G-patch protein Spp2 , into a catalytically activated complex ( B* ) [6–8] . Following the recruitment of the splicing factor Cwc25 [7 , 9] , step 1 is catalyzed , whereby the 2’-OH of the BS adenosine attacks the 5' SS of the pre-mRNA generating the cleaved 5’ exon and intron 3’ exon; concomitantly the C complex is formed . This then catalyses step 2 , in which the 3’ SS is cleaved , resulting in the excision of the intron and ligation of the 5’ and 3’ exons , after which the mRNA product is released . The excised intron lariat remains associated with U2 , U5 and U6 snRNPs , which then dissociate and take part in subsequent rounds of splicing . During the transformation of complex B into Bact , not only is the spliceosome's RNA network radically rearranged , but also its protein composition changes significantly; as a result , several proteins are released , while twelve Bact proteins are recruited . At the same time the NTC ( nineteen complex ) and the NTC-related proteins are stably integrated into the Bact complex [5] . Likewise , the three proteins comprising the RES ( pre-mRNA retention and splicing ) complex [10] are also stably integrated into the Bact complex . Despite the substantial restructuring that the spliceosome has undergone at this point , it does not yet have a functional active site . Previous studies showed that the binding affinity of several proteins is significantly changed during the Prp2-mediated transition of Bact spliceosomes to catalytically activated B* spliceosomes [7 , 11] . During this step , the essential splicing factor Cwc24 is quantitatively displaced from the B* complex . The U2-associated SF3a and SF3b proteins Prp11 and Cus1 and the RES protein Bud13 all remain bound to the B* spliceosome under near-physiological conditions , but their binding is reduced at high salt concentrations [11] . The destabilization of these proteins' binding by Prp2 and Spp2 indicates that the structure of the catalytic core of the spliceosome near the BS is remodeled . This could lead to a proper 5’SS and BS configuration for nucleophilic attack on the 5' SS phosphodiester bond during step 1 catalysis [7 , 12] . However , while it is clear that the affinity of the U2 and RES proteins for the spliceosome is significantly reduced during catalytic activation , it is not known whether this implies remodeling events involving contact of U2 and RES proteins with the pre-mRNA . Likewise , information about the set of spliceosomal protein-pre-mRNA contacts and their dynamics during splicing remains limited , but is crucial for unraveling potential functions of spliceosomal proteins for the formation and maintenance of the spliceosome's RNA–RNA network during catalysis . The U2 snRNA/BS interaction is established in the A complex and is thought to make the BS adenosine bulge out for nucleophilic attack on the 5’ SS during step 1 catalysis [13] . In human pre-spliceosomes and spliceosomes the U2 SF3a/b proteins help to recruit the U2 snRNP to the BS , and all of them except SF3b130 can be crosslinked , in a sequence-independent manner , to a region upstream of the BS ( the so-called “anchoring site” ) , to the BS itself and to a region downstream of it [14 , 15] . The BS sequence is highly conserved in yeast but only weakly conserved in metazoans . Given the short length of the BS sequence , and its degeneracy in metazoans , it has been suggested that spliceosomal proteins function together with the U2 snRNA/BS duplex to tether the U2 snRNP to pre-mRNA during spliceosome assembly . In yeast there is perfect complementarity between the BS sequence and the U2 RNA . Thus the anchoring/stabilization of the U2 snRNP to the BS sequence in yeast could be different from that in human , and it may not depend critically on protein–RNA interactions . Although most of the U2 proteins in the yeast S . cerevisiae are evolutionarily conserved [16–18] , it is not known whether they interact in a similar way with the BS region in yeast spliceosomes , or whether a site equivalent to the human anchoring site exists in yeast pre-mRNAs . So far , only Hsh155 , the yeast homologue of human SF3b155 , has been shown to crosslink to pre-mRNA between the BS and the 3'SS [19] . Furthermore , it was recently shown that the RES subunit Snu17 is in contact with the pre-mRNA downstream of the BS in proximity of U2-Hsh155 [20] . However , is it currently not known whether additional components of the yeast U2 SF3a/b and RES subunits make direct contact with the pre-mRNA in spliceosomal complexes . To address these questions , we have investigated protein–pre-mRNA interactions by UV-induced crosslinking of purified yeast spliceosomes stalled at the Bact assembly stage or after conversion of Bact to B* and C complexes , using a purified yeast splicing system [7] . Results of these studies revealed contacts in Bact complexes between pre-mRNA nucleotides directly upstream of the BS and the yeast U2 proteins Prp9 , Prp11 , Hsh49 , Cus1 and Hsh155; the latter were also in contact with the intron further downstream of the BS . Thus , these interactions are evolutionarily conserved between yeast and man . Consistent with previous results demonstrating a Snu17-pre-mRNA crosslink [20] , we observed that also the RES components Pml1 and Bud13 crosslinked to the intron downstream of the BS in Bact complexes . Subsequent UV crosslinking with purified spliceosomes that had been stalled after catalytic activation by Prp2/Spp2 and consecutive step 1 catalysis by Cwc25 revealed remodeling events involving contacts between U2 SF3a/b proteins upstream of the BS and the RES proteins downstream of it . Finally , concomitantly with these remodeling events , enhanced contacts of Cwc25 , Prp8 and the NTC-related protein Prp45 with the BS and/or 3'SS regions were observed . These studies thus provide novel insights into the extensive protein–pre-mRNA interactions and their dynamics within and surrounding the pre-mRNA BS and 3'SS regions during step 1 of splicing in yeast spliceosomes .
To obtain insights into the nature and number of proteins that are in direct contact with the region at the 3’ end of the intron in purified yeast spliceosomes , we constructed a pre-mRNA which was body-labeled with 32P-UTP during transcription in the 3’ third of the intron , including exon 2 and 47 nucleotides ( nts ) upstream of the BS ( termed hereinafter “3’-region-labeled pre-mRNA”; Fig 1A ) . The experimental strategy used to produce the 3’-region-labeled pre-mRNA is outlined in S1A Fig . Briefly , the 3’ fragment was obtained by T7 transcription . For this purpose , a T7 promoter was added by PCR and the PCR product was transcribed in vitro with an excess of GMP to ensure the presence of a monophosphate at the 5’ end and with α-32P UTP to randomly trace-label the entire RNA transcript ( see S1 Text for details ) . To produce the 5’ fragment , unlabeled actin pre-mRNA , prepared by transcription in vitro , was specifically cleaved between nucleotides 425 and 426 by a DNA enzyme based on the “8–17” catalytic motif [21] ( S1A Fig , upper panel ) . The 5’ cleavage fragment was dephosphorylated , gel purified and ligated to the T7-transcribed 3’ fragment by DNA splint directed RNA ligation [22] . The 3’-region-labeled pre-mRNA allows the analysis by UV crosslinking of protein–pre-mRNA interactions at the BS site , the region directly upstream of the BS as well as around the 3’SS . Protein–pre-mRNA interactions were analyzed initially in purified Bact spliceosomes , which were assembled in vitro by incubating heat-inactivated splicing extracts from a temperature-sensitive prp2-1 yeast strain with the 3’-region-labeled pre-mRNA that also contained an MS2 binding site at its 5’ end [5 , 7] . Bact spliceosomes were purified by glycerol-gradient centrifugation and MS2-MBP affinity chromatography and then were irradiated with UV light at 254 nm , and digested under denaturing conditions with a mixture of RNases T1 , A and I . The entire protein mixture was then separated by two-dimensional ( 2D ) gel electrophoresis as described for human spliceosomal complexes [23] . Our 2D gel electrophoresis method is based on charge-driven separation of proteins under denaturing conditions at acidic pH in the first dimension and further separated by molecular weight though SDS gradient PAGE in the second . In contrast to the commonly used isoelectric focusing ( IEF ) , this system prevents proteins from reaching zero charge and allows separation without in-gel precipitation over a wide range of isoelectric points ( IEPs ) and with masses greater than 300 kDa [23] . Fig 1B shows a RuBPS-stained 2D gel ( S1 Text ) of the total proteins isolated from non-irradiated Bact complexes . Individual protein spots were cut out of the gel and peptides were identified by mass spectrometry ( MS ) . Only a few contaminant proteins were found , such as Xnrn1/Kem1 and Hrb1/Tom34 , which are usually present in small amounts in preparations of yeast spliceosomes [5 , 7] . All the previously identified Bact complex proteins were observed [5]; these included nearly all of the U2 SF3a/b proteins ( i . e . SF3b: Rse1 , Hsh155 , Cus1 and Hsh49; SF3a: Prp9 , Prp11 and Prp21 ) , which could be well separated from each other . The low-MW U2 proteins Msl1 , Rds3 and Ysf3 and the Sm proteins D2 , E , F and G ran out of the gel but could be identified in 2D gels which were modified to improve the resolution of smaller proteins [23] . A subset of U5 proteins ( Prp8 , Brr2 and Snu114 ) , and most proteins of the NTC complex and NTC-related proteins , were also located as single , distinct spots . Proteins of the RES complex ( Ist3/Snu17 , Pml1 and Bud13 ) [10] were also identified . A comparison of previous MS analyses of purified yeast Bact spliceosomal complexes [5 , 7 , 24] with those of our 2D analysis indicates a general reliability of this method for separating and identifying proteins that co-purify with yeast spliceosomal complexes [23] . Fig 1C shows an autoradiography of the 2D gel performed as described above but with UV-irradiated Bact complexes . Exposure to 254-nm UV light is known to induce direct ( zero-length ) crosslinks between nitrogenous bases of nucleic acids and amino-acid side chains when they are in a favorable configuration . We observed prominent 32P-labeled spots of U2-Hsh155 and the NTC-related protein Prp46 , both of which could be superimposed on the RuBPS-stained spots ( indicated by circles in Fig 1C ) . This indicates that the covalent attachment of a few RNA nts to proteins larger than 50 kDa , after UV-irradiation , did not alter significantly their migration behavior . A predominant crosslink in the middle of the autoradiogram was due to the contaminating poly ( A ) -binding protein Hrb1/Tom34 , while other contaminant proteins were crosslinked to pre-mRNA at very low levels or not at all ( marked by asterisks in Fig 1C ) . Prominent radioactive spots were also observed for smaller U2 proteins ( MW < 50 kDa ) , such as U2-Hsh49 and also two proteins of the RES complex ( Pml1 and Snu17 ) . The covalent attachment of RNA nts to smaller proteins led to a shift of their crosslinked species to the acidic region ( i . e . left side ) of the gel in the first dimension; in addition , they were separated into several spots and did not co-localize with the RuBPS stained spot , which were located slightly below ( Fig 1C ) . Nevertheless , in all three cases crosslinked species migrated to the left side of the gel , where no other co-migrating proteins were visible in the RuBPS stained gel , indicating that the crosslinked proteins of interest were not contaminated with other proteins . We recently showed that the RES complex subunit Snu17 crosslinks in the Bact complex to a 14-nt-long region of the pre-mRNA intron downstream of the BS , as shown after digestion with RNase T1 [20] . Thus , the presence of several spots in the 2D gel may indicate that crosslinked pre-mRNA–protein species included shorter digestion products of the 14-nt-long RNA fragment ( note that treatment with three different RNases was performed for 2D gel analysis ) . The same is likely to be true for RES-Pml1 and U2-Hsh49 , whose crosslinked species showed a similar separation behavior ( Fig 1C ) . The unequivocal identification of these proteins will be demonstrated below . We also observed that low levels of additional proteins crosslinked to the 3’ part of the intron , such as the U5 proteins Prp8 and Brr2 , the U2 proteins Prp9 , Prp11 , Cus1 , the RES protein Bud13 and a few Bact-specific proteins ( i . e . Cwc22 , Cwc27/Cwc2 and Isy1 , indicated by dots; Fig 1C ) . These proteins crosslinked much less strongly to the 3’ part of the intron in the Bact spliceosome than those described above , suggesting that they are in contact with the 3’ region of the pre-mRNA but are in a conformation that does not favor the formation of UV-induced crosslinks . Alternatively , digestion with a mixture of three different RNases before 2D gel electrophoresis may lead to partial digestion of the crosslinking site . Next , we focused on the characterization of the crosslinked U2 SF3a/b and RES complex proteins with two major objectives: ( i ) identification of the candidate proteins crosslinked to pre-mRNA by pull-down of their tagged version , and ( ii ) identification of the region/position of crosslinks within the intron . Therefore , we generated yeast strains carrying a C-terminally TAP-tagged version of most of the U2 SF3a/b proteins and the two RES subunits Pml1 and Bud13 . In addition , to localize protein–pre-mRNA interactions to well-defined short regions in the RNA stretch directly upstream of the BS or around and downstream of it , we synthesized site-specifically labeled pre-mRNA ( S1B Fig ) . We prepared eight different actin–pre-mRNA constructs , each of which harbored a single 32P label directly 5’ of a distinct guanosine residue in the neighborhood of the BS ( G452–G516 , summarized in Fig 2A ) . For this purpose , full-length non-32P-labeled pre-mRNAs were cleaved into two pieces at a specific position by using a distinct DNA enzyme; after 5’ 32P-labeling of the 3' piece , the two fragments were ligated by using the DNA splint-directed RNA-ligation method of Moore and Sharp . In this way , full-length pre-mRNAs were recreated , each containing a 32P-label at the desired position [22 , 25 , 26] ( S1B Fig and Methods for details ) . Extracts from the prp2-1 strain harboring the TAP-tagged versions of proteins were then used for assembly of yeast Bact complexes on each of the site-specifically 32P-labeled pre-mRNAs . The purity of Bact complexes was determined by analyzing the composition of their associated snRNAs and pre-mRNA ( i . e . for the presence of U2 , U5L , U5S and U6 snRNA and the absence of splicing intermediates of the pre-mRNA; S2 Fig ) . Each purified Bact complex was irradiated with UV light at 254 nm and disrupted by incubating at 70°C in 3% SDS . After complete digestion with RNase T1 ( which cleaves 3’ of guanosine residues ) , the RNA fragments shown in Fig 2A were obtained , each of which contained a single radioactive phosphate 5’ of the terminal guanosine residue . We then immunoprecipitated TAP-tagged crosslinked proteins with IgG Sepharose beads and analyzed the immunoprecipitates by western blotting using the PAP complex ( peroxidase-anti-peroxidase ) ( Fig 2B and 2C , upper panels , western blot ) . Autoradiography of the membrane revealed the 32P-labeled RNA fragment crosslinked to each precipitated protein ( lower panels ) . Thus , we were able to assign a well-defined pre-mRNA region crosslinked to a known protein and could map the entire intron area spanning from nts 447–516 . We first analyzed the U2 SF3a/b proteins and initially focused on the region upstream and across the BS ( Fig 2B ) . For the RES complex proteins we focused on the region downstream of the BS ( Fig 2C ) because our earlier results showed that Snu17 is in direct contact with this region [20] . Western blotting confirmed that U2 proteins were immunoprecipitated either before irradiation ( –UV ) or after it ( +UV ) ; however , when UV irradiation was omitted , 32P-labeled fragments were not precipitated with the proteins ( Fig 2B , lower panels , autoradiography , lanes 1–4 ) . After UV irradiation , the U2 proteins Prp9 , Cus1 , Prp11 and Hsh49 were found crosslinked to the pre-mRNA fragments 32P-labeled at the G452 and G460 positions ( i . e . fragments 447–452 and 453–460; Fig 2B , lower panels , lanes 5 and 6 ) . None of the U2 proteins analyzed crosslinked to the downstream fragments 461–467 and 468–478 , with the exception of Hsh155 , which crosslinked to the RNA region 461–467 and with lower intensity to the BS region 468–478 ( Fig 2B , lanes 7 and 8 ) . The U2 proteins Rse1 and Prp21 and the two small proteins Rds3 and Ysf3 did not crosslink to pre-mRNA . This is consistent with earlier reports that the putative human homologue of Rse1 , SAP130 , could not be crosslinked to pre-mRNA [15] and Rds3 did not bind RNA in vitro [27] . Taken together , these results indicate that the U2 proteins Prp9 , Cus1 , Prp11 , Hsh49 and Hsh155 are in direct contact with the pre-mRNA in the Bact complex , and their interaction is confined to a 14-nt-long region of the intron upstream of the BS ( 447–460 ) , with the exception of Hsh155 , which interacts in addition with the intron downstream of the BS [19] ( see also below ) . Fig 2C shows a similar pull-down experiment performed with purified , irradiated and RNAse-T1-digested Bact complexes carrying TAP-tagged RES Pml1 and Bud13 proteins . Western blotting showed that Pml1-TAP and Bud13-TAP were immunoprecipitated both before and after UV-irradiation . Pml1-TAP crosslinked the most strongly to the pre-mRNA fragments 483–496 and 500–511 , while Bud13-TAP crosslinked to the fragment 500–511 ( Fig 2C , lower panels ) . Thus , Pml1 interacts directly with the 14-nt-long region of the intron downstream of the BS ( 483–496 ) , and both Pml1 and Bud13 interact further downstream than Snu17 [20] . Next , we expanded our analysis to the dynamics of protein–RNA interactions during catalytic activation and step 1 catalysis by using a purified yeast splicing system to investigate changes of UV crosslinking intensities in purified yeast spliceosomes stalled at specific assembly stages after Bact , such as the B* and C complex stages [7] . Bact spliceosomes were prepared as above by incubating distinct site-specifically labeled actin pre-mRNAs ( Fig 3A ) with prp2-1 heat-inactivated splicing extracts in which proteins were untagged . Each Bact spliceosome was affinity-purified as above . One portion of Bact spliceosomes was complemented with ATP and recombinant Prp2 and Spp2 ( whereby transformation of complex Bact to B* occurs ) and one portion was complemented with ATP plus Prp2 , Spp2 and Cwc25 ( whereby transformation of complex B* to C occurs and step 1 is catalyzed ) . Spliceosomes were then further purified by glycerol-gradient sedimentation . The actual conversion from Bact to B* to C was analyzed for each purified complex by gel electrophoresis ( S3 and S4 Figs ) . The presence of U2 , U5 and U6 snRNA , and the total ( or , for B* , nearly total ) absence of splicing intermediates of the pre-mRNA confirmed the identity of the Bact and B* complexes , while the presence of step 1 products confirmed the identity of the C complex ( S3C and S4C Figs ) . In addition , the efficiency of conversion of Bact to B* was determined by western-blot analysis , which revealed the nearly complete dissociation of the splicing factor Cwc24 from the B* complex during catalytic activation , as previously shown by MS and dual-color fluorescence cross-correlation spectroscopy ( dcFCCS ) [7 , 11] . The efficiency of conversion of B* to C complexes was monitored from the formation of step 1 splicing products , analyzed by 8% denaturing RNA PAGE and quantified by phosphorimager . The % of step 1 products ( compared to the total RNA in a lane ) was calculated to be ~40% ( S3D and S4D Figs ) . Peak fractions of purified Bact , B* and C complexes were irradiated with 254-nm UV light and–after denaturation and digestion with RNase T1 –the crosslinked 32P-labeled proteins were analyzed by SDS-PAGE . The gel was subjected to autoradiography ( Fig 3A ) . Each site-specifically labeled pre-mRNA showed a distinct crosslinking pattern , revealing bands of different intensities and masses . Fig 3A ( lanes 1 and 4 ) shows different degrees of crosslinking of four proteins with sizes consistent with the apparent molecular masses of untagged Prp9 , Cus1 , Prp11 and Hsh49 , which crosslinked to the pre-mRNA fragments 447–452 and 453–460 in the Bact complex . To ascertain that the four untagged crosslinked proteins corresponded to Prp9 , Cus1 , Prp11 and Hsh49 as shown in Fig 2B , we compared untagged and tagged proteins in parallel experiments . The molecular masses of tagged proteins are increased by a predicted 21kDa , along with the complete disappearance of the untagged version . S5 Fig lane 2 shows the patterns of untagged Hsh49 , Prp11 , Cus1 and Prp9 crosslinked to the pre-mRNA fragment 32P-labeled at G460 in the Bact complex . When the crosslinked proteins were compared with their tagged versions , we observed that Hsh49 shifted from 25kDa to ~50kDa ( compare lanes 2 and 5 , red arrow ) , Prp9 shifted from 60kDa to ~90kDa ( compare lanes 2 and 3 , yellow arrow ) , and Prp11 and Cus1 showed the expected size-shifts consistent with the addition of the TAP-tag ( compare lane 2 with lanes 4 and 6; green and blue arrows , respectively ) . Similar comparisons were performed for the identification of Hsh155 and the RES proteins ( S5B and S5C Fig ) . Taken together , these data allow assignment of the radioactive bands shown in Fig 3A to the proteins indicated ( on the left and right of the gel ) . The pattern in lane 1 of Fig 3A shows that Hsh49 crosslinked in highest yield to the ~6-nt-long region 447–452 , whereas Prp9 and Cus1 crosslinked at low levels to the same fragment . Prp11 was not clearly distinguishable from Hsh49; however , as shown by a light exposure of the gel in Fig 3B , its crosslinking yield was very weak . Although the chemistry of the different sites in the RNA and proteins may affect the intensity of the crosslinks , these results do suggest that Prp9 , Cus1 and Prp11 make no close contacts with this particular RNA region . Remarkably , however , the intensities of Prp9 and Cus1 crosslinks were much stronger in the ~8-nt-long region 32P-labeled further downstream ( i . e . at G460 ) in the Bact complex , whereas crosslinks of Hsh49 ( and Prp11 ) remained unchanged in this downstream ~8-nt-long region ( Fig 3A , lane 4; see also Figs 3B and S5A for the crosslinking intensity of Prp11 ) . Intriguingly , upon conversion of the Bact to the B* and C complexes , crosslinks of Hsh49 to the ~6-nt-long region ( 447–452 ) were greatly reduced , as shown by light exposure of the gel in Fig 3B ( lanes 1–3 ) . Quantification of the intensity of Hsh49 crosslinks indicated that it was reduced by 40% and 80% in the B* and C complexes , respectively , relative to the Bact complex ( S6A Fig ) . This indicates that remodeling of the spliceosome leads to a reduced interaction of Hsh49 with the ~6-nt-long region of the intron . Quantification of Prp11 crosslinks was difficult , as its weak signals did not resolve well from the strong signals of Hsh49 . Interestingly , reduced crosslinking yield of Hsh49 to the downstream ~8-nt-long fragment ( 453–460 ) was also observed during spliceosome remodeling . However , the yield of crosslinks of Hsh49 to this region were reduced to a less significant extent in the B* and C complexes compared with the Bact complex ( 20% and 40% , respectively; S6A Fig ) . This indicates that Hsh49 maintains a relatively strong interaction with the ~8-nt-long region of the intron during catalytic activation and step 1 catalysis , compared to the upstream ~6-nt-long region . Similarly , during spliceosome remodeling the levels of crosslinking of Prp9 to the ~8-nt-long fragment ( 453–460 ) decreased by ~40% in both B* and C complexes compared with the Bact complex , indicating that also the binding site of Prp9 is destabilized ( Figs 3A and 3B , lanes 4–6 , and S6A ) . Although Cus1 was difficult to quantify , the pattern of its crosslinking seemed reduced to a similar extent ( see Figs 3B , lanes 4–6 , and S6A ) . Taken together , these results revealed that contacts of Hsh49 , Prp11 , Cus1 and Prp9 with adjacent regions of the intron were reduced to various degrees during the spliceosome’s conformational changes , indicating remodeling of the binding sites of these proteins . To obtain independent evidence of the decrease over time of Hsh49 crosslinks , we performed 2D gel electrophoresis with UV-irradiated and RNase-digested B* and C complexes assembled on the 3’-region-labeled pre-mRNA , and compared the intensities of the signal from their crosslinked species with those observed in corresponding experiments with the Bact complex ( Fig 3C ) . Consistently with the data shown in Fig 3A , the crosslinking level of Hsh49 to the 3’-region-labeled pre-mRNA was high in the Bact complex . However , the crosslinking level of Hsh49 decreased by ~10% in the B* complex and by more than 40% in the C complex relative to the Bact complex ( Fig 3C ) , as determined by quantification of the radioactive spots ( S6B Fig for quantification of Hsh49 in 2D gels ) . Thus , these results confirmed that contacts of Hsh49 with the 3’ regions of the intron were reduced during the spliceosome’s conformational changes , indicating remodeling of its binding sites . A protein with the molecular mass of ~110 kDa , which was identified as Hsh155 ( S5B and S5C Fig ) , crosslinked upstream and downstream of the BS . Hsh155 crosslinked strongest to the pre-mRNA fragment immediately upstream of the BS ( 461–467 ) , and weakest to the BS fragment itself ( 468–478; Fig 3A , lanes 7–12 ) , consistently with the result of the pull-down experiment shown in Fig 2B . Furthermore , we observed enhanced crosslinking of Hsh155 to the entire region downstream of the BS ( Fig 3A , lanes 13–21 ) . These results indicate that Hsh155 is in contact with a large 50-nt-long region of the intron and it spans the BS; however , its pattern of interaction with the intron does not seem to change significantly during remodeling of the Bact to B* and C complexes , with a decrease in the yield of crosslinking of only ~20% ( S6A Fig ) . There were also additional crosslinks that were not compatible with any obvious U2 snRNP protein equivalent ( Fig 3A , marked with question marks ) . This suggests that there are additional proteins that also contribute to the protein–pre-mRNA interaction network in this region . Of interest is a 25 kDa protein that crosslinked with low intensity to the 11-nt-long BS fragment in the C complex ( see below for the characterization of this protein ) . In addition , a ~15 kDa protein was observed that crosslinked to the BS fragment in Bact , B* and C complexes; the identity of this protein could not be determined either by 2D gel electrophoresis or by tagging the small U2 proteins Rds3 or Ysf3 . Furthermore , crosslinking of Prp8 and Prp45 was identified ( see below for a detailed description ) . To shed some light on the dynamic interactions between the RES complex proteins and the intron , we investigated possible changes of their crosslinking pattern as described above for the U2 proteins . A protein of ~20 kDa was efficiently crosslinked to the pre-mRNA fragment 483–496 ( Fig 3A , lanes 16–18 ) . This protein , which was identified as Snu17 ( see S5B and S5C Fig , lane 3 ) , crosslinked strongest in the Bact complex; however , the intensity of crosslinking decreased by ~ 70% in the B* and C complexes relative to the Bact complex ( Figs 3B and S6C ) , indicating that the interaction of Snu17 with the 14-nt-long region 483–496 is weakened after activation of the spliceosome by Prp2 ( Fig 3A and 3B , lanes 16–18 , see S7 Fig for an independent experiment ) . Analysis of Snu17 crosslinks by 2D gel electrophoresis , confirmed that the binding of Snu17 to the 3' region of the pre-mRNA is drastically reduced during spliceosome remodeling ( Figs 3C and S6B for quantification of Snu17 in 2D gels ) The intensity of Pml1 crosslinking was much lower than that of Snu17 in the same region of the intron , suggesting that Pml1 either makes no close contacts with this region or simply binds in a manner unsuitable for forming crosslinks . Nonetheless , during transition of the spliceosome from Bact to B* the crosslinking intensity of Pml1 was reduced by ~50% and by another 10% from B* to C ( Figs 3A , lanes 16–18 and S6C ) . A similar decrease in crosslinking intensity was confirmed by 2D gel electrophoresis ( Figs 3C and S6B for quantification of Pml1 in 2D gels ) . Likewise , the levels of Pml1 and Bud13 crosslinks to the region further downstream ( 500–511 ) decreased by ~50–60% in the B* and C complexes relative to Bact ( Figs 3A , lanes 19–21 , and S6C ) . The intensity of Snu17 crosslinking was dramatically reduced in the 500–511 region of the intron , as compared with that upstream ( 483–496 ) , indicating that the closest interaction of Snu17 with the intron is with the 14-nt-long region 483–496 . This is consistent with results of our earlier pull-down experiments , which showed that Snu17 interacts directly with this region [20] . Taken together , these data indicate remodeling events involving RES protein contacts with the intron region downstream of the BS upon spliceosome conformational changes . A large protein with a molecular mass of ~250 kDa , which is the expected size for Prp8 , crosslinked in very low yield to regions 461–467 and 468–478 of the intron ( Fig 3A , lanes 7–12 , indicated by arrowheads ) . Quantification of these crosslinked species revealed that Prp8 crosslinked to the region 461–467 of the intron: first in the B* complex and with increased level ( by ~30% ) in the post-step 1 spliceosome ( Figs 3A , lanes 8 and 9 , and S6C ) . Interestingly , Prp8 crosslinked weakly also to the BS sequence 468–478 in the C complex , indicating that during/after step 1 catalysis , Prp8 is favorably positioned for interaction with the BS ( Fig 3A , lane 12 ) . A stronger Prp8 crosslink was observed further downstream , to the 14-nt-long region 483–496 in the B* complex , the intensity of which was enhanced in the C complex ( Fig 3A , lanes 16–18; see also S7 Fig for an independent experiment ) . Thus , our results indicate that Prp8 is favorably positioned for its interaction with the BS upon catalytic activation of the spliceosome by Prp2/Spp2 , and with the 3’SS region upon subsequent step 1 catalysis by Cwc25 . Again independent evidence of the temporal increase of Prp8 crosslinks was obtained by 2D gel electrophoresis performed with UV-irradiated and RNase-digested B* and C complexes assembled on the 3’-region-labeled pre-mRNA , and the intensities of their crosslinked species was compared with those observed in the Bact complex ( Fig 3D ) . Consistent with the data shown in Fig 3A , the crosslinking level of Prp8 to the 3’-region-labeled pre-mRNA was very low in the Bact complex , indicating that Prp8 makes no close contacts with the 3' region of the intron before catalytic activation by Prp2/Spp2 . However , the crosslinking level of Prp8 increased more than 60% in the B* complex and even more than 90% in the C complex relative to the Bact complex ( Figs 3D and S6B ) . Taken together , these results indicate that contacts of Prp8 with the BS and 3’SS regions begin during/after the catalytic activation by Prp2/Spp2 and the interaction with the 3'SS is enhanced after step 1 catalysis promoted by Cwc25 , and are consistent with previous results that showed contacts of Prp8 with the 3'SS subsequent to step 1 catalysis in yeast extracts [28–31] . To determine the identity of the 25kDa protein , which crosslinked to the BS fragment 468–478 in the post-step 1 spliceosome , we used recombinant full-length Cwc25 and truncated variants thereof in reconstitution of the C complex . Reconstituted C complexes were UV-irradiated and RNAse-T1-digested as above . Fig 4A shows that recombinant Cwc25 crosslinked to the BS fragment in the C complex ( lane 6 ) . A truncated variant of Cwc25 ( residues 1–168 ) , lacking 11 amino acids at the C-terminus , showed a similar crosslinking yield ( lane 5 ) , In contrast , the two variants Cwc25 1–102 and 1–125 , lacking 77 and 54 amino acids at their C-termini , did not crosslink to this region ( lanes 3 and 4 ) . Consistently with previous observations [32] , this experiment demonstrates that Cwc25 is in contact with the BS sequence and that at least the N-terminal 168 amino acids of Cwc25 are needed for this . Intriguingly , the addition of the truncated version of Cwc25 1–168 ( Fig 4A , lane 5 ) to B* spliceosomes promoted step 1 catalysis , which was even more efficient than that observed with the full-length version ( Fig 4B , compare lanes 5 and 2 ) . Surprisingly , Cwc25 1–125 promoted step 1 catalysis even in the absence of RNA crosslinking ( Fig 4B and 4A lanes 4 ) , indicating that Cwc25’s activity in promoting step 1 can be uncoupled from its activity in RNA-binding/crosslinking . This result suggests that Cwc25 1–125 may still interact with one or more proteins in the neighborhood of the BS and thus render the microenvironment of the catalytic center favorable for step 1 catalysis . Candidate proteins for interaction with Cwc25 are Prp8 and Hsh155 , which are shown here to crosslink to the BS region concomitantly with Cwc25 ( Fig 4A ) . Furthermore , Yju2 may also interact with Cwc25 , as it was previously shown to be involved in recruiting Cwc25 to the spliceosome [9] . In addition to the NTC , two splicing factors , namely Prp45 and Prp46 [33] that interact with components of the NTC , and whose function is related to NTC in human and yeast , were also observed in the 2D gel carried out with Bact spliceosomes ( Fig 1B ) . We observed that Prp45 did not crosslink in Bact complexes assembled on the 3’-region-labeled pre-mRNA after irradiation with UV light ( Fig 1C ) , in contrast , Prp46 , which was previously shown to interact with Prp45 in vitro and in vivo [33] , crosslinked in high yield to the 3' end of the intron in Bact spliceosomes ( Fig 1C ) . To determine whether this crosslink was retained during spliceosome remodeling , we prepared 2D gels from crosslinked , RNase-digested B* and C complexes assembled on the 3’-region-labeled pre-mRNA . Fig 5A shows that the crosslink of Prp46 was preserved with a similar yield in the B* complex but it increased by ~20% in the C complex ( S6B Fig ) . This indicates that Prp46 remains in contact with the 3’ end of the intron during remodeling of the Bact to B* and to C complexes . To map more precisely the RNA interaction site of Prp46 , we performed UV crosslinking of Bact spliceosomes assembled on site-specifically labeled pre-mRNAs in Prp46-TAP extract ( Fig 5B ) . After pull-down , we observed that Prp46-TAP crosslinked weakly to both RNA fragments labeled at G511 and G516 ( Fig 5B , lanes 7 and 8 ) . Despite the strong crosslink of Prp46 observed in 2D gels obtained from the Bact complex ( Fig 5A ) , we detected low levels of Prp46 crosslinks in the Bact complex assembled on each of the two site-specifically labeled pre-mRNAs ( lanes 7 and 8 ) . Taken together , these results suggest that the prominent crosslink of Prp46 in the 2D gel may be due either ( i ) to an additional crosslinked protein co-migrating with Prp46 or ( ii ) to interaction with a region located further upstream than the region 479–482 . Alternatively , or additionally , the TAP-tag fused to Prp46 may prevent efficient crosslinking of Prp46 to the intron ( Fig 5B ) . Although Prp45 did not crosslink in Bact complexes assembled on the 3’-region-labeled pre-mRNA after UV irradiation ( Figs 1C and 5A ) , we nonetheless observed a protein with the expected size of Prp45 ( i . e . ~42kDa ) , which crosslinked in low yield to the fragment 483–496 , in C complexes ( Figs 3A , lane 18 marked by a dot , and S7 ) . The identity of Prp45 was determined by pull-down of crosslinked and T1-digested complexes containing Prp45-TAP , assembled on pre-mRNA site-specifically labeled at G496 ( Fig 5C ) . Prp45-TAP crosslinked ( with low intensity ) only in the C complex ( lane 6 ) , indicating that Prp45 makes contact with the region of the intron 483–496 after step 1 catalysis . Independent evidence that Prp45 contacts the pre-mRNA upon step 1 catalysis was obtained again from the analysis of 2D gels obtained from crosslinked B* and C complexes that were assembled on the 3’-region-labeled pre-mRNA ( Fig 5A ) . Prp45 crosslinked in the B* complex at low levels , yet the intensity of this crosslinked species increased more than 80% in the C complex ( Figs 5A , S6B and S7 ) . This result suggests a temporal interaction of Prp45 with the intron's region near the 3' SS upstream of Prp46 , and indicates that Prp45 contacts the 3’ end of the intron after/during step 1 catalysis .
In the human system , the U2 protein-pre-mRNA interactions are already established in the early A complex but remain in the rearranged , activated spliceosome [15 , 34] . Here , we analyzed U2 protein–pre-mRNA interactions initially in purified Bact complexes , likely our data obtained with Bact complexes apply also to earlier complexes ( i . e . A and B complexes ) , which for practical reasons were not analyzed here . Using a combination of UV crosslinking and immunoprecipitation of TAP-tagged proteins , we were able to assign a number of U2 snRNP proteins crosslinked to specific sites using pre-mRNAs that were labeled at specific positions by a combination of DNA enzymes cleavage and splint-directed ligation [22] . Site-specific labeling of the RNA with 32P is a much more promising approach for UV crosslinking studies , because the RNA–protein interaction site can be precisely mapped on the RNA . For the first time , we were able to use purified yeast spliceosomes to perform a comprehensive protein–pre-mRNA interaction analysis and thus to assign a well-defined RNA region crosslinked to a known protein . In this way we were able to map an extensive area , spanning a 70-nt-long region of the intron . Consistent with previous studies with human spliceosomal complexes , crosslinking sites involving the yeast U2 SF3a proteins Prp9 and Prp11 , as well as SF3b proteins Cus1 and Hsh49 , were observed within a 14-nt-long region upstream of the BS of affinity-purified Bact complexes ( Fig 2B ) . Furthermore , consistent with previous results obtained in yeast [19] and human [14] spliceosomes , immunoprecipitation revealed contacts between a region located further downstream ( surrounding the BS ) and SF3b Hsh155 . These results provide evidence that the region directly upstream of the BS , and surrounding it , is the main interaction platform of the yeast U2 snRNP proteins . Likewise all human U2 snRNP-associated SAPs , except for SAP130 , were found in direct contact with a 20-nt-long region upstream of the BS in the isolated spliceosomal complexes A , B , and C [34 , 35] and SF3b155 was also found to bind to a site downstream of the BS [14] . Thus , our data furthermore suggest that U2 protein-pre-mRNA interactions with the regions upstream and downstream of the BS are conserved between yeast and human ( S1 Table ) . Furthermore , consistent with earlier findings [15] , an oligoribonucleotide complementary to the 14-nt-long region upstream of the BS inhibits formation of the yeast A , B and Bact complexes ( S8 Fig ) . Thus , interactions of SF3a and SF3b with the pre-mRNA appear to be a prerequisite for pre-spliceosome formation also in yeast , indicating that the perfect complementarity between the BS sequence and the U2 snRNA is not sufficient to anchor the U2 snRNP to the BS sequence , and that stable binding of U2 is largely dependent on U2 protein–pre-mRNA interactions . The RES complex is a conserved , spliceosome-associated module that has been shown to enhance splicing of a subset of transcripts and to promote the nuclear retention of unspliced pre-mRNAs in yeast [10] . Furthermore , it was shown to be required for efficient splicing of TAN1 pre-mRNA , and the intron sequence between the 5'SS and the BS was necessary and sufficient to mediate dependence upon RES [36] . Here , we identified low-yield crosslinks between the BS and the 3’SS of both Pml1 and Bud13 . Consistent with results from our earlier studies [20] , we show that the other RES complex protein , Snu17 , is in direct contact with the pre-mRNA in the region between the BS and the 3’SS within a 14-nt-long RNA stretch upstream of the G nucleotide at position 496 . Our data do not conflict with the result observed with the TAN1 pre-mRNA because the requirement of TAN1 intron nts upstream of the BS for RES dependence could be transient , and an interaction may occur earlier during spliceosome assembly . Our observation of direct contact between Snu17 and the intron is in agreement with earlier reports showing that Snu17 consists primarily of a RRM , which is probably involved in contacting the RNA . It was also reported that the RRM of Snu17 is atypical and acts as a central binding platform that provides two separate interaction surfaces , which interact with disordered parts of Bud13 and Pml1 at the same time [37–39] . While Bud13 and Pml1 do not harbor typical RNA-binding domains , Bud13 contains a conserved lysine-rich region that might bind RNA . In addition , Pml1p or U2 proteins interacting with RES in the spliceosome ( such as the SF3b Hsh155 ) might facilitate the recognition of RNA by RES . A recent NMR solution structure of the core of the RES complex revealed that complex formation leads to intricate folding of the three components that stabilize the RRM fold of Snu17 upon binding of Bud13 and Pml1 , while RNA binding efficiency is increased [20] . Taken together , our results indicate that Snu17 crosslinks directly to the intron between the BS and the 3’SS in the Bact complex , while Pml1 and Bud13 may make contact with the intron through their elaborated interconnection with Snu17 , but they may be in a conformation that does not favor the formation of UV-induced crosslinks ( see Fig 6 for a summary and S2 Table ) . As Hsh155 is in contact with nucleotides of the pre-mRNA between the BS and 3’SS ( Fig 3 ) , which are also in contact with all three components of RES , this indicates that Hsh155 and the RES proteins are in close proximity to one another in the Bact spliceosome . This is consistent with earlier studies showing by a yeast two-hybrid screen and co-immunoprecipitation experiments that Snu17 interacts with U2 SF3b proteins [18] . Furthermore , the RES complex subunit Snu17 was shown to bind to the U2 snRNP [41] . Taken together , all these studies indicate that there is a direct interconnection between RES , the U2 SF3b proteins and the pre-mRNA downstream of the BS . Examination of U2 protein–pre-mRNA interactions in purified spliceosomal complexes stalled after catalytic activation by Prp2/Spp2 ( B* complex ) and subsequent step 1 catalysis by Cwc25 ( to form the C complex ) , revealed that the spliceosome structure involving the region of the intron upstream of the BS and the SF3 proteins Prp9 , Hsh49 and Cus1 undergoes a conformational change during spliceosome activation and subsequent step 1 catalysis . That is , crosslinks of Prp9 , Hsh49 and Cus1 were significantly reduced in both spliceosomal complexes compared with those observed with the Bact complex , indicating that binding to pre-mRNA of these proteins is destabilized after ATP hydrolysis by Prp2 . This is consistent with the remodeling of the structure of the catalytic core of the spliceosome near the BS upon nucleophile attack on the 5' SS phosphodiester bond during step 1 catalysis . That is , alterations in U2 protein binding are probably due to conformational changes that destabilize the interactions of these proteins with the pre-mRNA upstream of the BS concomitant with step 1 . Intriguingly , our data reveal that contacts of Hsh49 , Cus1 and Prp9 ( and to a lesser extent Prp11 ) with two adjacent short regions of the intron upstream of the BS were reduced , to different degrees , during the remodeling of the spliceosome . This indicates that SF3 proteins remain in contact with the intron upstream of the BS even after step 1 catalysis , yet their binding affinity to the pre-mRNA is significantly reduced at a certain site and partially abolished at another . Our results further indicate that the complete set of U2 proteins remains in contact with the U2 snRNA via protein–RNA or protein–protein interaction . Indeed , it was shown that the U2 snRNP is released from the intron-lariat spliceosome in vitro as an integral snRNP , indicating that it remains intact during the entire splicing cycle and that none of its proteins are lost under physiological conditions in vitro [24] . Furthermore , as suggested by their decreased efficiency of crosslinking during spliceosome remodeling ( Fig 3 ) , the binding of the RES complex as a whole is reduced . Remodeling events involving the RES complex proteins are intriguing because the same stretch of the intron is also bound by Prp2 and is essential for Prp2- and Spp2-mediated catalytic activation [8 , 42] . Indeed , Prp2 was crosslinked to the same nucleotides of the intron as the RES proteins [8]; thus , it may be possible that Prp2 recognizes this stretch of RNA "productively" only when it is in contact with the RES proteins . The RES complex could be recognized as an entry point or primary target by Prp2/Spp2 to initiate translocations along the intron [42] , thereby destabilizing RNA-bound proteins and acting as a classical RNPase . Alternatively , it was recently suggested that Prp2 , in addition to binding the intron , is probably involved in several protein–protein interactions in the spliceosome [8] . This would lead to the formation of a relay system that could transmit a power stroke within the motor module of Prp2/ATP , through the various anchor points that Prp2 shares with other components of the spliceosome [8] . Thus , the RES–the binding of which is destabilized upon Prp2/Spp2-mediated B* formation [11] ( and this work ) –could be an important primary element of this communication system . Importantly , it was recently reported that a prp2 mutant was suppressed by deletion of PML1 , indicating that Pml1 stabilizes an interaction that Prp2 destabilizes [43] . To date , Prp8 is the only spliceosomal protein that has been shown to crosslink to all the three regions in pre-mRNA that are required for splicing ( 5’SS , 3’SS , and BS ) , as well as to U5 and U6 snRNAs [44 , 45] . Here , by 2D gel electrophoresis of affinity-purified Bact complexes , we provide evidence that Prp8 –although already stably associated with Bact–makes no close contacts with the 3’ region of the intron before catalytic activation by Prp2/Spp2 . Using affinity-purified spliceosomes stalled at the B* and C stages , we show that conformational changes leading to step 1 catalysis bring Prp8 to a position near the BS and the 3’SS ( Figs 3 and S7 ) [28–31] . That is , remodeling at the catalytic core of the spliceosome accompanies stabilization of Prp8–pre-mRNA contacts . Intriguingly , previous work showed that high-affinity binding sites are created in the B* complex–also for additional factors required for step 1 catalysis such as Yju2 and Cwc25 –during catalytic activation [11] . Thus , the ATP-dependent Prp2-driven activation of the spliceosome leads not only to reduced contacts with the pre-mRNA of U2 and RES proteins , but also to stabilization of other protein–pre-mRNA interactions by promoting direct contact with the pre-mRNA . Taken together , these results provide new insight into the dynamics of protein–pre-mRNA interactions ( simultaneous reduction of some and enhancement of others ) within the spliceosome during its catalytic activation and catalysis . Step 1 catalysis cannot occur efficiently without Cwc25 [7 , 9] . After Prp2-mediated catalytic activation of the spliceosome , a strong binding site is created on the B* spliceosome for the step 1 factor Cwc25 . While Cwc25 only shows background binding to complex Bact , its binding to complex B* has a Kd value in the subnanomolar range [11] . Consistent with the enhanced binding of Cwc25 upon the action of Prp2 , we show here that Cwc25 crosslinks to the 11-nt-long BS fragment . These data are in agreement with earlier reports showing that Cwc25 crosslinks to the intron sequence three bases downstream of the BS [32] . Interestingly , using truncated versions of recombinant Cwc25 for reconstitution of C complexes , we observed that Cwc25 1–125 ( lacking 54 amino acids at its C-terminus ) promoted step 1 catalysis even in the absence of RNA crosslinking , indicating that Cwc25’s step-1-promoting activity is not coupled to its pre-mRNA interacting activity . This indicates that contacts of Cwc25 1–125 would theoretically occur with one or more proteins in the proximity of the BS , thus making the microenvironment of the catalytic center suitable for step 1 catalysis . This would be consistent with Cwc25 being one of the intrinsically disordered proteins , which are highly connected or “promiscuous” proteins that undergo several simultaneous or sequential interactions and use regions of disorder as a scaffold for assembling an interacting group of proteins [46] . Thus , Cwc25 might act as an important hub in the catalytic center of the spliceosome . Indeed , we observed Cwc25’s contacts in the BS region of C complexes concomitant with enhanced crosslinking of Prp8 and Prp45 to the same or a slightly downstream region ( Figs 3–5 ) , suggesting that Cwc25 co-ordinates the catalytic center through protein–protein interaction . We showed that Prp45 contacts the pre-mRNA only after step 1 catalysis ( Figs 5C and S7 ) , although it is already associated with the spliceosome at the Bact stage ( Fig 1B ) . Earlier results showed that , in addition to their interaction in two-hybrid screens , Prp45 and Prp46 interact in vitro , most probably through direct protein–protein contact [33] . Here , Prp45 crosslinked to the pre-mRNA region of the intron 483–496 and Prp46 to the region immediately downstream ( i . e . 500–516 ) , indicating that the two proteins are also in close contact during spliceosome remodeling . Interestingly , Prp45 crosslinked during or after step 1 catalysis to the same pre-mRNA region of the intron ( i . e . 483–496 ) where Prp8 was also found to crosslink with highest yield ( S7 Fig ) , indicating that profound remodeling events involving this region occur . Indeed , a simultaneous reduction of Snu17 and Pml1 contacts was also observed ( Figs 6 and S7 for a summary ) . The contact of Prp45 to this region is consistent with earlier work that showed that Prp45 interacts with Prp22 , a DEAH-box RNA helicase involved in spliceosome disassembly [33 , 47] , which was also crosslinked to the 3’SS in post-step 1 spliceosomes [48] . In addition , a temperature-sensitive allele of Prp45 was shown to be synthetically lethal with alleles of several second-step splicing factors ( Slu7 , Prp17 , Prp18 and Prp22 ) and with several NTC components . Thus , Prp45 may be required for Prp22 as well as the recruitment or stabilization of additional step 2 factors , and the positioning of Prp45 close to the 3’SS after step 1 may determine the timing of this event . The timing of the direct interaction of Prp45 with the pre-mRNA may explain the contribution of this protein to step 2 catalysis: this could be effected either ( i ) by participating in maintaining the step 2 conformation , or ( ii ) by binding and regulating the Prp22 ATPase/translocase activity [47] .
In a first step , actin pre-mRNA was prepared by transcription in vitro with T7 RNA polymerase ( S1 Text ) . The transcription reaction was not gel-purified , but instead was precipitated with ethanol . After washing twice with 70% ethanol , the precipitated RNA was dried , dissolved in 50μl CE buffer ( 10 mM cacodylic acid-KOH , pH 7 . 0 , 0 . 2 mM EDTA-KOH , pH 8 ) or water and applied to a G 50 spin column ( GE Healthcare ) . The eluted RNA was then subjected to DNA enzyme cleavage , essentially as described previously [25] . First , a threefold molar excess of the DNA enzyme over the pre-mRNA was added to the reaction mixture . The solution was then adjusted to 15 mM NaCl and 5 mM TRIS-HCl , pH 7 . 7 . After denaturation at 70°C for 2 min the mixture was kept at room temperature for 5 min . Finally , 150 mM NaCl , 50 mM Tris-HCl , pH 7 . 7 and 2 mM of both MgCl2 and MnCl2 were added and the mixture was incubated at 30°C for 3 hrs . To remove the cyclic phosphate produced at the 3’ end of the 5’ fragment by the DNA enzyme , the intrinsic 3'-phosphatase activity at low ATP concentration of T4 polynucleotide kinase ( T4 PNK ) was used [26] . The reaction was supplemented with 2 unites/μl T4 PNK , PNK buffer and 0 . 4 mM ATP and incubated for 1 h at 37°C [26] . The RNA digestion fragments were gel-purified as described for in vitro transcriptions ( S1 Text ) . For the production of site-specifically labeled pre-mRNAs , the purified 3’ pre-mRNA fragment created by DNA enzyme cleavage was 5’-phosphorylated with 2 μM γ-32P ATP , T4-PNK buffer , 2 units/μl RNAsin and 1 unit/μl T4-PNK in a total volume of 20 μl or more , depending on the experiment . The reaction mixture was incubated for 1 h at 37°C and then purified by using a G 50 spin column , followed by phenol-chloroform-isoamyl alcohol ( PCI ) extraction and ethanol precipitation . RNA fragments were ligated by DNA splint-directed RNA ligation [22] . The 5’-ligation fragments were prepared by DNA enzyme cleavage followed by 3’-dephosphorylation as described above . For site-specific labeling , the 3’ ligation fragment was labeled at the 5’ end with γ-32P ATP as described above . For region-specific labeling , the 3’ fragment was produced by radioactive in vitro transcription using GMP as a starting nucleotide ( S1 Text ) . The 5’ ligation fragment , the DNA splint and the 3’ ligation fragment were mixed in a 1 . 4:1 . 2:1 ratio . After addition of T4 DNA ligase buffer and water , the reaction was denatured for 2 min at 70°C and the sample was then cooled to 30°C at 6°C per min . Thereafter 1 mM ATP , 2 units/μl RNAsin and 3 units/μl T4 DNA ligase were added and the reaction was incubated for 3 hrs at 30°C . Finally , the ligation product was gel-purified . The efficiency of ligation was ~30–60% .
|
The spliceosome is a highly dynamic RNP machine that during the catalytic cycle undergoes many changes in composition and conformation . The pre-catalytic Bact spliceosome contains the U2 , U6 and U5 snRNAs and ~40 proteins , which are evolutionarily conserved between budding yeast and metazoans . The Bact spliceosome is converted to a catalytically-activated B* spliceosome and following recruitment of the Cwc25 protein , step 1 of splicing is catalyzed and the C spliceosome is generated . The U2 snRNP plays an essential role in branch-site selection and pre-mRNA splicing catalysis . During the Bact to B* transition the affinity of several U2 SF3a/b proteins for the spliceosome is significantly reduced . Whether this is due to remodeling events affecting U2 snRNP contacts with the pre-mRNA is not known . Information about conserved spliceosomal protein-pre-mRNA contacts and their dynamics during splicing remains limited . Here we investigated pre-mRNA–protein contact sites in yeast Bact spliceosomes by UV-induced crosslinking . We detected contacts of nucleotides surrounding the branch-site with several of the U2 SF3a/b proteins , and we show that these interactions are evolutionarily conserved . We carried out a similar investigation with B* and C spliceosomes and provide important insights into the dynamics of pre-mRNA–protein interactions involving the essential U2 , RES , Cwc25 , Prp8 and Prp45 proteins .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Dynamic Contacts of U2, RES, Cwc25, Prp8 and Prp45 Proteins with the Pre-mRNA Branch-Site and 3' Splice Site during Catalytic Activation and Step 1 Catalysis in Yeast Spliceosomes
|
HTLV-1/2 are transmitted sexually , by whole cell blood products and from mother-to-child ( MTC ) , mainly through breastfeeding . HTLV-1/2 prevalence in pregnant women is high in Rio de Janeiro , however there were no local studies addressing the rate of adverse pregnancy outcomes ( APO ) and MTC transmission . The aim was to study sociodemographic characteristics which may be associated to HTLV-1/2 infection and describe pregnancy outcomes and MTC transmission in HTLV-1/2-positive women . The cross-sectional study screened 1 , 628 pregnant women in of Rio de Janeiro ( 2012–2014 ) and found 12 asymptomatic carrier mothers ( prevalence = 0 . 74% ) . Pregnancy outcome information was retrieved from medical records . Sociodemographic characteristics were similar between the positive and negative groups except for maternal age , which was higher in carrier mothers . The incidence of adverse pregnancy outcomes was similar in infected and non-infected patients ( p = 0 . 33 ) , however there was a high rate of premature rupture of membranes ( PROM ) amid infected mothers ( 3/12 ) . Multilevel logistic regression found that for each additional year of age , the chance of being HTLV-1/2-positive increased 11% and that having another sexually transmitted infection ( STI ) increased 9 times the chance of being infected . Carrier mothers had more antenatal visits ( OR = 5 . 26 ) . Among the children of HTLV-1/2-positive mothers there was one fetal death , one infant death and one loss of follow-up . After two years of follow-up there was one case of MTC transmission ( 1/9 ) . The mother reported breastfeeding for one month only . Knowledge about factors associated to HTLV-1/2 infection , its impact on pregnancy outcomes and the MTC transmission rate is important to guide public health policies on antenatal screening and management .
Human T-lymphotropic virus types 1 and 2 ( HTLV-1/2 ) are human oncogenic retroviruses first identified in the early 1980’s [1] . There are six subtypes of HTLV-1 ( A to F ) , which have no impact on the clinical expression of the disease [2] . There are two other types of HTLV ( 3 and 4 ) ; but there is no evidence of their pathogenicity in humans [3] . HTLV-1/2 viruses are globally distributed and there may be up to 10 million infected worldwide [4] . Prevalence is characterized by endemic clusters occurring next to low prevalence areas . It also varies considerably according to the ethnical and social background of the population . Since transmission occurs through infected body fluids , intravenous ( IV ) drug users and sex workers have been reported as high-risk groups [4] . The association between low social and economic level and lower education is not homogeneous among studies and most likely represents a bias . Endemic HTLV-1 clusters are found in Sub-Saharan Africa , South-western Japan , Central and South America as well as the Middle East and Melanesia [4] . Regardless of the area , seroprevalence increases with age , particularly in women due the excess efficiency of the male-female sexual transmission . HTLV-2 is endemic in Pygmy tribes of Central Africa and in several Native American populations , particularly in the Amazon area [5 , 6] . It is also frequent in IV drug users , often in co-infection with HIV [5 , 7] . Brazil may be the country with the highest absolute number of HTLV-1/2 carriers in the world . Estimates range from 800 , 000 to 2 . 5 million people [5 , 8 , 9] . Such variation in numbers can be explained both by the epidemiological characteristic of the infection and by the lack of data , with large areas of the country unmapped . Infection is perennial and most of the patients are asymptomatic reservoirs , sustaining the chain of transmission . In contrast , up to 8% of HTLV-1 carriers develop severe diseases , mainly the highly aggressive adult T-cell leukaemia/lymphoma ( ATLL ) and the painful and disabling HTLV-1-associated myelopathy/Tropical Spastic Paraparesis . Type 1 virus also causes a spectrum of inflammatory conditions , such as dermatitis and uveitis [10] . In turn , HTLV-2 has been associated to erythrodermatitis , neurologic disorders and opportunistic infections [7] . Literature about the effect of HTLV-1/2 infection on pregnancy outcomes is scarce . Only one study on the subject was published . It was conducted in Africa , between 1986 and 1988 , involving 45 HTLV-1/2 positive pregnant women and 90 negative ones . No statistically significant differences were found between the groups regarding neither sociodemographic profile , pregnancy and neonatal outcomes [11] . Only four Brazilian researches on HTLV-1/2 in pregnant women and puerperae assessed previous obstetric history and pregnancy outcome , and they reported only on miscarriage [12–15] . Large regional studies which altogether included over 130 , 000 pregnant women report miscarriage rates between 22% and 30% , notably higher than that observed in non-infected patients [13–15] . On the other hand , the only research conducted on an endemic area found a pregnancy loss rate of 10% , similar to the general population [12] . Dal Fabbro’s study was the only to report the frequency of two or more previous miscarriages , which was 0 . 8% in HTLV-1/2 infected women . This number is equivalent to the incidence of recurrent pregnancy losses in the general population [13] . HTLV-1/2 is transmitted via whole cell containing body fluids , mainly through sexual contact , exposure to blood products or viscera and from mother-to-child ( MTC ) . The relative importance of each mode of transmission is still largely unknown and most likely it varies with the population involved . In endemic areas such as Japan MTC transmission has been described as the main source of transmission , mainly through breastfeeding [12 , 16] . Only 2 . 5–5 . 0% of children are seroconverted in the absence of breastfeeding while up to 25% are infected if breastfed for over 12 months [16–19] . In fact , a recent Brazilian study found a vertical transmission rate of 50% in children who were breastfed for over 24 months [17] . This research also detected an increased risk of infection in siblings , confirming the trend for familial clustering of the disease [17] . Higher proviral load and antibody titers in maternal blood and breastmilk are also associated with increased MTC transmission rate [17 , 20–22] . On the other hand , peripartum transmission has been shown to have little impact on the burden of disease [19 , 23] . The aim of the research was to describe the epidemiological profile of pregnant women diagnosed with HTLV-1/2 in the metropolitan area of Rio de Janeiro , the occurrence of adverse pregnancy outcomes ( APO ) and the rate of mother-to-child transmission .
The study population consisted of 1 , 628 pregnant women . The first 1 , 204 were enrolled at admission for delivery as part of a research on HTLV-1/2 prevalence conducted at two public hospitals in the metropolitan area of Rio de Janeiro: the ‘Pedro Ernesto’ University Hospital of the Rio de Janeiro State University ( Universidade do Estado do Rio de Janeiro–HUPE/UERJ ) and the ‘Hospital Estadual da Mãe’ ( HEM ) . HUPE is a referral centre for high-risk patients while HEM , situated at the adjacent city of Mesquita , assists low and medium-complexity cases [24] . As a result of the relevant prevalence found in the local population ( 0 . 66% ) at the first part of the study , routine HTLV-1/2 screening was instituted in HUPE’s perinatal unit in July 2013 . The other 424 women were recruited at their first antenatal visit . During the first part of the study , roughly 24% of the women attending the HEM were recruited while the uptake at UERJ was 40% . During the antenatal screening period at UERJ only four approached women refused to take part in the study . The study was conducted between November 2012 and December 2017 . Subjects were recruited until 2014 and the children of HTLV-1/2 positive women were followed for three years . Women who were mentally unable to give consent or who declined to take part in the research were excluded . A structured questionnaire with socio-epidemiological , clinical and reproductive data was applied at recruitment . Women found to be HTLV-1/2 positive were counseled by a multidisciplinary team which provided health information and psychosocial support . They were advised about the risk of vertical transmission though breastfeeding and formula milk was provided to safeguard the infants’ nutrition . Children of carrier mothers were monitored for at least three years after the birth on the paediatric infectology department of HUPE . There were two exceptions: one who was followed up at her local health care center and another who was lost to follow up . Blood for HTLV-1/2 screening was collected either during the routine antenatal care or at the admission for delivery by chemiluminescent microparticle immunoassay ( CMIA—Architect rHTLV-I/II , Abbott ) . Children’s samples were also screened by CMIA within the month after birth , at six months , one and two years of age . Infection was confirmed if the child remained seropositive after 24 months , and additional yearly exams were performed for follow-up . Reactive samples were confirmed by Western blot ( WB , Inno Lia HTLV-I/II score Biomerieux ) . Two pregnant women with positive screening tests and negative WB results were considered false positive and allocated to the negative group in the statistical analysis . Routine antenatal screening for sexually transmitted infections ( STI ) in Brazil consists of VDRL test for syphilis and ELISA tests for HIV , hepatitis B and C . In case of positive screening the confirmation tests are FTA-Abs for syphilis and western blot for HIV . All tests were done at HUPE/UERJ’s Clinical Analysis Laboratory . This research complies with the Declaration of Helsinki and the Resolution 466 of December 12 , 2012 of the Brazilian Ministry of Health . The project was approved by the Rio de Janeiro State University Research and Ethics Committee ( COEP-UERJ , process 034 . 3 . 2012 ) and sponsored by the Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro ( FAPERJ , E-26/110 . 351/2012 ) . Written informed consent was obtained from all the subjects and from the legally authorized representatives of the minors who agreed to take part in the research . Anonymity and data confidentiality were guaranteed . Means , medians , standard deviations and percentages were used to describe the results . Medians were used to define the cutoff point used to convert numerical variables to categorical ones ( age , family income , number of partners and number of pregnancies ) . The following variables were considered as adverse pregnancy outcomes: fetal demise , perinatal death , maternal hypertensive syndrome , preterm birth and admission to the neonatal intensive care unit . A composite variable named ‘adverse outcome’ was created in view of the low frequency of each adverse obstetric outcome studied . Missing data were excluded from the statistical analyses . Mann-Whitney and Fisher’s exact tests were used to compare categorical variables between the HTLV-1/2 positive ( G1 ) and negative ( G2 ) groups . Multilevel logistic regression was performed to assess the factors associated to HTLV-1/2 infection . Age , number of pregnancies , coinfection with other STI , condom use in pregnancy and number of antenatal appointments were included in the model as fixed effects and "Hospital" was included as random effects ( random intercept ) . The Epi-info software version 3 . 5 . 2 and the R-Project version 3 . 3 . 1 were used for building the database and performing the statistical analyses .
Prevalence of HTLV-1/2 infection in this study population was 0 . 74% ( 12/1628 ) , with no significant difference between the hospitals ( 0 . 82% in HUPE , 0 . 67% in Mesquita; p = 0 . 78 ) . Among the sociodemographic characteristics ( Table 1 ) only maternal age was significantly different between the HTLV positive and negative groups . The age was over 24 years in 83% ( n = 10 ) of the HTLV+ group ( G1 ) and in 53% of G2 ( p = 0 . 03 ) . Most women in both groups reported being non-white ( p = 0 . 99 ) ; having at least 10 years of formal education ( p = 0 . 99 ) and being in a stable marital relationship ( p = 0 . 74 ) . About half the patients in both groups had monthly household income higher than two minimum wages ( p = 0 . 99 ) . No subjects reported behavioural risk factors like the use of IV drugs , having multiple sexual partners and being a sex worker . Sexual and reproductive characteristics are summarized in Table 2 . HTLV-1/2 positive women ( G1 ) were more frequently coinfected with another STI than patients in G2 ( 41 . 7% vs . 8 . 9% , p< 0 . 01 ) . Three of the HTLV-1/2 positive women had syphilis , one had HIV and another had histopathologically confirmed HPV condilomata . Most women in both groups had their first antenatal visit before 12 weeks of pregnancy ( 66 . 6% and 59 . 2% , p = 0 . 46 ) . The number of antenatal appointments was higher in the HTLV-1/2 positive group , even after the multivariate analysis ( 81 . 8% vs . 52% , p = 0 . 03 ) . Regarding the obstetric history , around 2/3 of infected women had more than two previous pregnancies ( 66 . 7% ) , while the proportion was inverse in G2 ( 34 . 9% ) . Number of sexual partners , previous reproductive history and frequency of condom use were similar in both groups . The prevalence of comorbidities was high in the study population ( 37 . 6% ) , being significantly more frequent in HUPE ( 78 . 1% ) than in HEM ( 4 . 8% , p<0 . 001 ) . However , the rate was similar between the infected and non-infected groups ( p = 0 . 16 ) . There were five adverse pregnancy outcomes in the HTLV positive group . The difference was not significant when compared to the negative group ( p = 0 . 33 ) ( Table 3 ) . There were three cases of coinfection with syphilis ( 3/12 ) and three of premature rupture of membranes ( PROM -3/12 ) at term . One patient with PROM also had syphilis , but the other two cases had no other risk factors ( 2/12; 16 . 6% ) . The prevalence of PROM in the HTLV-1/2 negative group was 4 . 7% ( 71/1 , 518 ) . Although the difference seems significant , it was not confirmed by Fisher’s exact test ( p = 0 . 26 ) , probably due to the small number of cases . One HTLV-1/2-infected newborn was admitted to the neonatal intensive care unit ( NICU ) due to prolonged PROM . Among non-infected babies , 137 were admitted to the NICU ( 8 . 3% vs . 10 . 0% , p = 0 . 66 ) . There were missing data on 210 seronegative pregnancy outcomes . Differences in maternal age , STI coinfection and the number of antenatal visits remained significant after multivariate logistic regression ( Table 4 ) . Multilevel Logistic regression model–dependent variable: HLTV; fixed effects: age , number of pregnancies , coinfection with other STI , condom use in pregnancy , number of antenatal appointments; random effects: hospital . Each additional year of maternal age increased the chance of being HTLV-1/2 positive in 11% ( OR = 1 . 11 ) . Having another STI increased 9 times the chance of being infected ( OR = 9 . 27 ) . In G1 there was a higher frequency of antenatal visits ( OR = 5 . 31 ) ( Table 4 ) . Among the children of the 12 HTLV-1/2 infected mothers , there was one fetal demise , one infant death and one loss of follow-up . The fetal death occurred at 24 weeks of pregnancy and the infant died at two months of age due to pneumonia . One child had its seroconversion ( 1/9 ) confirmed after two years of age . She was born at January 2013 and remains asymptomatic under medical surveillance at the HUPE to date . Her mother referred breastfeeding for less than one month . The other eight children were periodically monitored , seven at the pediatric infectology department of UERJ and one at its local health care center . Despite their economic difficulty , their mothers reported avoiding breastfeeding since they were aware of their carrier status .
The HTLV-1/2 prevalence found was consistent with the study previously published by our group which recruited only women admitted for delivery [24] . The sociodemographic profile of both groups was similar except for the older age found in the infected group . This is in accordance with the international literature on the disease’s epidemiology [4] and with the studies performed in Brazilian areas with high prevalence of HTLV-1/2 [17 , 25 , 26] . On the other hand , in a large research performed in Gabon , where HTLV-1/2 prevalence is over 10% , there was no sociodemographic difference between the groups [11] . The multivariate analysis of sexual and reproductive characteristics found two significant differences between the infected and non-infected groups: STI coinfection and number of antenatal visits . The increased frequency of STI coinfection in HTLV-1/2 carriers , particularly syphilis , was consistent with studies from endemic areas [12 , 14 , 27 , 28] , although it’s not a universal finding across publications [11 , 13 , 15 , 26 , 29] . On the Gabon research there were twice as many cases coinfected with syphilis than controls , however statistical significance was not reached , probably because insufficient sample size [11] . The latest Salvador study found that 21 . 5% of HTLV-1/2 infected subjects also had syphilis ( OR = 36 . 7 ) [28] . As for the higher number of antenatal visits in the HTLV-1/2 positive group , it cannot be explained by the knowledge of the carrier status itself since only one patient was aware of the infection at the beginning of the pregnancy . There was also no significant correlation with the presence of comorbidities ( p = 0 . 38 ) . The hypothesis of the more frequent antenatal visits being due to these women having more previous adverse pregnancy outcomes seemed significant ( p = 0 . 04 ) , but its confidence interval was too wide ( 0 . 91–10 . 2 ) . It is true , though , that the lack of significance could be caused by the small number of cases . A previous case-control study which addressed this variable also failed to find any difference between infected and non-infected patients [11] . Regarding the reproductive history , it’s striking that almost half of the carrier mothers ( 5/12 ) had previous adverse obstetric outcomes ( three first trimester miscarriages , one fetal demise and one FGR with neonatal death ) . Three of these women had no comorbidities . It must be said however , that no causal link to HTLV-1/2 can be inferred since their infectious status was unknown during the previous pregnancies . The prevalence of early miscarriages was similar ( circa 21% ) among HTLV-1/2-positive and negative patients ( p = 0 . 86 ) . This was in accordance with data from other endemic areas [11 , 12] . Regarding the adverse pregnancy outcomes studied , no difference was observed between infected and non-infected patients . It’s important to stress that this finding cannot be generalized in view of the small sample size and the low incidence of the outcomes . The two studies reporting on obstetric results of HTLV-1/2 infected women did not find association between the infection and adverse pregnancy outcomes as well [11 , 12] . The Gabon research found a trend for preterm delivery and complicated pregnancies in HTLV-1/2 positive women [11] . Unfortunately , even this study , which followed 45 infected patients and 90 controls , was underpowered for this statistical analysis . There were no cases of preterm delivery or low birth weight among HTLV-1/2 positive patients , in accordance with the findings of Bittencourt et al [12] . There was one case of fetal growth restriction and intrauterine demise in an otherwise healthy HTLV-1 infected mother . This woman reported two previous adverse pregnancy outcomes , but it’s unknown whether she was already infected at the time . At the Salvador case-control study[12] , there were also fetal deaths on the HTLV-1/2 carrier mothers’ group , but those happened in patients with additional comorbidities such as hypertension and falciform anaemia . Additionally , that study reported three cases of hypertension in pregnancy and one admission to the NICU due to neonatal sepsis after PROM . Among the infected mothers there were three cases of term PROM ( 25% ) . In a recent Brazilian study using data from the Ministry of Health [30] , PROM was found to complicate approximately 4 . 2% of all livebirths in the country . This number is in line with the prevalence found in our HTLV-1/2 negative group . Uterine inflammation and sexually transmitted infections have been shown to be associated with obstetric complications such as PROM [31–34] . HTLV-1/2 infection is also known to be linked to different inflammatory and infectious manifestations . Thus , it seems reasonable to interrogate whether HTLV-1/2 infection may increase the risk of PROM . Unfortunately , the largest study on HTLV-1/2 pregnancy outcomes , performed in Gabon , didn’t assess the incidence of PROM [11] . This finding prompts the need for further research , adequately powered to elucidate the matter . In our study , the mother of the only infected child reported breastfeeding for less than one month . Mother-to-child transmission of HTLV-1/2 occurs mainly through breastfeeding , ranging from 3 . 9% to 22% in endemic areas [35] . The policy of universal HTLV-1/2 antenatal screening and contraindication of breastfeeding for infected mothers at the Nagasaki province reduced the local MTC transmission rate from 20 . 3% to 2 . 5% [16] , which are the known seroconversion rates of prolonged breastfeeding ( > 6 months ) and exclusive bottle-feeding , respectively . However , even with shorter periods of breastfeeding the MTC transmission rate is greater than using only infant formula ( 7 . 4% vs . 2 . 5% ) [22] . Other possible reasons for the MTC transmission in this case are peripartum infection or additional risk factors such as: high maternal antigenemia , concentration of gp46 HTLV-1/2 antibodies , the presence of anti-Tax antibodies , or the human leucocyte antigen system ( HLA ) type concordance between mother and child [17 , 18 , 35 , 36 , 37 , 38 , 39] . The hypothesis of peripartum infection seems unlikely since the child was delivered by caesarean section due to hypertensive syndrome without PROM or labour . A limitation of this study is that it could not assess the other variables mentioned , such as proviral load and HLA type . Other two Brazilian studies report MTC transmission after less than a month of breastfeeding . In both cases the mothers’ proviral loads were extremely high [17 , 37] . Another limitation of the study was the small sample of infected patients and the low frequency of adverse pregnancy outcomes , which were grouped for the statistical analysis . A study which is properly powered for statistical analysis on this matter would require a much greater sample size , and that may prove impeditive in areas of intermediate prevalence such as ours . On the other hand , a strength of the study is that the children of infected mothers were followed up for three years , a gold standard set by the Nagazaki study group [16] . This was proposed since some cases of seropositivity in children are caused by maternal antibodies , which generally disappear after 12 months of life . Our MTC transmission rate was 1/9 ( 11% ) , similar to the ones found in Haiti and Guyana [11 , 35] . The study confirmed the high prevalence of HTLV-1/2 in pregnant women at the metropolitan area of Rio de Janeiro and found no sociodemographic difference between infected and non-infected patients . Carrier mothers frequently reported previous adverse pregnancy outcomes ( 5/12 ) , but at the current pregnancy there was only one unexplained fetal demise ( growth restricted ) and one admission to the NICU due to sepsis . There was a significant association to other STI , but the intriguing point was the number of PROM cases among infected women ( 3/12 ) . Since there is no treatment or immunization for HTLV-1/2 , preventive measures are currently the only effective way to break the chain of transmission . Thus , it is vital to increase awareness about the infection among health providers and the population . Safe sex campaigns are already extensive , therefore tackling MTC transmission is likely to have the most significant effect on the longitudinal perpetuation of the virus and the reduction of HTLV-1/2 associated diseases , particularly in endemic areas . There are no studies on antiretroviral therapy or mode of delivery to address the potential for reducing MTC transmission . Thus , avoidance of breastfeeding remains the only effective way to block the MTC transmission of the virus . The use of this strategy as a public policy in low income areas is still not a consensus , since breastfeeding also plays a role in reducing infant mortality and morbidity through immunity boosting and protecting against infections . Since there is no clear difference in the sociodemographic profile of HTLV-1/2 carriers , routine prenatal screening at endemic areas is supported by several research groups [9 , 10 , 13 , 15–17 , 21 , 25 , 27 , 28 , 30 , 31 , 40–42] . Considering the number of livebirths at the metropolitan area of Rio de Janeiro ( 236 , 960 LB in 2016 ) and the HTLV-1/2 seroprevalence in pregnant women found in this study , the introduction of local routine antenatal screening could avoid over one thousand cases of MTC transmission in a year[41] . A similar estimate was found by a recent epidemiological study [42] . Further studies are needed on the cost-effectiveness of such strategy across different prevalence areas and socioeconomic resources . Moreover , epidemiological mapping of prevalence and mother-to-child transmission is needed to guide public health policies on antenatal screening and management . Additionally , better understanding of the burden of the infection in pregnancy may help to improve HTLV-1/2 infected mothers’ antenatal care and their children’s outcome .
|
HTLV-1/2 are retroviruses transmitted by sex , blood products and from mother to child ( MTC ) , mainly through breastfeeding . There is currently no vaccine , treatment or cure . Although it’s mostly asymptomatic it can cause disabling and even lethal diseases in carriers . The prevalence of HTLV-1/2 in pregnant women at the metropolitan area of Rio de Janeiro is high ( 0 . 74% ) . Our aim was to study the sociodemographic characteristics which may be associated to HTLV-1/2 infection and describe pregnancy outcomes and MTC transmission in the infected population . Apart from being slightly older , there were no differences in the carrier mothers’ sociodemographic profile . Pregnant women with sexually transmitted infections had a 9-fold chance of being HTLV-1/2 positive . Although adverse pregnancy outcomes were not increased , infected mothers had a high rate of ruptured membranes . Among the children of HTLV-1/2-positive mothers there was one fetal death , one infant death and one loss of follow-up . There was one case of MTC transmission ( 1/9 ) , after one month of breastfeeding . Knowledge about factors associated to HTLV-1/2 infection , its impact on pregnancy , and the MTC transmission rate is important to guide further research and public health policies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"neonatology",
"urology",
"medicine",
"and",
"health",
"sciences",
"maternal",
"health",
"obstetrics",
"and",
"gynecology",
"geographical",
"locations",
"tropical",
"diseases",
"pediatrics",
"treponematoses",
"bacterial",
"diseases",
"women's",
"health",
"sexually",
"transmitted",
"diseases",
"pregnancy",
"neglected",
"tropical",
"diseases",
"pregnancy",
"complications",
"families",
"infectious",
"diseases",
"south",
"america",
"antenatal",
"care",
"fetal",
"death",
"brazil",
"people",
"and",
"places",
"breast",
"feeding",
"genitourinary",
"infections",
"co-infections",
"population",
"groupings",
"mothers",
"syphilis"
] |
2019
|
Pregnancy outcomes and mother-to-child transmission rate in HTLV-1/2 infected women attending two public hospitals in the metropolitan area of Rio de Janeiro
|
The most widespread measures of human brain activity are the blood-oxygen-level dependent ( BOLD ) signal and surface field potential . Prior studies report a variety of relationships between these signals . To develop an understanding of how to interpret these signals and the relationship between them , we developed a model of ( a ) neuronal population responses and ( b ) transformations from neuronal responses into the functional magnetic resonance imaging ( fMRI ) BOLD signal and electrocorticographic ( ECoG ) field potential . Rather than seeking a transformation between the two measures directly , this approach interprets each measure with respect to the underlying neuronal population responses . This model accounts for the relationship between BOLD and ECoG data from human visual cortex in V1 , V2 , and V3 , with the model predictions and data matching in three ways: across stimuli , the BOLD amplitude and ECoG broadband power were positively correlated , the BOLD amplitude and alpha power ( 8–13 Hz ) were negatively correlated , and the BOLD amplitude and narrowband gamma power ( 30–80 Hz ) were uncorrelated . The two measures provide complementary information about human brain activity , and we infer that features of the field potential that are uncorrelated with BOLD arise largely from changes in synchrony , rather than level , of neuronal activity .
Most measurements of activity in the living human brain arise from the responses of large populations of neurons , spanning the millimeter scale of functional magnetic resonance imaging ( fMRI ) and electrocorticography ( ECoG ) to the centimeter scale of electro- and magneto-encephalography ( EEG and MEG ) . Integrating results across methods is challenging because the signals measured by these instruments differ in spatial and temporal sensitivity , as well as in the manner by which they combine the underlying neuronal population activity [1–3] . Differences in scale can be partially bridged by bringing the measurements into register . For example , EEG and MEG sensor data can be projected to cortical sources subject to constraints from simultaneously recorded fMRI data [4] or from independent fMRI localizers [5] , and ECoG electrodes can be aligned to a high-resolution anatomical MRI image [6] and compared to the local fMRI signal . Yet , even when electrophysiological and fMRI data are spatially registered , striking differences in the sensitivity to stimulus and task are often observed , indicating differences in how neuronal responses contribute to the measured physiological signals . For example , the fMRI blood-oxygen-level dependent ( BOLD ) signal and EEG evoked potentials differ in which brain areas are most sensitive to visual motion ( area MT+ with fMRI [7] versus V1 and V3A with EEG [8] ) . Within the same visual area , fMRI and source-localized EEG evoked potentials can show different effects of task in similar experimental paradigms , such as the effect of spatial attention on the contrast response function ( additive in fMRI [9] , multiplicative in EEG [10] ) . Even when the spatial scale of the two signals is approximately matched at acquisition , such as ECoG electrodes and fMRI voxels ( both at approximately 2 mm ) , systematically different patterns of responses can be obtained , such as compressive spatial summation in fMRI versus nearly linear summation in ECoG steady-state potentials ( but not ECoG broadband signals ) [11] . Such fundamental functional differences cannot be explained by numerical measurement-to-measurement transformations . Rather , these differences must reflect the fact that the measurements are based on different aspects of the neural population response . To explain the differences in measurement modalities requires a computational framework that derives each of these signals from the neuronal responses . One approach toward developing such a framework has been to measure the BOLD signal and electrophysiological signals simultaneously , or separately but using the same stimulus and task conditions , and to ask how features of the electrophysiological response compare to the BOLD signal . This approach has revealed important patterns , yet after several decades of careful study , some apparent discrepancies remain . A number of studies comparing band-limited power in field potential recordings to the BOLD signal have shown that increases in power between 30 Hz and 100 Hz ( gamma band ) are more highly correlated with BOLD amplitude than power changes in other bands [12–17] . Yet , power changes in this band do not fully account for the BOLD signal: very large power changes can occur in the gamma band without a measurable BOLD signal change [18 , 19] , and power changes in lower frequency bands can be correlated with the BOLD signal independently of power changes in the gamma band [20–23] . It therefore cannot be the case that field potential power in the gamma band is a general predictor of BOLD , even if the two measures are often correlated . Another source of disagreement is that within the gamma band , some reports claim that BOLD is best predicted by synchronous ( narrowband ) signals [13] , and others claim that BOLD is best predicted by asynchronous ( broadband ) neural signals [11] . Moreover , in some cases , it has been reported that no feature of the local field potential ( LFP ) predicts the intrinsic optical imaging signal ( closely related to BOLD ) as accurately as multiunit spiking activity [24] . Consistent with this claim , a comparison of both motion and contrast response functions measured with single units and with BOLD suggested a tight coupling between BOLD and single-unit responses [25–27] . To our knowledge , there is currently no single model linking the electrophysiological and BOLD signals that accounts for the wide range of empirical results . The numerous studies correlating features of electrophysiological signals with BOLD provide constraints in interpreting the relationship between the two types of signals , yet the approach has not led to a general , computational solution . We argue that one reason that correlation studies have not led to computational solutions is that any particular feature of the field potential could be caused by many possible neuronal population responses . For example , a flat field potential ( minimal signal ) could arise because there is little activity in the local neuronal population or it could arise from a pair of neuronal subpopulations responding vigorously but in counterphase , resulting in cancellation in the field potential . The same field potential in the two situations would be accompanied by different levels of metabolic demand and presumably different levels of BOLD signal . Similarly , any particular BOLD measurement could be due to many different patterns of neural activity . For example , stimulation of a neuronal population that inhibits local spiking can cause an elevation in the BOLD signal [28] , as can stimulation of an excitatory population that increases the local spike rate [29] . In short , there can be no single transfer function that predicts the BOLD signal from the field potential because the field potential does not cause the BOLD signal; rather , the neuronal activity gives rise to both the field potential and the BOLD signal . We propose that many of the different claims pertaining to the relationship between BOLD amplitude and features of the field potential can be accounted for by a modeling framework in which BOLD and field potential measurements are predicted from simulated neuronal population activity , rather than by predicting the BOLD signal directly from the field potential . In this paper , we model fMRI and ECoG responses in two stages , one stage in which we simulate activity in a population of neurons , and a second stage in which we model the transformation from the population activity to the instrument measures . By design , the model employs a minimal set of principles governing how the instruments pool neuronal activity , rather than a biophysically detailed description of neuronal and hemodynamic events . This approach enables us to ask whether this minimal set of principles is sufficient to guide simulations of neuronal population activity , such that the parameters of the simulations are fit to ECoG measurements from human visual cortex , and the output of the simulations predicts fMRI BOLD responses in the same regions for the same stimuli .
The fMRI BOLD signal and the LFP measure neuronal population activity in a fundamentally different manner . The goal of this analytic framework is to capture these differences in simple mathematical expressions and from these expressions , derive the relationship between the two instrument measurements . We purposely omit a large number of biophysical details , such as cell types , neuronal compartments , the dynamics of blood flow , and so forth , both for tractability and in order to emphasize the basic principles of how different measures integrate neuronal activity . In the sections that follow , we then show that , when coupled to simulated neural responses , the model can account for many important patterns observed in fMRI and ECoG data from human visual cortex . For this analytic framework , we consider how a population of n neurons responds to a stimulus or task during a brief epoch ( time 0 to T ) , assumed to be on the order of a second . Each neuron will produce a time-varying dendritic current , denoted as Ii ( t ) for the ith neuron , resulting from the trans-membrane potential . We would like to know how these currents , I ( t ) , relate to the fMRI BOLD signal and to the LFP signal measured by an ECoG electrode . We assume that the LFP arises primarily from dendritic membrane currents [2] . We ignore output spikes . ( Although spikes can influence the LFP [30] , it is generally thought that the influence is smaller than synaptic and dendritic currents [2] , and including spikes would not change the logic of our arguments . ) For the ith neuron , the contribution to the LFP is then αi × Ii ( t ) . The constant αi depends on the distance and orientation of the neuron with respect to the electrode , as well as the electrode’s impedance . For simplicity , we assume that each neuron is equidistant from the electrode and has the same orientation , like pyramidal neurons perpendicular to the cortical surface , and therefore its contribution to the electrode measurement is scaled by the same constant , α . These neurons act together like a single , equivalent circuit , and hence the LFP time series will sum the contribution from each neuron , LFP ( t ) =α∙∑inIi ( t ) ( Eq 1 ) Field potential recordings are often summarized as the power ( or band-limited power ) in the time series [31] . Here we summarize the LFP response within a short time window as the power in the signal summed over the time window T: LFPpower=α∙∫0T ( [∑inIi ( t ) ]2 ) dt]powerofsum ( Eq 2 ) Importantly , Eq 2 is a linear/nonlinear ( L/N ) computation , since the LFP power is computed by first summing the signals ( L ) , and then computing the power ( N ) . The BOLD signal pools neural activity in a fundamentally different manner because it depends on metabolic demand [e . g . , for reviews , see 1 , 32] . ( Recent work on the neurobiology of neurovascular coupling indicates that much of the BOLD signal is not caused directly by changes in the level of metabolic products such as glucose , but rather by signaling molecules that tend to correlate with metabolic demand [33] ) . For simplicity , we discuss the BOLD signal throughout in terms of metabolic demand and return to this issue explicitly in the Discussion , ( 3 . 6 The role of a simple model in understanding the relation between BOLD and LFP ) . The metabolic demand of each neuron will increase if the cell depolarizes ( excitation ) or hyperpolarizes ( inhibition ) [28 , 34 , 35] . Hence , the metabolic demand of a neuron is a nonlinear function of its membrane potential: either a positive or negative change in voltage relative to resting potential causes a current , thereby resulting in a positive metabolic demand . There are many possible nonlinear functions one could assume to summarize the metabolic demand from the dendritic time series , such as the rectified signal ( absolute value ) or the power ( squared signal ) . For tractability , we assume the metabolic demand of the ith neuron is proportional to the power in the time-varying trans-membrane current integrated over time: βi × ( POWER ( Ii ( t ) ) or βi×∫0T ( Ii ( t ) 2 ) dt , with βi a scaling constant for the ith neuron . ( Similar results were obtained if we used the absolute value rather than the power ) . For the entire population of n neurons , we then assume the BOLD signal will sum the metabolic demand of each neuron . For simplicity , we use the same scaling constant for each neuron: BOLD=β∙∑in ( ∫0T ( Ii ( t ) 2 ) dt ) sumofpower ( Eq 3 ) Importantly , Eq 3 is an N/L computation , since the power is computed first ( N ) and then the signals are summed ( L ) , opposite to the order of operations for the LFP in Eq 3 ( Fig 1 ) ( personal communication from David J Heeger ) . In other words , we approximate the BOLD signal as the sum of the power , and LFP as the power of the sum , of the separate neuronal time series . The difference in the order of operations can have a profound effect on the predicted signals , as in the simple example with two neurons depicted in Fig 1C and 1D . The BOLD signal pooled over the two neurons is the same whether the time series from the two neurons are in phase or out of phase , whereas the LFP power is large when the time series are in phase and small when they are out of phase . These approximations allow us to make predictions about the relation between LFP and BOLD . By theorem , we know that the power of the sum of several time series is exactly equal to the sum of the power of each time series plus the sum of the cross-power between the different time series ( Eq 4 ) : ∫0T ( [∑inXi ( t ) ]2 ) dt=∑in ( ∫0T ( Xi ( t ) 2 ) dt ) +∑i≠jn ( ∫0T ( Xi ( t ) ∙Xj ( t ) ) dt ) PowerofsumSumofpowerSumofcross-power ( Eq 4 ) Applying this theorem to Eqs 2 and 3 shows the relationship between our models of BOLD and LFP power: LFPpower=αβ∙BOLD+α∙∑i≠jn ( ∫0T ( Ii ( t ) ∙Ij ( t ) ) dt ) ( Eq 5 ) We can now see that the LFP power depends on two quantities , one of which is related to the BOLD signal , and one of which is unrelated to the BOLD signal ( Eq 5 ) . The first quantity summarizes the total level of neural activity ( summed across neurons ) , and the second quantity summarizes the relationship between neural time series ( the cross-power , similar to covariance ) . If and when the second term tends to be large compared to the first , then the LFP power will not be closely related to the BOLD signal . One cannot deduce from first principles whether the first term in Eq 4 ( summed power ) or the second term ( summed cross-power ) will dominate . However , the number of elements contributing to the two terms is quite different: For n neurons , the first term has n numbers ( the power in each neuron’s time series ) , whereas the second term has nearly n2 numbers ( all the pairwise cross-powers ) . Hence , if there is any appreciable covariance , then the LFP power will be dominated by the second term , and the correlation with BOLD will be weak ( except in cases where the cross-power and power are highly correlated ) . To see how these equations translate to quantitative measures of BOLD and LFP , we consider a small neuronal population whose time series conform to a multivariate Gaussian distribution . We assume that each neuron’s time series has the same mean , m; the same variance , σ2; and all of the pairwise correlations have the same value , ρ: X∼N ( μ , Σ ) μ= ( m⋮m ) Σ=[σ2⋯σ2ρ⋮⋱⋮σ2ρ⋯σ2] ( Eq 6 ) X is the population time series , μ is the mean of each time series , and ∑ is the covariance matrix . We can now rewrite the simulated BOLD signal ( the sum of the power ) and the LFP ( power of the sum ) in terms of the parameters of the multivariate Gaussian ( and arbitrary scaling factors α , β ) , BOLD=β∙[n∙ ( m2+σ2 ) ]LFPpower=α∙[n∙ ( m2+σ2 ) + ( n2−n ) ( m2+σ2ρ ) ] ( Eq 7 ) where n is the number of neurons . This enables us to visualize how the BOLD signal and the LFP power depend on just three values: the variance , correlation , and mean in the neural time series , rather than on all the individual time series ( Fig 2 ) . For these neuronal time series , the LFP , modeled as the power of the sum of neuronal time series ( panel A ) , is dominated by the neuronal cross-power ( panel C ) . The BOLD signal , modeled as the sum of the power in the neuronal time series ( panel B ) , makes little contribution to the LFP , except when the correlation between neurons is low ( ρ is close to 0 ) ; in this case , there is no cross-power , and BOLD and LFP power are correlated . In section 2 . 1 , we proposed formulae to derive instrument measures from neuronal population activity . Here , we ask how we might simulate neuronal activity with a small number of parameters . A low-dimensional characterization of the population activity is useful since we normally do not have access to the time series of an entire population of neurons . Moreover , a low-dimensional representation can lead to better understanding and generalization even when high-dimensional data are available [36 , 37] . After simulating the population activity , we then use the analytic framework from section 2 . 1 to compute the BOLD and LFP signals . The parameters for the simulations were fit to ECoG recordings from human V1 , V2 , and V3 [38] . Because there were recordings from multiple electrodes and multiple stimuli , we ran multiple simulations fit to the different ECoG responses . We then used these simulations to predict the BOLD signal and compared these predictions to the measured BOLD signal for the same stimuli and same cortical locations ( but in different observers ) . The steps for simulating the neuronal population data and the derived LFP and BOLD , and for comparing the simulations to empirical data , are summarized in Table 1 . In the ECoG experiments , there were four grating stimuli of different spatial frequencies , three noise patterns with different power spectra , and one blank stimulus ( mean luminance ) . For each of the 8 stimuli and each of 22 electrodes in V1 , V2 and V3 , we decomposed the measured ECoG responses into three spectral components: broadband , narrowband gamma , and alpha ( Fig 3 ) . An important feature of this dataset is that the three components of the ECoG responses showed different patterns across stimuli [38]: stimuli comprised of noise patterns caused large broadband increases but little to no measureable narrowband gamma response , whereas grating stimuli elicited both broadband increases and narrowband gamma increases . Gratings and noise stimuli both resulted in decreases in alpha power compared to baseline ( also see S1 Fig ) . Had all three responses been tightly correlated with each other , it would not be possible to infer how each relates separately to the BOLD signal . The simulations were structured to approximate the experimental design and the results of our ECoG experiments . To match the design of our ECoG experiments , a simulated experiment consisted of 240 trials , each of which were 1 second long ( 30 repeats of 8 conditions ) . The LFP time series were transformed to power spectra , which were averaged across the 30 repeated trials of the same condition . The simulation parameters—i . e . , the level of the three inputs , C1 , C2 , and C3—were fit to the measured ECoG summary metrics ( broadband , gamma , and alpha ) for each of the 8 conditions for a particular electrode ( Fig 6 ) . To verify the validity of this procedure , we asked whether the simulations using the fitted parameters produce simulated spectra , which , when analyzed like the ECoG spectra , reproduce the original values of broadband , gamma , and alpha . In other words , do we close the loop from measured spectral components ( broadband , gamma , and alpha ) to inferred input parameters ( C1 , C2 , C3 ) to simulated population activity , to simulated spectral components ( broadband , gamma and alpha ) ? The original values are not reproduced exactly because the simulations are stochastic , but overall , the original broadband , gamma , and alpha values are recovered with high accuracy ( S9 Fig ) . As described above , the fitting of the parameters for C1 , C2 , and C3 was constrained by the assumptions that for C1 , the correlation between neurons was 0 ( and the amplitude was varied for fitting ) ; for C2 , the amplitude was fixed at a nonzero value ( and the correlation was varied for fitting ) ; and for C3 , the correlation was fixed at a nonzero value ( and the amplitude was varied for fitting ) . Results from alternative models with different constraints show poorer fits and are described briefly below and more extensively in S7 Fig . Importantly , the parameter fits did not take into account the measured BOLD responses . Hence the simulations provided a test: if the input parameters were chosen to produce outputs that match the measured ECoG responses ( training data ) , does the simulated BOLD signal accurately predict the measured BOLD signal ( test data ) ? We measured BOLD responses in four healthy subjects to the same visual stimuli as used in ECoG ( subjects are different from the ECoG subjects ) and extracted the signal from regions of interest in visual cortex matched to the previously recorded ECoG electrode locations ( S2 Fig and S3 Fig ) . For an example V1 site , the predicted BOLD signal accurately matched the measured BOLD signal , with 89% of the variance in the measured BOLD signal explained by the prediction , as quantified by R2—the coefficient of determination ( Fig 7A ) . Across V1 sites , the predicted BOLD signal from the simulations accounted for a median of 80% of the variance in the measured data ( Fig 7C ) . For an example V2 site , the predicted BOLD signal also matched the measured BOLD signal ( R2 = 0 . 74 , Fig 7B ) . Across V2/V3 sites the simulations explained a median of 40% of the variance in the data . The explained variance in V2/V3 is substantial but lower than in V1 . One likely reason for the higher variance explained in V1 is that for the particular stimuli used in these experiments ( gratings and noise patterns ) , the BOLD response reliability was higher in V1 . For example , the median R2 computed by using half the BOLD data as a predictor for the other half ( split half by subjects ) was 86% for V1 and 63% for V2/V3 . Similarly , the stimulus-evoked BOLD responses in V1 were larger than in V2 and V3 , with more stimulus-related variance to explain: a mean of 1 . 8% signal change in V1 versus 1 . 2% in V2 and 0 . 8% in V3 ( S4 Fig ) . It is possible that a stimulus set more tailored to extrastriate areas , such as textures or more naturalistic scenes , would have evoked more reliable responses in extrastriate cortex . For each of the 22 simulations , the three input parameters C1 , C2 , and C3 defining each of the 8 stimulus conditions were fit to produce the LFP data from the corresponding ECoG electrode . By design , the C1 ( broadband ) and C3 ( alpha ) inputs were fit to ECoG data by varying the level per neuron , whereas C2 was fit to data by varying the correlation across neurons . In principle , for any of the three inputs , the ECoG data could have been fit by varying either the level per neuron or correlation across neurons . For completeness , we tested all 8 combinations of models ( S7 Fig ) . The most accurate model , quantified as the R2 between the measured BOLD and the simulated BOLD ( median across the sites in V1 or in V2/V3 ) , was the simulation type used in the main text , in which C1 and C3 varied in the level per neuron and C2 varied in the correlation across neurons . Models in which the broadband correlation rather than level was used to fit the ECoG broadband power were much less accurate . The models in which the gamma LFP power was fit by modulating the level rather than the correlation in the simulated population caused a small drop in R2 . The previous analysis showed that when simulations were fit to ECoG data , the simulated BOLD response predicted the measured BOLD response . Here , we used regression analyses to assess how the simulated LFP predicted the simulated BOLD response and how the measured LFP predicted the measured BOLD response .
Many studies have reported correlations between BOLD and power in the gamma band LFP ( 30–130 Hz ) ( review for human studies: [52] ) . Yet , changes in gamma band power do not reflect a single biological mechanism . For example , several recent studies have emphasized that LFP power changes in the gamma band reflect multiple distinct neural sources , including narrowband oscillations and broadband power shifts , with very different stimulus selectivity and biological origins [38 , 53 , 54] . Broadband changes have been proposed to reflect , approximately , the total level of Poisson-distributed spiking ( or spike arrivals ) in a local patch of cortex [40] . In contrast , the narrowband gamma response is caused by neural activity with a high level of cell-to-cell synchrony [55] and likely depends on specialized circuitry [56] . While the two responses are sometimes distinguished as “high gamma” ( referring to broadband signals ) and “low gamma” ( referring to oscillatory signals ) , this distinction is not general . Broadband signals can extend into low frequencies [11 , 57] so that the two signals can overlap in frequency bands . Hence , separating gamma band field potentials into an oscillatory component and a broadband ( nonoscillatory ) component is not reliably accomplished by binning the signals into two temporal frequency bands , one low and one high , but rather requires a model-based analysis , such as fitting the spectrum as the sum of a baseline power law ( to capture the broadband component ) and a log-Gaussian bump ( to capture the oscillatory component ) [38] . There is strong experimental support for the idea that increases in broadband LFP power primarily reflect increases in asynchronous neural activity rather than increases in coherence . First , experiments have shown that broadband power is correlated with multiunit spiking activity [54 , 58] . Second , unlike the case of narrowband gamma LFP , changes in broadband LFP are not accompanied by changes in broadband spike-field coupling ( Figure 1A-B in [43] ) . The possibility that neuronal synchrony sometimes affects broadband signals cannot be ruled out , for example , as shown in cases of pharmacological manipulations in nonhuman primates [59] . In such cases , there would not be a simple relationship between broadband power and BOLD . The prior literature has not shown definitively whether broadband LFP , narrowband gamma , or both predict the BOLD signal . The first study that directly compared simultaneously recorded BOLD and electrophysiology showed that both LFP power in the gamma frequency range ( 40–130 Hz ) and multiunit spiking activity ( MUA ) predicted the BOLD signal [16] and further , that when the LFP power diverged from MUA , the gamma band LFP predicted the BOLD signal more accurately than did spiking . This study however did not separately test whether a narrowband ( oscillatory ) or a broadband ( nonoscillatory ) component of the LFP better predicted the BOLD response . Other studies , too , have shown a variety of patterns when correlating LFP power changes in the gamma band with BOLD . Some reported that BOLD amplitude correlates with narrowband gamma activity [13] , while others showed that BOLD correlates with broadband changes [11] , and many did not distinguish narrowband from broadband power in the gamma band [60] . Simultaneous recordings of hemodynamic and neuronal activity in macaque V1 showed that BOLD signals from intrinsic optical images can occur in the absence of gamma band LFP changes [61] and that , in some circumstances , multiunit activity predicts the BOLD signal more accurately than gamma band LFP [24 , 62] . Here , we separately quantified the broadband power ( spanning at least 50–150 Hz ) and narrowband gamma power . We found that the amplitude of broadband changes accurately predicted the BOLD signal in V1 . The empirical results and the models help resolve the question of why “high gamma” has been shown to correlate with BOLD , and “low gamma” sometimes does not [24] . The likely reason is unrelated to the difference in frequency range , nor to the size of the spectral perturbation in the LFP . In fact , the elevation in broadband power is relatively small ( 2- or 3-fold ) compared to the elevation in power often observed in narrowband gamma oscillations ( 10x or more ) [38] . Instead , “high gamma” is predictive of the BOLD signal in many cases not because of the specific frequency range , but because this signal captures the level of asynchronous neuronal response; this signal happens to be most clearly visible in the high-frequency range ( >100 Hz ) in which it is not masked by rhythmic lower-frequency responses . Hence , the distinction in predicting the BOLD response is not about “high” versus “low” gamma but rather synchronous versus asynchronous responses , and the broadband signal , sometimes labeled high gamma , maps onto the first term on the right-hand side of Eq 4 , the portion of the field potential measurement that sums the energy demand of each neuron . Our model fits and data support this view . When we captured the stimulus-related broadband response by simulating a change in broadband coherence across neurons rather than a change in the level of response in each neuron , our predicted BOLD response was highly inaccurate ( S7 Fig ) . In contrast , we propose that “low gamma” often does not predict the BOLD response because “low gamma” reflects narrowband oscillatory processes , which largely arise from a change in neuronal synchrony across the population rather than a change in the response level per neuron . This corresponds to the second term in the right-hand side of Eq 4 , the portion of the field potential measurement that reflects the cross-power arising from currents in different neurons , which in our model , is independent of the signals giving rise to the BOLD signal . Our results and model do not argue that narrowband gamma oscillations will never be predictive of the BOLD signal . If , in a particular experiment , narrowband gamma oscillations were to covary with broadband increases , we would expect both signals to correlate with BOLD . This might occur in an experiment with gratings of different contrast; with increasing contrast , narrowband gamma responses , broadband responses , and BOLD responses all increase [21 , 63] , and all three measures would correlate across stimuli . In such an experiment , if narrowband gamma oscillations had a higher signal-to-noise ratio than the broadband response , then the oscillatory signal would likely show a higher correlation with BOLD . In contrast , when the choice of stimulus or task can independently modulate broadband power and gamma oscillations so that the two LFP measures are not correlated , as in the experiments presented here and previously [38] , then gamma oscillations will not strongly correlate with BOLD . Our simulation and empirical results are consistent with studies that varied chromatic contrast and spatial frequency while measuring MEG and BOLD . These studies found that BOLD and narrowband gamma activity did not match in stimulus specificity [18 , 19] . It is likely that these stimulus manipulations , like ours , independently modulated narrowband gamma power and broadband power , although the studies did not quantify broadband fields , which are more challenging to measure with MEG than with ECoG [64] . We speculate that broadband fields spanning the gamma range would have shown a higher correlation with BOLD . In our model , the LFP measures are highly sensitive to neuronal synchrony , whereas BOLD is not . In our simulations , increases in neuronal synchrony drove narrowband gamma oscillations in the field potential . There are other cases of population activity with a high degree of neuronal synchrony . One example is the steady-state evoked potential associated with a periodic stimulus [65 , 66] . Previous studies have described discrepancies between evoked potentials and the BOLD signal , such as in the case of spatial summation [11] , directional motion selectivity [7 , 8] and spatial attention [9 , 10] . Our modeling framework suggests that the neural sources generating the steady-state potential ( synchronous neural activity ) are likely to be only weakly related to the BOLD signal ( depending largely on asynchronous signals ) , as these sources will primarily affect the second term on the right-hand side of Eq 4 ( cross-power ) . This does not imply that the two measures are always or even usually discrepant; the BOLD signal and steady-state potentials are likely to correlate any time that the steady state signals correlate with other measures of neural activity . When measures do dissociate , we do not conclude that one measure is more accurate; instead , the measures offer complementary views of the population activity , emphasizing the degree of synchrony or the average level of the response . An intriguing question is how each of the two signals contributes to perception and behavior [67] . Neural synchrony can also emerge without being time-locked to the stimulus , often called “induced synchrony” or “induced oscillations” [68] . In our simulation , we assumed that narrowband gamma LFP changes were induced by increases in synchrony between neurons and not by changes in the level of gamma power within the individual neurons . In contrast , we assumed that broadband LFP increases were induced by increased broadband activity in individual neurons and not by increased broadband coherence between neurons . ( In Eq 4 , a change in the left-hand side , LFP power in the gamma band , can be produced by a change in either the first or second term on the right ) . This explains why , in our simulation , the broadband power was correlated with BOLD , whereas the LFP gamma power was not , findings that were also confirmed by the data . Were our assumptions justified ? In principle , an increase in narrowband gamma power in the LFP could arise because the neurons synchronize in the gamma band or because ongoing gamma oscillations within each neuron increase in amplitude , independent of coordination between neurons . There is strong experimental support for the former . Experiments that measure both intracellular membrane potential from single neurons and the extracellular LFP show that when there is an increase in narrowband LFP gamma power , the gamma power from individual neurons becomes more coherent with the LFP [47] . Moreover , the coherence between local spiking and the LFP also increases in the gamma band when LFP gamma power increases [43] . These results are consistent with our assumption that a significant part of the increase in gamma LFP power arises from a change in population coherence . To our knowledge , it is not certain whether there is also some increase in the level of gamma signals within individual neurons when the narrowband gamma band LFP power changes . However , since we can attribute a large part of the change in gamma LFP to a change in coherence , we infer that we can only attribute , at most , a small part of the change in gamma LFP to the level of gamma power within neurons . In our simulation , we made two simple but extreme assumptions . First , we assumed that gamma oscillations occur with no change in the total level of neural activity , and hence no change in metabolic demand or BOLD . Second , we assumed that broadband responses occur with no change in neural synchrony . While these assumptions are likely incorrect at the limit , the simulations nonetheless captured the pattern of ECoG and fMRI results obtained in our datasets . Alternative models in which the broadband response was caused by a change in synchrony were much less accurate ( S7 Fig ) . Models in which gamma responses were caused by a change in level were only slightly less accurate and cannot be ruled out entirely ( S7 Fig ) . However , the regression models fit to our data ( Fig 9 ) show that the power of narrowband gamma oscillations does not predict the BOLD response . Hence the most parsimonious explanation is that these responses in the LFP are caused in large part by changes in synchrony . Both our measurements and our simulations showed that broadband electrophysiological responses were related to , but did not fully account for , the BOLD signal . This was especially evident in Simulation 2 and extrastriate data ( V2/V3 ) . In these cases , the amplitude of low-frequency oscillations ( 8–15 Hz ) was negatively correlated with the BOLD signal , independent of broadband signals . Numerous previous studies have reported that low-frequency oscillations are anticorrelated with BOLD , including measurements in motor , visual , and language areas [20–22 , 69–71] . This result may appear to conflict with the prior discussion , since we argued that oscillations ( to the degree that they reflect neuronal synchrony ) should have little to no effect on metabolic demand or the BOLD signal . It is therefore important to ask why low-frequency oscillations sometimes correlate with the BOLD signal , both in data and in simulation . One explanation is that alpha oscillations , or a mechanism that generates the oscillations , affect the BOLD signal indirectly by inhibiting cortical activity . According to this explanation , an increase in alpha power results in a decrease in local spiking activity , in turn reducing metabolic demand and the BOLD signal [72] . Alpha oscillations may indeed co-occur with reduced cortical excitation [73] . However , if this coupling between alpha power and spiking was the only explanation for the relationship between alpha power and BOLD , then a more direct measure of neuronal excitation , such as broadband or multiunit activity , would adequately predict the BOLD signal; alpha power would negatively correlate with the BOLD signal but would provide no additional predictive power . Our data and model do not support this explanation , as we find that for most cortical sites , the most accurate predictor of the BOLD signal is a combined model including both the amplitude of alpha oscillations and broadband power . We therefore propose that in addition to the indirect effect of modulating cortical excitability , alpha oscillations are also accompanied by a mean shift in membrane potential , making it less depolarized ( i . e . , closer to the equilibrium potential ) , and this shift reduces metabolic demand . Indirect evidence for a mean shift comes from MEG and ECoG studies [49 , 50 , 74] , which refer to alpha oscillations as being asymmetrical ( i . e . , they are not centered at 0—there is a mean shift ) . This can be explained by a simple process: if alpha oscillations reflect periodic inhibitory pulses , then on average , they will cause a hyperpolarization ( or less depolarization ) . If the neuron was slightly depolarized before the inhibitory alpha pulses , then the pulses would push the neuron toward equilibrium and hence a lower-energy state . In this view , large alpha oscillations reflect larger inhibitory pulses , reducing depolarization . We suggest that this reduced depolarization affects metabolic demand in two ways: by reducing spiking ( as discussed above ) and by maintaining a less-depolarized state , reducing metabolic demand . In our model , the contribution to the BOLD signal from each neuron is the power in the time series ( Eq 3 ) , and the mean contributes to power . The idea that a mean shift in the membrane potential affects metabolic demand ( in addition to altering excitability ) is consistent with the observation that slowly changing currents ( <0 . 5 Hz ) correlate with BOLD fluctuations [12 , 75] . Moreover , if alpha oscillations are associated with a mean shift in membrane potential , this would explain why cortical excitability depends on the phase of the alpha cycle: at one phase , the membrane potential is more depolarized , and hence cortex is more excitable , and in the opposite phase , cortex is more hyperpolarized and hence less excitable . This is consistent with the observations that the threshold for eliciting a phosphene with TMS changes with alpha phase [76 , 77] and that the alpha phase at the time of stimulus presentation influences the size of the BOLD response in visual cortex [78] . Inhibition takes two neurons—one to inhibit and one to be inhibited . In our simulations , the alpha oscillations ( C3 ) were associated with inhibitory fluctuations in the membrane potential ( mean below 0 ) , which in turn was associated with decreases in BOLD . It is important to note that these fluctuations are meant to capture the effect of local inhibition on the postsynaptic neurons ( the neurons being inhibited ) . The inhibitory neurons themselves are presynaptic , and the action of inhibiting other neurons is presumably an active process that consumes energy . Therefore , inhibition is expected to increase energy demand in some neurons ( the presynaptic neurons ) and decrease energy demand in other neurons ( postsynaptic neurons ) . We did not model the inhibitory neurons explicitly; however , the neural activity associated with active inhibition would be expected to contribute to the measured broadband signal in the ECoG data and is implicitly included in the broadband inputs in our simulations ( C1 ) . More complex models ( see paragraph 3 . 6 ) in which the circuitry of excitatory and inhibitory neurons is explicitly represented ( such as [63 , 79 , 80] ) may provide insight into how the balance between excitation and inhibition influences the field potential and the BOLD signal . We found that the relationship between the BOLD signal and features of the ECoG data differed across cortical areas . For example , broadband changes in ECoG responses explained more variance in the BOLD data in V1 than in V2/V3 . Conversely , low-frequency power decreases ( alpha , 8–13 Hz ) explained more variance in the BOLD signal in V2/V3 than in V1 . In the absence of a model , we might have interpreted the results as evidence that neurovascular coupling differs across sites . Many previous studies have reported differences in the relation between LFP and BOLD as a function of site or condition , for example , between cortical and subcortical locations [81] , across cortical regions [82 , 83] , between cortical layers [84] , and as a function of medication [85] . Here , we showed that a difference in the relationship between LFP and BOLD need not arise because of a difference in neurovascular coupling . In our results , Simulations 1 and 2 , like V1 compared to extrastriate areas , showed differences in the relationship between LFP and BOLD , yet we used the identical model of neurovascular coupling in all simulations . The systematic differences in the two simulations arose because of differences in the neuronal population activity , not because of differences in neurovascular coupling . While our results do not exclude the possibility of differences in neurovascular coupling across locations or states , they do caution against interpreting differences in the relationship between field potentials and BOLD as evidence for a difference in neurovascular coupling , since they show that a single model can account for a variety of patterns . More generally , the V1 versus V2/V3 discrepancies bolster the argument that one cannot predict the exact relationship between BOLD and field potentials without also specifying the neuronal population activity . A complete description of the biophysical processes giving rise to the BOLD signal and the field potential is far beyond the scope of this paper and is likely premature given the enormous complexity in the nervous system , the vascular system , and the coupling mechanisms between them . Instead , the purpose of our modeling framework was to first begin with a general principle , namely that BOLD and field potentials sum neural activity according to a different sequence of operations; second , to instantiate this principle in simple mathematical rules; third , to combine these rules with a minimal model of neural population activity; and finally , to ask to what extent such a model can account for the patterns in our data . Our model omits many biophysical components , such as compartmentalized neurons , multiple cell types and vessel types , neurotransmitters , the dynamics of blood flow , and so on; hence , it is not a detailed simulation of the nervous system or vascular system . We modeled the BOLD signal as a function of dendritic currents summed across neurons within an imaging region . The logic motivating this is straightforward . Neuronal activity consumes a large amount of energy , and this energy demand is dominated by the cost of restoring the membrane potential following ionic flows from synaptic potentials and action potentials [86 , 87] . As a result , the increased energy demand from neuronal responses is related to the dendritic currents . Neurovascular coupling is the process of increasing blood flow to meet this energetic demand; a failure of the hemodynamic response such as stroke can cause neuronal and glial cell death , highlighting the importance of the relationship between blood supply regulation and neuronal activity [33] . We modeled the hemodynamic response as being proportional to the energy demand from dendritic currents . This model was proposed as a computational summary of the approximate relationship between the BOLD signal and neuronal activity , not as a hypothesis about a causal mechanism . Recent work suggests that energy consumption per se ( e . g . , the change in the cerebral metabolic rate of oxygen consumption ) is not the triggering mechanism for the increased blood flow , rather neurotransmitters and other molecules associated with synaptic events are part of a complex cascade that causes vasodilation and changes in blood flow [33 , 88 , 89] . The exact biological mechanism responsible for neurovascular coupling is an area of highly active , ongoing research [35] . The key assumptions in the model—that the BOLD signal is correlated with changes in membrane potential and that the order of operations differs for the BOLD signal and the LFP—makes accurate predictions for our dataset . A separate and important research question is how closely the biophysical mechanisms match this computational-level description , and what these mechanisms are . The simplicity of the model has benefits . It facilitates an understanding derived from basic principles , similar to the advantages in building computational , rather than biophysical , models of neural responses [90–93] . Both types of models and empirical studies are valuable . Here , we emphasize that even with a highly simplified model of the BOLD signal , the field potential , and neuronal population activity , we are able to reconcile a wide range of findings in a complicated and technical literature . The model accounts for differences in how broadband field potentials and gamma oscillations relate to the BOLD signal . It can explain differences between cortical areas in the relationship between field potentials and BOLD . The model also provides an explanation for why the amplitude of alpha rhythms is negatively correlated with BOLD , even after accounting for the relationship between broadband signals and BOLD . We note that drastic simplifications are the norm in many fields of neuroscience , such as receptive field modeling of visual neurons; most such models omit fixational eye movements , optical properties of the eye , retinal and cortical circuitry , etc . , instead modeling responses as a few simple mathematical computations of the stimulus ( filtering , thresholding , and normalization ) [94] . These highly simplified models will certainly fail under some conditions [95] , yet they have proven to be of immense value to the field [93] , in part due to their simplicity and in part because the alternative in which the responses of visual neurons are computed from a complete , neurobiologically realistic model of the nervous system simply does not exist . To test competing computational theories about the relation between the visual input , the LFP , and the BOLD response , it is essential to make sample data and code available for others [38 , 53] . Following standard practices of reproducible research [96–98] , the Matlab code of the simulation and sample data and code to reproduce the Figs in this manuscript can be downloaded at https://github . com/dorahermes/Paper_Hermes_2017_PLOSBiology . To understand how the electrophysiology and BOLD responses are related , it is necessary to specify both the manner in which population activity transfers to the two signals and the neuronal population activity itself . The former shows that the covariance between neuronal time series has a large influence on the field potential and not the BOLD signal . Based on our simulations and empirical results , we made several inferences about the neuronal population responses mediating the BOLD signal and the LFP: that narrowband gamma oscillations in visual cortex likely arise more from synchronization of neural responses than a change in level of the neural response and hence have a large influence on the field potential and little influence on the BOLD signal , that responses that are asynchronous across neurons manifest in broadband field potentials and an elevated BOLD signal , and that low-frequency oscillations observed in field potentials are likely accompanied by a widespread hyperpolarization , which in turn reduces metabolic demand and the BOLD signal . Our model-based approach brings us a step closer to a general solution to the question of how neural activity relates to the BOLD signal .
Informed , written consent was obtained from all subjects . The fMRI protocols were approved by the New York University IRB and ECoG protocols were approved by the Stanford University IRB , according to the principles expressed in the Declaration of Helsinki . Simulations were computed for a population of 200 neurons . Each simulation trial was 1 second long with millisecond sampling . The time series for each neuron was derived by summing three inputs , each 1 second long , followed by leaky integration with a time scale of 10 milliseconds to simulate temporal integration in the dendrite ( Fig 4 ) . Each simulation was fit to ECoG data from one electrode and consisted of 240 trials , 8 repeats of 30 stimulus conditions . A condition in the simulation was defined by the parameter settings for the three inputs ( Fig 4 ) : C1 ( broadband ) , C2 ( gamma ) , and C3 ( alpha ) . Variations in these three inputs resulted in power changes in the broadband , gamma , and alpha LFP . The inputs were fit to data such that the simulated LFP power changes matched the ECoG data power changes for a particular electrode and stimulus . Stimuli for ECoG experiments were reported previously [38] . In brief , for one subject , the stimuli came from 8 classes of patterns ( 30 exemplars per class , 20x20° ) , including high contrast vertical gratings ( 0 . 16 , 0 . 33 , 0 . 65 , or 1 . 3 cycles per degree square wave ) noise patterns ( spectral power distributions of k/f4 , k/f2 , and k/f0 ) , and a blank screen at mean luminance ( S2 Fig ) . For the second ECoG subject , there were the same 8 classes as well as two other stimulus classes–a high contrast white noise pattern and a plaid at 0 . 65 cpd . The fMRI subjects had the same 10 stimulus classes as the second ECoG subject . ECoG data were measured from two subjects who were implanted with subdural electrodes ( 2 . 3 mm diameter , AdTech Medical Instrument Corp ) for clinical purposes at Stanford Hospital . Informed , written consent was obtained from all subjects . ECoG protocols were approved by the Stanford University IRB . In 22 electrodes in V1 V2 and V3 , broadband and narrowband gamma responses were quantified as before [38] , and alpha power changes were calculated . fMRI data was measured from four subjects ( three female , ages 22–42 ) with normal or corrected-to-normal vision at the Center for Brain Imaging at NYU . Informed , written consent was obtained from all subjects . The fMRI protocols were approved by the New York University IRB . fMRI data were preprocessed and analyzed using custom software ( http://vistalab . stanford . edu/software ) . Disc regions of interest ( ROIs ) ( radius = 2 mm ) were defined in fMRI subjects to match the position of the electrodes in ECoG subjects using a combination of anatomy , pRF centers , and visual field maps . The similarity between the ROI position and electrode position was compared via visual inspection of anatomical images and pRF centers ( S3 Fig ) . The relationship between fMRI and ECoG signals was analyzed using a linear regression model . The cross-validated coefficient of determination ( R2 ) was used as a metric for model accuracy , and the regression coefficients were used to test whether ECoG predictors ( broadband , gamma , and alpha ) had a positive or negative relation with BOLD . The relationship between fMRI and ECoG signals was analyzed using a linear regression model: y=Xb+c+ε where y is a vector of fMRI amplitudes ( beta estimates ) , with n entries for the n different stimuli; X is a matrix of ECoG responses , n by 1 , 2 , or 3 , where the columns correspond to one or more of broadband , gamma , and alpha estimates; b are the 1 , 2 , or 3 beta weights for the broadband , gamma , and alpha estimates; c is a constant ( the y-intercept ) ; and ε is the residual error term . The model was fitted separately for each cortical site ( electrode/ROI pair ) and for different combinations of predictors—broadband alone , gamma alone , alpha alone , each pairwise combination , and all three predictors together . The n stimuli in the regressions included the contrast patterns and the blank stimulus . Inclusion of the blank stimulus is important for capturing the sign of the mean response . For example , if all contrast patterns induced a BOLD response of a particular level ( say , +1 ) and induced ECoG responses of a particular level ( say , –1 ) and we did not include the blank stimulus in the regression , then after subtracting the mean from each measure , all beta estimates would be approximately 0 . This would mask a systematic relationship between ECoG and BOLD measures ( in this example , an anticorrelation ) arising from viewing stimuli with contrast compared to viewing a blank screen . Models were evaluated by split-half cross-validation . First , the regression model yi = Xibi + ci + ε was fit using half of the fMRI subjects ( 1 and 2 ) and half of the ECoG stimulus repetitions ( even repetitions ) . To cross-validate this model , the beta values ( bi ) were then applied to the left out half of the ECoG data ( odd stimulus repetitions ) to predict the left out half of the fMRI data ( fMRI subjects 3 and 4 ) . The same procedure was applied by reversing the training and testing data . This resulted in two testing datasets with BOLD responses predicted from ECoG for each stimulus condition ( Xibi + ci ) and an actual measured BOLD value . For each cortical site , the coefficient of determination ( see below ) was calculated between the concatenated predictions and BOLD data values of the two test sets . All R2 values reported in the results are cross-validated in this manner . The same pattern of results was achieved if instead of cross-validation , we solved the models on the complete datasets and computed the R2 adjusted for the number of regressors . To test whether different ECoG predictors ( broadband , narrowband , alpha ) had a positive or negative relation with BOLD , we tested whether the regression coefficient was significantly larger or smaller than 0 . The regression coefficient was considered to be significantly different from 0 using a bootstrap statistic: for each model , the median of the beta values across sites was calculated after resampling 10 , 000 times . If <2 . 5% of the resampled statistics were smaller than zero , the beta values were considered significantly positive , and similarly , if <2 . 5% of the resampled statistics were greater than 0 , the beta values were considered significantly negative . All model predictions were quantified using the coefficient of determination on cross-validated predictions . For predicting BOLD data from simulations of population neuronal activity ( Fig 7 , S7 Fig ) , the predicted BOLD has arbitrary units . In these cases , the observed BOLD and the predicted BOLD were both normalized by subtracting the mean and then dividing by the vector length . When predicting BOLD responses from features of the LFP data ( broadband , gamma , and alpha ) by regression , the predicted BOLD data were in the same units as the measured BOLD , and no normalizing or rescaling was done . To quantify the accuracy of the models , we calculated the cross-validated coefficient of determination , R2: R2=1–SSresidualsSSdata SSresiduals=∑i ( yi−fi ) 2 SSdata=∑i ( yi−y¯ ) 2 where y are the data values and f are the prediction values . Because the model fits are cross-validated , it is possible for the model errors ( residuals ) to be larger than the data values , hence R2 can be lower than 0 , and spans ( −∞ , 1] . In the case in which the model predictions and the data are unrelated and each are normally distributed with equal variance , R2 will tend to –1 .
|
There are several methods for measuring activity in the living human brain . Here , we studied functional magnetic resonance imaging ( fMRI ) , which depends on the vascular response to neuronal activity , and surface field potentials , which measure electrical activity from many neurons . These two widely used measurements of human brain activity often provide different and potentially conflicting results . We propose a quantitative model for how these two measurements integrate activity from neuronal populations . The fMRI signal is highly sensitive to the average level of local neuronal activity but not the degree of synchrony between neurons . In contrast , the field potential is most sensitive to synchronous neuronal signals . Our model accounts for several observations seen in fMRI and field potential data: some very large features of field potential recordings , such as gamma oscillations , can occur with little to no associated fMRI signal . The model predicts this because the gamma oscillations result more from increased neuronal synchrony than increased neuronal activity . Other field potential signals , such as broadband changes , which are likely driven by the level of neuronal activity rather than a change in synchrony , are highly correlated with fMRI . The two measures thus provide complementary information about human brain activity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"diagnostic",
"radiology",
"functional",
"magnetic",
"resonance",
"imaging",
"membrane",
"potential",
"brain",
"electrophysiology",
"neuroscience",
"magnetic",
"resonance",
"imaging",
"simulation",
"and",
"modeling",
"brain",
"mapping",
"electrocorticography",
"neuronal",
"dendrites",
"neuroimaging",
"research",
"and",
"analysis",
"methods",
"imaging",
"techniques",
"animal",
"cells",
"visual",
"cortex",
"cellular",
"neuroscience",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"cell",
"biology",
"anatomy",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neurophysiology"
] |
2017
|
Neuronal synchrony and the relation between the blood-oxygen-level dependent response and the local field potential
|
The lymphatic system is responsible for transporting interstitial fluid back to the bloodstream , but unlike the cardiovascular system , lacks a centralized pump-the heart–to drive flow . Instead , each collecting lymphatic vessel can individually contract and dilate producing unidirectional flow enforced by intraluminal check valves . Due to the large number and spatial distribution of such pumps , high-level coordination would be unwieldy . This leads to the question of how each segment of lymphatic vessel responds to local signals that can contribute to the coordination of pumping on a network basis . Beginning with elementary fluid mechanics and known cellular behaviors , we show that two complementary oscillators emerge from i ) mechanical stretch with calcium ion transport and ii ) fluid shear stress induced nitric oxide production ( NO ) . Using numerical simulation and linear stability analysis we show that the newly identified shear-NO oscillator shares similarities with the well-known Van der Pol oscillator , but has unique characteristics . Depending on the operating conditions , the shear-NO process may i ) be inherently stable , ii ) oscillate spontaneously in response to random disturbances or iii ) synchronize with weak periodic stimuli . When the complementary shear-driven and stretch-driven oscillators interact , either may dominate , producing a rich family of behaviors similar to those observed in vivo .
To maintain fluid homeostasis , interstitial fluid drains into the lymphatic system through initial lymphatic vessels that carry it to the collecting lymphatic vessels . The collecting lymphatic vessels transport the fluid ( known as lymph ) both passively and actively to lymph nodes and back to the systemic blood circulation . Collecting lymphatic vessels are surrounded by specialized lymphatic muscle cells ( LMCs ) [1] and sub-divided by valve structures that define individual segments called lymphangions ( Fig 1 ) [2] . Lymphangions serve as both pumps and conduits . In contrast to the blood circulation , where a single pump drives flow through relatively passive conduits , each lymphangion has the ability to pump lymph through the converging network to lymph nodes and eventually to the thoracic duct . Pumping occurs when expansions in radius draw fluid into the upstream end of the lymphangion and then expel it downstream during a contraction . Directional flow is enforced by intraluminal valves , which favor flow toward the thoracic duct . Lymphatic vessel contractions are triggered when cytosolic Ca2+ entering from intravascular stores and outside the cell surpasses a threshold concentration in the cytoplasm of the LMC , resulting in actin and myosin cross-bridging within the LMCs [3] . The contraction phase ends as transmembrane pumps restore cytoplasmic Ca2+ concentration to equilibrium allowing actin-myosin binding to relax , and the trans-wall pressure and the passive elastic properties of the wall to reopen the vessel . The effects of Ca2+ on LMC contraction are moderated by endothelial-derived relaxation factors ( EDRFs ) that act as potent dilators of lymphatic and blood vessels when produced by the vessel-lining lymphatic endothelial cells ( LECs ) in response to dynamic fluid shear stresses . The best known EDRF is nitric oxide although others such as histamine have been shown to be important [4 , 5] . For notational simplicity we represent the entire class of EDRFs herein as NO . The NO and Ca2+ levels are both subject to mechanical regulation; Ca2+ can enter the cell through stretch-activated ion channels [6 , 7] , and NO is produced by LECs when they are exposed to increased fluid shear stress [8] . Although rhythmic contractions can be produced by purely chemical oscillations in Ca2+ within the LMCs [9 , 10] , it is likely that feedback regulation is necessary for robust homeostasis . Indeed , we previously used a relatively complex numerical simulation of lymphatic pumping to demonstrate that a wide spectrum of oscillatory behaviors is possible , and that the behavior is very sensitive to local levels of stretch and stress [11] . Our present aim is to reduce the complexity of our previous model to examine how the observed oscillations arise from the integration of simple mechanical and chemical processes within a physiological system . Once simplified , we employ tools such as linear stability analysis to identify key parameter groups that determine the qualitative dynamic behaviors of such systems . Linear stability analysis seeks to determine which parameter values cause a small disturbance to grow rapidly from an initial state , or alternatively decay back to equilibrium . In addition , our approach allows us to show how oscillations can arise from the interactions between mechanical and chemical processes that lack intrinsic oscillators when considered separately . Here we develop generic formulations of the mechano-chemical processes in the muscular lymphatic vessel wall based on Ca2+ and NO signaling , then explore the dynamics of each chemical species while holding the effects of the other constant . Finally we examine the behavior of the fully coupled system . Linear stability analysis reveals a new class of oscillator arising from the dynamics of shear and NO that can act alone or in concert with the better recognized Ca2+ dynamics . The most remarkable feature of the shear-NO mechanism is its ability to offer distributed control of the pumping process , which is essential for managing a decentralized network of pumps and conduits .
Our model is based on a single lymphangion ( Fig 1 ) bounded at each end by one-way valves . The radius of the lymphangion is governed by the radial forces which are determined by the contractile ( Ca2+ ) and dilatory ( NO ) signaling molecules . Neglecting inertial effects , the radial forces on the vessel wall balance as DdRdt=−E ( R ) −F ( R , CCa , CNO ) +AR ( t ) ( 1 ) where the left hand side represents rate dependent effects with D incorporating the visco-elastic material properties of the vessel wall as well as viscous losses in the flow and lags in the transduction of concentrations into force , E is a restoring force-including elastic forces and vessel “tone” -imparted by the material properties of the vessel wall and F is a dynamic inward acting contractile force produced by muscle cells that surround the vessel . To a first approximation , we assume that the concentration of Ca2+ is transduced into a contractile force as F=FCaCCa/ ( 1+αCNO ) where α scales the possible desensitizing effect of NO [12] . The activation term AR can include a steady component from the mean transmural pressure difference pm that influences the baseline radius as well as extrinsic disturbances to the radius from the surrounding tissue and adjacent lymphangions . The restoring force is typically highly nonlinear[13–15] . Here we adopt the form E ( R ) =AeaR−poffset with stiffening coefficient a , scaling coefficient A , and offset pressure poffset selected to give a good fit of Shirasawa and Benoit ( see the third figure of reference [15] ) at typical operating pressures . In our numerical simulations we retain the full nonlinear form of E ( R ) , but for the stability analysis that follows we linearize the elastic force near an equilibrium radius R1 in the absence of dynamic increments to Ca or NO as E ( R ) ≈E0+E1 ( R−R1 ) where E0 is the elastic force at equilibrium and a Taylor series expansion near equilibrium yields E1=AaeaR1 . We find the equilibrium radius by solving 0=DdRdt=− ( AeaR1−poffset ) −FCaSCa0/KCa+pm for R1 where pm is the mean transmural pressure . Given the stiffening behavior of the wall ( E1∝eaR1 ) , we expect that appropriate values of E1 will be larger at higher mean transmural pressures where the equilibrium radius will be somewhat larger . The concentrations of the signaling molecules ( i ∈ {Ca , NO} ) are governed by the generic conservation law dCidt=−Ki ( CCa , CNO ) +Si ( R , R˙ , CCa , CNO ) +Ai ( t ) ( 2 ) where all concentrations are taken to be dimensionless ratios relative to a suitable reference , Ki is clearance of the signaling species through chemical reaction , transmembrane ion pumps and advective-diffusive transport , Si is a dynamic source term for the signaling molecule and Ai is an additional source term that can include the effects of imposed flow from upstream fluid pressure , inflammation , pace-making signals from adjacent cells , neural signaling , random disturbances , etc . Since our focus is on the interactions between Ca2+ and NO , we introduce a minimal representation of the Ca2+ dynamics rather than a fully detailed model of Ca2+ oscillations as may be found in the literature [10 , 16] . We retain the following features: i ) at rest , Ca2+ is at a low concentration in the cytoplasm of LMCs; ii ) a contraction is initiated when Ca2+ is rapidly admitted to the cytoplasm through ion-selective channels thereby triggering cross bridge formation between actin and myosin chains creating a contractile force [17]; iii ) relaxation of LMCs coincides with a drop in cytoplasmic Ca2+ concentration due to a drop in the rate of influx and the restoration of baseline conditions by ion pumps in the cell membrane and sarcoplasmic reticulum; and iv ) the LMC is refractory to a new contraction cycle until Ca2+ levels have returned to near equilibrium . As the Ca2+ levels approach their threshold level , we hypothesize that the membrane acquires sensitivity to small perturbations . Furthermore , the sensitivity is enhanced when the membrane is stretched to a larger radius . This models stretch-sensitive ion channels found in LMCs [6 , 18] . Each step in the process has the potential for modulation by NO . Alternatively , each form of modulation by NO can be disabled to demonstrate behaviors that have been observed in experimental preparations , for example after removal of LECs ( which produce NO ) or the genetic or pharmacological suppressions of NO [19–21] . We mathematically express the release of Ca2+ into the cytoplasm from intracellular stores and the extracellular fluid as the sum of a steady source Sca0 needed to maintain the baseline Ca2+ concentration and a transient component that is sufficiently rapid to be modeled as an impulse function δ ( t ) where t is the time since the Ca2+ concentration most recently passed the threshold necessary to trigger another contraction CCaThresh . This is expressed as , SCa=SCa0+SCa1δ ( t ) ( 1+γCNO ) ( 3 ) where SCa1/ ( 1 + γCNO ) is the magnitude of a bolus of Ca2+ with a possible reduction due to NO that is scaled by γ . We model the clearance of Ca2+ from the cytoplasm with KCa=KCa ( 1+βCNO ) ( CCa−CCaThresh ) ( 4 ) where KCa is a rate constant and β scales the possible enhancement of Ca2+ clearance attributed to NO [12 , 16] . At high concentrations of Ca2+ the clearance rate may be limited by the membrane pump capacity , but near the threshold required to trigger a contraction we assume clearance rates proportional to the concentration increment . The threshold itself may include a random component that we incorporate into the activation term . The linearized form of the model for a constant level of NO can now be written as DdRdt=−E1R−FCaCCa+AR ( t ) ( 5 ) and dCCadt=−KCaCCa+SCa1δ ( t ) +ACa ( t ) ( 6 ) where the constants have been absorbed into the activation terms so that the radius and concentration now represent the increments from baseline values . The parameters SCa1 and KCa now include the adjustments due to NO introduced in Eqs 3 and 4 . In the previous section we developed the model so that it reproduces Ca2+ induced contractions . We next considered how NO , created when LECs experience increased shear stress , can modulate contractions when it diffuses rapidly into adjacent LMCs . A suitable form for the NO source term can be obtained by considering steady laminar flow in a circular tube with negligible inertia [22] . Conservation of mass in a tube of time-varying radius requires that ∂Q∂z=−2πRdRdt which when integrated with respect to z along a vessel yields Q ( z ) =−2πRdRdtz+C1 where the constant of integration depends on the end conditions for the lymphangion . When the segment is contracting dRdt<0 and the upstream valve is closed [Q ( 0 ) = 0] we have Q ( z ) =−2πRdRdtz . Alternatively , if the segment is expanding dRdt>0 and the downstream valve is closed ( Q ( L ) = 0 ) , we obtain Q ( z ) =2πRdRdt ( L−z ) . We can express the mean flow along the length more compactly as Q¯=πR|dRdt|L . When additional flow Q0 is imposed on the segment by an axial pressure gradient we have Q¯=πR|dRdt|L+Q0 . Approximating the velocity profile with that of steady laminar flow with negligible inertia [22] we relate the mean shear stress τ to the flow rate by τ=4μQ¯πR3 . The mean shear stress along the segment due to dynamic changes during contraction or expansion is therefore approximately τ=4μLR2|dRdt|+4μQ0πR3 . Recent studies show that valves in collecting lymphatic vessels are biased toward the open condition [23] , but the simplification employed here allows us to study the basic stability of the system , at the possible expense of some accuracy in the predictions of pumping efficiency . There may be levels of shear stress below which NO production is negligible and above which NO production saturates at a maximum , but here we linearize the transduction of shear stress into the production of NO in an intermediate range to yield SNO=SNOR2|dRdt|+SNO0 ( 7 ) where SNO has absorbed the remaining constants in the shear stress expression and SNO0 represents NO released due to the through-flow term Q0 or chronic sources of NO such as might arise during inflammation . The source term SNO0 can be time varying , but arises from the local environment of the lymphangion and mathematically acts as an input to our model of a single lymphangion rather than as an interaction within the system itself . SNO0 therefore can serve as an external trigger to the system or as a steady offset , but does not directly impact the dynamics of an individual lymphangion , except by parametrically ( rather than dynamically ) changing the equilibrium radius . We can examine the effects of fluid viscosity on the pressure by using the same set of assumptions . The pressure will vary due to viscous flow effects according to ∂p∂z=−8μQπR4 . When contracting we have ∂p∂z=16μR3dRdtz , which when integrated along the length gives p ( z ) =16μR3dRdtz2+p ( 0 ) . Averaging over the length of the segment yields p¯=16μL23R3dRdt+p ( 0 ) where the first term gives the magnitude of the pressure decrement ( or increment for vessel expansion ) due to flow induced by the contraction of a single lymphangion . We see that the pressure increment due to flow induced by the single lymphangion also multiplies dRdt , so it can be absorbed into the overall damping term D . For typical vessel sizes , we find that the lag due to viscosity is orders of magnitude smaller than that from chemical and mechanical lags which are on the order of one second . NO does not produce a true outward force . However , it is conceptually equivalent to consider an effective force produced by NO that has the effect of countering FCa and the elastic effects . Mathematically , for small αCNO , we can write this as F=FCaCCa01+αCNO≈FCaCCa0 ( 1−αCNO ) ( 8 ) By defining FNO ≡ FCaCCa0α , we can write FNO=−FNOCNO ( 9 ) And the net force from Eq 8 becomes F=FCaCCa0−FNOCNO ( 10 ) where the net contractile force is decomposed into a positive term set by the baseline Ca2+ levels and a negative term that represents how the Ca2+ levels are modulated by NO . Thus , the NO-dependent term is not a true outward force , but arises mathematically from a reduction in the Ca2+-dependent contractile forces . The parameter values used in the simulations that follow are given in Table 1 . The parameter values were based on experimental data where possible , but were chosen to demonstrate a wide range of mathematical behaviors for the system rather than to mimic a particular experimental data set in detail . As a representative example , we show simulations based on measurements in rats [24] which offer data relating Ca2+ concentrations to lymph vessel diameter and contractile tension . Our own experiments discussed later [20] were done on mice which have smaller collecting lymphatic vessels than rats . Specific parameters governing the effects of NO are difficult to estimate , but fortunately may not be necessary here . As will be shown in the results section , we require only estimates of combinations of parameters such as SNO and FCaCCa0α , rather than values for each parameter individually . Unlike the geometrically-detailed continuum model of lymphatic NO transport in Wilson et al [25] that includes shear-induced production and clearance by diffusion , convection and reaction , our present model employs averages over a single lymphangion and combines the sensitivity to shear stress with the rate of production of NO . To that end , we employ parameter values that yield diameter changes due to NO on the order of 10% as observed in [20 , 21] . Moreover , we expect the effects of NO to be rapid . NO is released by endothelial cells about 2 seconds or less after increases in shear stress as observed previously [26–29] . Lymphatic vessels can be expected to dilate faster than blood vessels [30] because their muscle cells contain more rapid-acting contractile proteins than those of blood vessels [31] . We take the clearance of NO to be similar to , but somewhat faster than , that of Ca2+ [32] . Parameters such as the contractility , NO production and the mechanical stiffness appear to depend on anatomical location , species and age [13 , 28 , 33–36] , suggesting that the full range of possibilities realizable in vivo awaits further investigation .
In the absence of dynamic activation , Eq 6 implies the Ca2+ concentration during each contraction cycle will decay as CCa ( t ) =SCa1e−KCat . Using this as an input to Eq 5 , we find that the radius varies from its baseline value during each contraction cycle as ΔR ( t ) =FCaSCa1E1KCa ( e−t/tCa−e−t/tmech ) ( tmech−tCa ) ( 11 ) where we see that the return to equilibrium depends on two characteristic times , one set by the rate of Ca2+ clearance tCa = 1/KCa and the other by the mechanical lag tmech = D/E1 which can include the lag between the concentration increase and force production . As a reference time scale we have selected tCa = 1/KCa = 1 s which is in the range observed by Shirasawa and Benoit [15] . They observed a similar lag between the rise in Ca2+ concentration and the peak force generation . Here we use this lag as an estimate of tmech which we take to incorporate the visco-elastic and chemical-mechanical transduction lags . While both time constants contribute to the overall response , the slower of the two characteristic times gives the dominant time constant tc that determines the return to equilibrium . Experimental observations of the magnitude of radius change as a function of pressure show that it decreases with increased internal pressure [21] . This phenomenon is reproduced by our model as the vessel wall stiffens ( larger E1 in the denominator ) at greater radius . The amplitude may be further modified if the myosin cross bridging is length dependent as seen in skeletal muscle [36] . The frequency of contractions at constant NO levels is set by the interplay between the characteristic time for calcium tCa and the magnitude of the random activation term . In addition to the steady source of Ca2+ that establishes the vascular tone , we include a random component ACaRand ( R , t ) with zero mean and a standard deviation σ that can be applied to either the concentration itself or to the threshold level at which a new release of Ca2+ is triggered . A higher radius leads to more stretch in the LMC membrane and therefore greater sensitivity of ion channels . This can be modeled by increasing the noise level ( for example let σ ∝ pm ) . This will lead to a higher frequency at larger radii as found experimentally [21] . Exponential decay of Ca2+ near equilibrium leads to a latency period between contractions that varies as T = tc log ( Cmax/σ ) where Cmax is the magnitude of the Ca2+ increment from baseline . We further investigated random activation with the aid of numerical simulations implemented with the Euler-Maruyama method [37] , which properly scales the computational time step with the standard deviation of the noise ( Fig 2 ) . The simulation presented in Fig 2d–2f results from a higher mean pressure than Fig 2a–2c . Thus , the baseline radius is larger in Fig 2d–2f , which in turn yields a stiffer wall ( larger E1 ) leading to a smaller mechanical time constant ( smaller tmech ) and reduced amplitude for change in the radius . We note that our simple model of stretch-Ca2+ dynamics replicates important features of the contraction cycles observed in vivo [20] where contractions are generally similar to one another in magnitude and duration , but may be separated by inconsistent periods of latency . In Fig 2 , we see that the simulated contractions are nearly identical to each other ( that is , the trajectories nearly retrace one another in the phase portraits shown Fig 2c and 2f ) , but occur on inconsistent intervals . Even though the intervals between contractions are not perfectly uniform , they are well estimated by tc log ( Cmax/σ ) . We also see that increased transmural pressure can reduce the interval between contractions by stiffening the wall ( pm ↑⇒ R1 ↑⇒ E1 ↑⇒ tmech ↓⇒ tc ↓⇒ T↓ ) and also by increasing the sensitivity of the Ca2+ channels by stretching the vessel wall ( pm ↑⇒ σ ↑⇒ T ↓ ) yielding higher frequency contractions ( compare Fig 2b and 2e ) . We also find that our model of the stretch-Ca2+ process readily synchronizes when we impose extrinsic rhythmic pace-making since only small variations in the Ca2+ concentration relative to the threshold level are needed to initiate the next contraction cycle ( Fig 3 ) . Such small variations in Ca2+ concentration can be readily introduced by diffusion or voltage signals from adjacent LMCs . Alternatively , the vessel may be locally stretched by lymph arriving from upstream , which can also trigger a local contraction . In this way , neighboring LMCs can synchronize contractions to coordinate flow along a series of lymphangions throughout a connected network of collecting lymphatic vessels [38 , 39] . The conditions for oscillations in radius to arise near baseline Ca2+ levels in the absence of sharp spikes in Ca2+ as considered in the previous section are available from linear stability analysis of the shear-NO process near a point R1 which yields [R˙CNO . ]=[−E1DFNOD−SNOE1DR12 sgn ( R˙ ) SNOFNODR12−KNO][RCNO]+[inputs] ( 12 ) where the inputs include all extrinsic disturbances from the adjacent lymphangions and surrounding tissue . We treat small variations in R parametrically so that the dynamics of the system may be characterized by the eigenvalues of the Jacobian matrix [40] which are roots of the characteristic polynomial: λ2+ ( E1D+KNO−sgn ( R˙ ) SNOFNODR12 ) λ+E1KNOD=0 ( 13 ) Since all of the coefficients are positive , stability requires only that the second term be positive . Thus the system is always stable during contraction ( R˙<0 ) . However during dilation ( R˙>0 ) the second term can be positive or negative which allows the system to switch between stability and instability ( Fig 4 ) . Herein is the key feature from which NO can induce spontaneous oscillations in the radius without the sharp spikes in Ca2+ concentration described in Eq 6 . If the radius is large enough so that E1/D + KNO > SNOFNO/DR2 , then the fixed point is inherently stable . If instead , the radius is small enough that E1/D + KNO < SNOFNO/DR2 then the radius will unstably increase when perturbed . The instability arises because a slight increase in radius ( from point 0 on Fig 5 ) pulls fluid into the lymphangion , increasing shear and temporarily creating a runaway effect wherein more NO is released from the LEC further increasing the radius and drawing in still more fluid ( upper branch from point 0 to point 2 on Fig 5b ) . The instability persists until the radius becomes large enough at point 2 that the shear stresses begin to drop because more cross-sectional area is available for lymph flow . Thereafter the release of NO occurs more slowly than its degradation so that the system can return stably to equilibrium along the lower branch of the trajectory from point 2 to 0 . Mathematically , the unstable increase in radius persists until the sign of R˙ changes at point 2 . A change in the sign of R˙ does not require that the eigenvalues move to the left half of the complex plane at point B on Fig 4 as required for inherent stability , but rather requires only that the radius increase sufficiently to move the eigenvalues off of the real axis beyond point A , thus permitting at least a partial cycle of oscillation that includes a time at which R˙=0 . As the vessel begins to contract , the sign of SNOFNO/DR2 changes at the R-nullcline where R˙=0 , leading to an unconditionally stable return to the original radius . The time scale for contraction is approximated by tc ≈ tmech + tNO + tFNO where the three contributions arise from mechanical lag tmech as before , the clearance of NO tNO = 1/KNO , and the rate of force modulation by NO tFNO = SNOFNO/E1KNOR2 . To a similar degree of approximation , the instability of the NO-shear dynamics requires tmech + tNO ≤ tFNO . In other words , the change in force elicited by shear stress must persist longer than processes that tend to dissipate its effects . Exact algebraic expressions for the eigenvalues may be employed if desired , but this approximation captures the key dependencies . See Table 2 . The NO cycle can be generalized into a controllable and synchronizable oscillator . Fig 6 shows the behavior of the NO cycle in response to small random disturbances . During the stable contraction process , the vessel remains refractory to disturbances until close enough to equilibrium for a random disturbance to trigger another cycle , much as we found with the stretch-Ca process . Here the period of the NO-induced oscillations varies as tc log ( Cmax/σ ) as before but tc and Cmax now refer to NO rather than Ca2+ . As with Ca2+ , a relatively quiet environment or reduced sensitivity to disturbance will elicit longer latency periods between cycles but will not significantly change the shape of the shear-NO cycle . The NO dynamics can also readily synchronize with externally-imposed , small-amplitude sinusoids ( Figs 7 and 8 ) . Fig 7b and 7c show how the radius oscillates at precisely the input frequency for frequencies reasonably close to the response when noise triggered ( Fig 6 ) . However , when the input frequency is too high ( Fig 7a ) or too low ( Fig 7d ) synchronization occurs , but at half or double the input frequency , respectively . Fig 8 shows a parametric study of synchronization over a wide range of input frequencies and amplitudes where we find that synchronization can include a variety of integer ratios between input and output frequencies as explained further in the Discussion . Having explored the system dynamics when Ca2+ and NO are taken to be constant relative to each other we now consider their combined , dynamic effects . Our model includes three possible interactions reported in the literature: NO may i ) desensitize the LMCs to Ca2+ as modeled by Eq 10 , ii ) modify the availability of Ca2+ by 1/ ( 1 + γCNO ) or iii ) speed clearance of Ca2+ by 1 + βCNO [41 , 42] . Our simulations ( Fig 9 ) show that the dynamic effects of NO are most pronounced when the shear-NO dynamics are unstable . When the shear-NO dynamics are unstable , the radius can overshoot the nominal radius before or after a Ca2+-induced contraction , yielding oscillations in radius that are more symmetrical about equilibrium than when shear-NO is stable . At marginal stability ( Fig 9d ) , the NO concentration rings at a frequency determined by the point where the eigenvalues of the shear-NO oscillator cross the imaginary axis f = ( E1KNO/D ) 1/2/2π . At larger radii , the shear-NO mechanism is inherently stable , but can still reduce the magnitude of the oscillations driven by the stretch-Ca2+ process . This process is important in the presence of an assisting pressure gradient because the dilation induced by the forced flow can put the vessel into the range of radii where the shear-NO mechanism can inhibit contractions that would otherwise tend to restrict free flow through the vessel . The overall frequency is set by a complex interplay of stretch-Ca2+ and shear-NO mechanisms , but will typically be dominated by the faster of the two processes . Long latency intervals between Ca2+-induced contractions can permit NO to produce an unstable dilation , whereas , short intervals due to Ca2+ can suppress the autonomous oscillations possible through the NO mechanism . Interestingly , the published clearance rates for Ca2+ and NO cover a wide enough range that either possibility exists in vivo [15 , 32] . Experimental observations of diameter in vivo show cycles consistent with the model predictions ( Fig 10 ) ( data from [20] ) . In the absence of direct measurements of concentrations , we employ an alternative phase portrait of diameter plotted against the rate of change of diameter . Fig 10a and 10b are from a wild-type mouse in which Ca2+ and NO effects can operate normally . Here , we observe complex oscillations that include both rapid contractions and occasional strong dilations above the baseline diameter as expected from the shear-NO mechanism . In contrast , when the NO effects have been genetically deleted in eNOS-/- mice in Fig 10c and 10d , we see wave forms that are nearly identical to each other but dominated by contraction with the dilatory effects of NO appearing to be substantially weakened . In all cases , we see cycles occurring on irregular intervals as we expect from noise-triggered oscillators .
While the correspondence between the experiments and the model is encouraging , we should not expect a model of an isolated lymphangion to reproduce all features of a vessel in an intact , in vivo vascular network . For example , the effects of flow introduced from upstream or disturbances from surrounding tissue are inputs present in the animal , but are not included in the modeled dynamics . We have also not yet included effects due to nonlinear valve efficiencies or the bias of the check valves toward the open position [43] . Nonetheless , phase portraits , such as those newly employed here , promise to assist further study of the nonlinear dynamics that govern vascular oscillations . While our results await further experimental validation and improved estimates of key parameter values , it is interesting to consider the newly identified shear-NO oscillator more generically from a nonlinear dynamics perspective . The shear-NO oscillator has some important similarities and differences from the well-known Van der Pol oscillator [44] which has the form x¨−A ( 1−x2 ) x˙+x=g ( t ) ( 14 ) The form of the characteristic polynomial in Eq 13 implies that the shear-NO oscillator may be written as x¨+A ( 1−sgn ( x˙ ) B/x2 ) x˙+ ( x−x1 ) =g ( t ) ( 15 ) where the generic variable x fills the role of the radius in the shear-NO model , x1 is the nominal operating point and g ( t ) is a forcing function that can include steady , random or periodic components . The stability-determining second term in both oscillators can change sign based on the magnitude of the variable with a positive second term implying stability . The Van der Pol oscillator is known to self-sustain oscillations about the origin in phase space ( x , x˙ ) = ( 0 , 0 ) when g ( t ) = 0 , as its second term changes sign during different phases of each cycle . In contrast , the shear-NO oscillator operates near ( x , x˙ ) = ( x1 , 0 ) , but with x > x1 . Therefore , the second term in Eq 15 can be either i ) always be positive ( x12>B ) regardless of the sign on x˙ implying inherent stability or ii ) can be conditionally positive depending on both the magnitude of B and the sign of x˙ . As a result , the shear-NO oscillator cannot produce self-sustained oscillations for large radius ( x12>B ) . Furthermore , even when the radius is sufficiently small ( x12<B ) , the radius will return unequivocally to equilibrium as long as x˙<0 unless a non-zero forcing function is present to change the sign of x˙ . However , we find that when x12<B the magnitude of the forcing needed to start a new cycle can be arbitrarily small and in the form of either random noise or a periodic stimulus provided that enough time has passed for the system to approach its equilibrium point . In the context of the shear-NO dynamics , the key to oscillations is the inverse dependence on radius for the NO source due to shear stress in Eq 7 . As long as the exponent on R−2 remains negative ( increasing radius leading to lower shear stress and less NO production ) , then the NO-shear mechanism will be capable of a mathematical transition from unstable to stable as seen above in the generic oscillator in Eq 15 . The physiological impact of this result then depends on the relative magnitude of the time scales identified herein , not on any single parameter value . For example , the stability of the NO-shear mechanism depends on groups of parameters such as tFNO = SNOFNO/E1KNOR2 , which combines the sensitivity of the vessel to shear stress , the contractile force , the wall stiffness and the NO clearance rates . Inputs of constant magnitude have the effect of adjusting the equilibrium point . Using the shear NO oscillator as an example , an increase in transmural pressure will dilate the vessel , as will a pressure gradient that assists flow by inducing NO production via a steady shear stress . Likewise , a steady source of NO from local inflammation will chronically dilate the vessel [20] . If the vessel becomes sufficiently large , the stability criterion found above suggests that the shear-NO process will not support self-sustaining oscillations , in part due to the direct effect of radius on the stability criterion , but also due to greater stiffness of the wall at larger radius ( higher E1 ) . Nonlinearities in the force production and chemical source/elimination terms may also alter the stability in similar ways . Numerical simulation and examination in the phase plane reveal that the stretch-Ca2+ and shear-NO processes possess numerous symmetries that offer intriguing possibilities when the processes act together ( Figs 2 and 6 , Table 3 ) . Most notably , we see that the shear-NO process produces rapid and unstable dilation toward a larger radius , followed by stable contraction , while the stretch-Ca2+ process causes the vessel to contract rapidly and unstably toward a smaller radius and then to dilate stably . An essential feature of both the Ca2+ and NO mechanisms is that taken separately they do not produce traditional , self-sustaining limit cycles , but instead have a one-sided stability near equilibrium from which a new cycle begins only with a perturbation from the local environment . Interestingly , a suitable trigger for the stretch-Ca2+ oscillator can be an increase in radius produced by the shear-NO mechanism . And conversely , the shear-NO oscillatory can be triggered by a contraction arising from the stretch-Ca2+ mechanism . Balanov et al [45] reviews a variety of similar , so-called “noise-induced” oscillators in contexts outside of lymphatic physiology such as neurons and electrical monovibrators , but to our knowledge , the coupling of symmetric , noise-induced oscillators described in the present study has not been previously investigated . Balanov et al [45] also review how nonlinear oscillators can synchronize with small-amplitude sinusoidal inputs . Here we found that synchronization of either oscillator can occur over a wide range of frequencies ( shear-NO shown in Fig 8 , similar behaviors for stretch-Ca2+ acting alone and in combination with shear-NO can be observed ) . The synchronization behavior seen here is similar to that for the forced Van der Pol oscillator in its ability to produce so-called Arnold tongues which are broad domains within which the input and output frequencies are locked in ratios of m:n where m and n are small integers [44 , 46] . Kornuta et al [47] recently showed that lymphatic vessels studied ex vivo synchronize their contractions in a 1:1 fashion with imposed oscillatory variations in shear stress when the amplitude of the stimulus is sufficient large and the frequency of the input is relatively close to the autonomous frequency . Interestingly , they also observed that small amplitude variation in transmural pressure did not yield 1:1 frequency locking . However , our examination of their results ( Fig 8 in [47] ) suggests that 2:3 locking may have occurred . In the absence of imposed flow , they also found that the vessel continued to contract , but with a lower and more erratic frequency consistent with our simulated noise-triggered oscillator ( Figs 2 and 3 ) in the absence of the shear-NO mechanism . Ohhashi et al [48] also examined sinusoidal variations in transmural pressure at frequencies well away from the spontaneous frequency . Here too , 1:1 frequency locking did not arise , but the frequency of the contractions responded strongly to the input waveform . Given the subtlety of identifying non-1:1 synchronization , further examination of the experimental record may be warranted . In conclusion , we have presented a model of a vascular oscillator . The present analysis is sufficiently general to point toward several features that are likely found in other systems . The linear stability analysis shows: ( i ) complementary mechanisms for dilation and contraction of collecting lymphatic vessels , ( ii ) a fast , unstable process that recovers slowly and stably to a one-sided equilibrium , ( iii ) disturbance-based triggering that facilitates either synchronization with a cyclic pacemaker or spontaneous oscillations from random disturbances and ( iv ) the capability for reciprocal modulation between contractile and relaxation effects . Those features are not only limited to the presented example of Ca2+ and EDRFs but can be extended into other fields . The ability of the Ca2+ and NO based oscillators to respond to each other and external stimuli explains how lymphatic pumping can be coordinated along extended lengths of collecting lymphatic vessels without the need for higher order coordination . This new class of coupled , noise-driven oscillator can help to explain the diverse pumping behavior of lymphatic vessels .
|
For decades , cardiovascular physiology has been an area of intense research , and we have a fundamental understanding of the mechanisms the heart uses to drive blood flow through the distributed network of vessels in the body . The lymphatic system is now receiving similar attention as more is learned about its functional role in disease processes . The importance of the lymphatic system in collecting excess fluid from tissues and returning it to the blood is well known , but how the lymph flow is regulated without a central pump is poorly understood . Each segment of collecting lymphatic vessel can independently contract yielding a network of distributed pump/conduits . This paper shows how the lymphatic muscle cells that squeeze fluid along the lymphatic vessels can be effectively regulated using only chemical and mechanical signals that they receive from their immediate microenvironment . Using stability theory and the tools of nonlinear dynamics we identify two complementary oscillators that respond to stretch of the vessel wall and shear of fluid flowing over the vessel wall . Numerical simulations of the combined oscillators show that they have characteristics well suited to the regulation of distributed systems in general and may have application in other biological and physical contexts .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"acoustics",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"body",
"fluids",
"classical",
"mechanics",
"fluid",
"mechanics",
"radii",
"mechanical",
"stress",
"geometry",
"neuroscience",
"muscle",
"contraction",
"mathematics",
"algebra",
"sound",
"pressure",
"lymph",
"muscle",
"physiology",
"neurochemicals",
"nitric",
"oxide",
"fluid",
"dynamics",
"continuum",
"mechanics",
"fluid",
"flow",
"physics",
"shear",
"stresses",
"biochemistry",
"linear",
"algebra",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"eigenvalues"
] |
2016
|
Synchronization and Random Triggering of Lymphatic Vessel Contractions
|
Many nematodes form dauer larvae when exposed to unfavorable conditions , representing an example of phenotypic plasticity and a major survival and dispersal strategy . In Caenorhabditis elegans , the regulation of dauer induction is a model for pheromone , insulin , and steroid-hormone signaling . Recent studies in Pristionchus pacificus revealed substantial natural variation in various aspects of dauer development , i . e . pheromone production and sensing and dauer longevity and fitness . One intriguing example is a strain from Ohio , having extremely long-lived dauers associated with very high fitness and often forming the most dauers in response to other strains´ pheromones , including the reference strain from California . While such examples have been suggested to represent intraspecific competition among strains , the molecular mechanisms underlying these dauer-associated patterns are currently unknown . We generated recombinant-inbred-lines between the Californian and Ohioan strains and used quantitative-trait-loci analysis to investigate the molecular mechanism determining natural variation in dauer development . Surprisingly , we discovered that the orphan gene dauerless controls dauer formation by copy number variation . The Ohioan strain has one dauerless copy causing high dauer formation , whereas the Californian strain has two copies , resulting in strongly reduced dauer formation . Transgenic animals expressing multiple copies do not form dauers . dauerless is exclusively expressed in CAN neurons , and both CAN ablation and dauerless mutations increase dauer formation . Strikingly , dauerless underwent several duplications and acts in parallel or downstream of steroid-hormone signaling but upstream of the nuclear-hormone-receptor daf-12 . We identified the novel or fast-evolving gene dauerless as inhibitor of dauer development . Our findings reveal the importance of gene duplications and copy number variations for orphan gene function and suggest daf-12 as major target for dauer regulation . We discuss the consequences of the novel vs . fast-evolving nature of orphans for the evolution of developmental networks and their role in natural variation and intraspecific competition .
Phenotypic ( developmental ) plasticity describes the ability of an individual organism to develop distinct phenotypes from the same genotype . Besides numerous examples in plants and insects , nematode dauer development represents one key example of phenotypic plasticity ( Fig 1 ) [1] . The nematode model organisms Caenorhabditis elegans and Pristionchus pacificus undergo direct development through four larval stages under favorable environmental conditions , reaching adulthood in as little as three days under standard laboratory conditions ( 20°C ) ( Fig 1A ) . In contrast , unfavorable conditions , such as high temperature , low food availability , and high population density , result in the formation of long-lived dauer larvae [2] . Dauer larvae are resistant to many environmental stresses and show several morphological and behavioral adaptations . They have a closed mouth and a thick cuticle , enabling survival under harsh conditions . In addition , many dauer larvae show a nictation or waving behavior ( Winkverhalten ) , which is usually considered to represent a dispersal strategy , allowing dauer larvae to attach to and disperse with various invertebrates . For example , P . pacificus is associated with scarab beetles in the wild and shows a necromenic association with its beetle hosts ( Fig 1B ) [3] . On the living beetle , nematodes are exclusively found in the dauer stage , and they resume development only after the beetle´s natural death by feeding on developing microbes on the carcass [4] . Therefore , the nematode dauer stage is usually considered to represent the most important dispersal and survival strategy that has contributed enormously to the evolutionary success of this taxon [5] . In C . elegans , dauer development is regulated by a complex genetic network involving pheromone , insulin , TGF-β , and endocrine signaling [6] . The dauer pheromone consists of a modular library of small-molecule signals , the ascarosides , which induce the formation of dauer larvae [7] . Several signaling pathways , including insulin , TGF-β , and cGMP signaling , have been shown to transmit various environmental signals and are thought to converge on an endocrine signaling module . Endocrine signaling consists of the nuclear hormone receptor DAF-12 that acts as a developmental switch [8] . In its ligand-free state , DAF-12 induces dauer formation and consistently , daf-12 loss-of-function mutants are dauer-formation-defective ( Daf-d ) . In contrast , in the presence of the DAF-12 ligand , dauer formation is inhibited . Several derivatives of the steroid hormone dafachronic acid ( DA ) were shown to act as DAF-12 ligands , including ∆7-DA , ∆4-DA , and ∆1 , ∆7-DA [9 , 10] . Besides these detailed genetic and molecular studies that made dauer formation an important model in biomedical research , various studies in P . pacificus have established nematode dauer formation as a model system for investigating natural variation and its consequences on the evolutionary ecology of the organism . First , Bose and coworkers ( 2014 ) showed substantial natural variation in pheromone signaling [11] . In P . pacificus , dauer pheromones consist of a blend of ascarosides and paratosides with chemically very diverse building blocks from all major metabolic pathways [7 , 12] . The comparison of the composition of the pheromones of six natural isolates of P . pacificus revealed tremendous variation in pheromone composition , even among three strains from the same habitat on La Réunion Island in the Indian Ocean . Second , when exposed to individual pheromone components different strains showed enormous variation in their dauer formation response with little correlation between pheromone production and pheromone sensing [11] . These results have been interpreted as cross-preference and indication for intraspecific competition , a phenomenon to be described below . Third , dauer larvae were also shown to differ in their survival properties [13] . Specifically , we showed that dauer larvae of eight tested P . pacificus wild isolates survived under standardized laboratory conditions for 25 to 50 weeks . The strain RS5134 isolated from a scarab beetle in Ohio ( USA ) showed one of the most extreme survival rates of 50 weeks in distilled water at 8°C ( Fig 1C ) . Finally , these strains not only survived but were also able to reproduce after dauer recovery as indicated by brood size tests [13] . Again , variation in fitness was observed with the strain from RS5134/Ohio producing approximately 100 progeny after 46 weeks in the dauer stage ( Fig 1C ) . Together , these studies support the notion that all tested aspects of dauer induction and exit show substantial natural variation . One of the most striking findings of these natural variation studies was the lack of correlation between small-molecule production and sensing . The analysis of pheromone extracts of 16 P . pacificus wild isolates had already indicated that most dauer pheromones induce higher dauer formation in other P . pacificus genotypes rather than in their own ( Fig 1D ) [13] . For example , the strain RS5134 from Ohio formed more dauers in response to the pheromone of the wild-type strain RS2333 from California than in response to its own pheromone . This response pattern , which was observed in 13 of 16 tested strains , has been described as “cross-preference” ( Fig 1D ) [13] . Follow-up analysis indicated that the strains RS2333/California and RS5134/Ohio differ in the exact composition of their dauer pheromones [11] . Inspired by these surprising results on natural variation in dauer pheromone production and sensing in P . pacificus , a novel assay was established to analyze if natural isolates can compete for dauer induction [11] . While our competition experiments support intraspecific competition in nematode dauer formation , the underlying molecular mechanisms and the potential ecological consequences of these results remain largely unknown . In evolutionary terms , intraspecific competition has been suggested to be associated with evolutionary arms races and to represent a strong selective force driving the divergence among populations [14 , 15] . Indeed , studies in bacteria have provided detailed insights into competitive interactions often involving toxin-antitoxin systems [16] , but little is known about the genetic mechanisms underlying intraspecific competition in animals . Here , we investigate the molecular mechanisms underlying the observed cross-preference and competition between RS2333/California and RS5134/Ohio . Using a recombinant-inbred-line ( RIL ) and quantitative-trait-loci ( QTL ) approach , we made the surprising finding that intraspecific competition relies on an orphan gene , dauerless , that acts by copy number variation . dauerless is exclusively expressed in CAN neurons , and animals in which the CAN neurons have been ablated , as well as dauerless deletion mutants generated by the CRISPR/Cas9 system , show increased dauer formation . Finally , epistasis analysis indicates that dauerless acts downstream or in parallel of steroid-hormone signaling but upstream of daf-12 . Our findings reveal the importance of gene duplications and dosage effects and indicate that novel or fast-evolving genes can have key functions in developmental regulatory networks .
The molecular mechanisms underlying cross-preference and intraspecific competition among nematode populations can best be investigated by RIL and QTL approaches . We selected the two strains RS2333/California and RS5134/Ohio for molecular investigations because of their large differences in dauer induction and because RS5134/Ohio has the highest dauer formation in response to the pheromone of RS2333/California out of eight tested strains [13] . Specifically , RS5134/Ohio shows nearly 60% dauer formation in response to the RS2333/California pheromone , but only 35% dauer formation in response to its own pheromone ( Fisher's exact test , P<0 . 006 ) ( Fig 1D ) . In contrast , RS2333/California has a low dauer formation phenotype in response to its own and the RS5134/Ohio pheromone ( Fig 1D ) , indicating that the California—Ohio pair has robust phenotypic differences that allow QTL analysis . The cross-preference between RS2333/California and RS5134/Ohio was originally observed in dauer pheromone assays ( Fig 1D ) [13] . To test for the existence of intraspecific competition between RS2333/California and RS5134/Ohio we performed competition assays in Ussing chambers ( Fig 2A ) and measured dauer formation over time ( Fig 2B ) . Starting at days 10 and 11 , RS5134/Ohio formed more dauers when exposed to the RS2333/California pheromone than when only exposed to its own pheromone in the control ( Fisher's exact test , P<0 . 003 ) ( Fig 2B ) . In contrast , RS2333/California showed approximately the same dauer formation phenotype when exposed to the RS5134/Ohio pheromone as in the control experiment after the same amount of time ( Fig 2B ) . In the following , up to day 14 ( Fig 2B ) , RS2333/California dauer formation stayed low at approximately 15% . In contrast , RS5134/Ohio showed increased dauer formation reaching 30% in the control , but 60% when exposed to the RS2333/California pheromone ( Fisher's exact test , P<0 . 0001 ) ( Fig 2B ) . These results indicate that cross-preference of dauer pheromones leads to a higher dauer formation phenotype in RS5134/Ohio when exposed to the RS2333/California pheromone possibly indicating that strains can compete for dauer entry . To elucidate the molecular mechanisms underlying intraspecific competition , we generated 911 RILs of RS2333/California and RS5134/Ohio ( S1A Fig ) . For each RIL , we determined the dauer formation phenotype in response to both parental pheromones ( see Materials and Methods ) . Next , we selected 136 RILs , covering all phenotypic classes ( high , intermediate and low dauer formation phenotypes ) for genotyping with simple sequence length and conformation polymorphism markers [17] . Using QTL mapping , we indentified six QTL peaks with significantly high logorithm-of-odds ( LOD ) scores ( S1B Fig ) . Fine mapping enabled us to narrow down the QTL peak associated with the marker ME25944 on chromosome I to a 10 kb region based on the high recombination frequency in this interval of the P . pacificus genome ( Fig 3 ) . This 10 kb region contains only two gene predictions , the orphan gene Contig44-snap . 18 and the globin-like gene Contig44-snap . 19 ( Fig 3A ) . Comparison of expression levels by RNA-seq between RS2333/California and RS5134/Ohio showed a strong upregulation of Contig44-snap . 18 in RS2333/California ( FDR<0 . 05 ) , which makes Contig44-snap . 18 the prime candidate for the causative gene within the QTL peak . Contig44-snap . 18 contains 10 exons with a conceptual translation into a polypetide of 314 amino acids ( Fig 3A ) . There is no sequence similarity at the DNA and protein level of Contig44-snap . 18 to sequences outside the genera Pristionchus and Parapristionchus ( S2 Fig ) . Also , we did not find any signal peptide or other known sequence motifs using the programs SignalP4 . 1 and prosite release 20 . 111 . Thus , Contig44-snap . 18 represents a true orphan ( pioneer ) gene ( for details see below ) . Orphan genes are common in P . pacificus and other nematodes [18 , 19] , and a large number of the P . pacificus orphan genes have been shown to be expressed in transcriptomics and proteomics studies [20 , 21] . Further investigation of Contig44-snap . 18 suggests copy number variation ( CNV ) at this locus between RS2333/California and RS5134/Ohio . In the RS2333/California genome assembly , Contig44-snap . 18 is located in a 4 . 3 kb region for which the coverage is approximately twice as high as for the adjacent regions ( Fig 3B ) . This difference in coverage does not exist in RS5134/Ohio , suggesting that the region was duplicated in RS2333/California but not in RS5134/Ohio ( Fig 3B ) . Additional support for a local and recent duplication event comes from the re-sequencing project of 104 wild isolates of P . pacificus [22] . The strain RS106 from Poland , which represents the strain most closely related to RS2333/California , is the only other strain with a difference in coverage of Contig44-snap . 18 ( Fig 3B ) . Using inverse PCR , we determined the location of the duplicated region on chromosome III of RS2333/California ( Fig 3C ) . The second copy of Contig44-snap . 18 on chromosome III is identical to Contig44-snap . 18 on chromosome I over a 4 . 3 kb interval , as expected for a recent duplication event . This includes 1 . 1 kb of upstream and 1 . 5 kb of downstream sequences ( Fig 3C ) . In contrast , the Contig44-snap . 18 genes on chromosome I of RS2333/California and RS5134/Ohio contain a total of 5 single nucleotide polymorphisms ( SNPs ) in the 1 . 7 kb open reading frame and a total of 13 SNPs in the complete 4 . 3 kb region ( Fig 3C ) . Interestingly , the second copy of Contig44-snap . 18 gene on chromosome III of RS2333/California is also associated with one of our QTL peaks . However , due to the much lower recombination frequency in this area of the P . pacificus genome , our attempts at finemapping of this QTL were unsuccessful . Together , these findings further support the role of the orphan gene Contig44-snap . 18 as prime candidate for the QTL on chromosome I and let us hypothersize that the observed phenotypic differences between the two strains are caused by a recent CNV . The CNV hypothesis for Contig44-snap . 18 suggests that the presence of one gene copy results in high dauer formation , whereas two gene copies cause low dauer formation in response to dauer pheromones . To test this hypothesis and to determine if Contig44-snap . 18 indeed plays a role in dauer formation , we generated transgenic lines that carry multiple copies of Contig44-snap . 18 . According to the CNV hypothesis , a further increase in Contig44-snap . 18 copies should further decrease dauer formation after exposure to dauer pheromone . We injected a 9 . 5 kb genomic construct of RS2333/California or RS5134/Ohio , consisting of a 3 . 3 kb upstream ( promoter ) region , the 1 . 7 kb Contig44-snap . 18 open reading frame , and a 4 . 5 kb downstream region , into RS2333/California or RS5134/Ohio animals . We generated a total of five transgenic lines carrying the RS2333/California version of Contig44-snap . 18 in either the RS2333/California or RS5134/Ohio genetic background and the RS5134/Ohio version of Contig44-snap . 18 in RS2333/California ( Table 1 ) . Strikingly , transgenic animals of all five lines do not form dauers at all ( Table 1 c1 , d1 , e1 , f1 , g1 ) . In contrast , non-transgenic nematodes that have lost the transgenic array show the wild-type RS2333/California or RS5134/Ohio dauer formation phenotype ( Table 1 c2 , d2 , e2 , f2 , g2 ) . Furthermore , in a transgenic control line , which only expresses the red-fluorescent-protein ( RFP ) injection marker , both transgenic and non-transgenic animals form dauers at wild-type frequencies ( Table 1 h1 , h2 ) . These results support the CNV hypothesis for Contig44-snap . 18: One copy results in the high dauer formation phenotype of RS5134/Ohio , two copies lead to the low dauer formation phenotype of RS2333/California , and multiple copies eliminate dauer formation in transgenic animals . Since the expression of multiple copies of Contig44-snap . 18 completely inhibits dauer formation , we named the gene dauerless , dau-1 . 1 for the copy on chromosome I and dau-1 . 2 for the copy on chromsome III . Given the absence of any sequence similarity of dau-1 to genes in other organisms , we next wanted to determine the expression pattern of dau-1 to obtain additional insight into its function . We generated three independent transgenic lines by injecting an RS2333/California translational reporter construct , containing a 4 . 7 kb promoter region and the dau-1 open reading frame driving RFP expression , into RS2333/California animals . Surprisingly , we found that dau-1 is exclusively expressed in the CAN neurons , which are a pair of neurons in the mid-body region born in late embryogenesis ( Fig 4A ) . Specifically , we observed RFP expression in CAN neurons in the J2 , J3 , and J4 stages of all three transgenic lines . dau-1 expression in CAN neurons suggests a previously unknown role for these neurons in P . pacificus dauer formation . In C . elegans , little is known about the function of CAN neurons other than that they are essential for survival . Specifically , C . elegans animals in which CAN neurons have been ablated die within 24 hours . In contrast to C . elegans , CAN ablation in P . pacificus is viable . Therefore , to analyze the function of CAN neurons in dauer formation , we ablated the CAN neurons after hatching and performed dauer pheromone assays with J2 larvae ( Fig 5; Table 1 q-v ) . Surprisingly , after CAN ablation RS2333/California animals show 94% and 75% dauer formation in response to RS2333/California and RS5134/Ohio pheromone , respectively ( Fig 5; Table 1 r ) ) . This dauer formation phenotype is not only higher than the wild type RS2333/California response but also higher than the wild type RS5134/Ohio response , suggesting that CAN neurons are part of a network repressing dauer formation . Similarly , CAN ablation in RS5134/Ohio resulted in extremely high dauer formation in response to both pheromones ( Fig 5; Table 1 t ) . Most surprisingly however , a transgenic line expressing the dau-1 . 1::RFP reporter construct also showed an extremely high dauer formation phenotype in pheromone response , indicating that dau-1 expression in CAN neurons is necessary for the dau-1-mediated inhibition of dauer formation ( Fig 5; Table 1 v ) . The observed CAN ablation phenotype is consistent with the CNV hypothesis of dau-1: Expression of an increased number of dau-1 copies reduces dauer formation . In contrast , eliminating dau-1 by CAN ablation increases dauer response above wild type levels , suggesting that dau-1 represents a strong suppressor of dauer development . To further support the idea that dau-1 is a suppressor of dauer formation , we tested if the specific elimination of one of the two copies of dau-1 in RS2333/California would increase the dauer formation phenotype to wild type RS5134/Ohio levels . We generated deletion mutants in the RS2333/California background using the CRISPR/Cas9 system [23] and obtained two lines ( tu490 and tu491 ) with a deletion in dau-1 . 1 and one line ( tu492 ) with a deletion in dau-1 . 2 . All three mutant lines form more dauers than wild type RS2333/California ( Fig 5; Table 1 w1 , w2 , w3 ) . Furthermore , the mutant dauer formation phenotype mimics the wild type RS5134/Ohio phenotype in response to both dauer pheromones . Thus , eliminating one dau-1 copy in RS2333/California is sufficient to increase the level of dauer formation to the RS5134/Ohio phenotype . Next , we wanted to know if a dau-1 double mutant in the RS2333/California background would further increase the dauer formation phenotype to the level observed after CAN ablation . We generated a double mutant by crossing the single mutant lines dau-1 . 1 ( tu490 ) and dau-1 . 2 ( tu492 ) . Indeed , the resulting double mutant dau-1 . 1 ( tu490 ) ; dau-1 . 2 ( tu492 ) showed a further increase in dauer formation after treatment with both pheromones ( Fig 5; Table 1 x1 ) . Specifically , after treatment with the RS2333/California pheromone double mutant animals formed 91% dauers as compared to 94% dauer formation after CAN ablation ( Fig 5; Table 1 r , x1 ) . Similarly , in response to RS5134/Ohio pheromone , double mutant and CAN ablated animals formed approximately 80% dauers , whereas both single mutants formed significantly fewer dauers after the same treatment ( Fig 5; Table 1 r , x1 ) . Note , that statistically significant differences were only observed after treatment with the RS5134/Ohio pheromone , most likely due to the lower baseline of dauer formation in comparison to the RS2333/California pheromone . Together , these findings provide final support for the CNV hypothesis and indicate that the role of the CAN neurons in dauer formation is exclusively regulated by the dau-1 genes . To determine if dau-1 acts upstream or downstream of endocrine signaling , we performed epistasis analysis with a daf-12 mutant and cholesterol depletion experiments . Dauer formation in P . pacificus is known to involve cholesterol-derived steroid hormones and therefore , we made use of the fact that dauer induction by pheromones can be enhanced by cholesterol depletion [24] . In control experiments , wild type RS2333/California and RS5134/Ohio animals showed increased dauer formation in the absence of cholesterol in response to both pheromones ( Table 1 i , j ) . However , when we tested the dau-1 transgenic lines in dauer pheromone assays using agar plates without cholesterol , all transgenic animals still showed 0% dauer formation ( Table 1 k1 , l1 , m1 , n1 , o1 ) . In contrast , the non-transgenic nematodes that have lost the transgenic array and the transgenic control line show increased dauer formation ( Table 1 k2 , l2 , m2 , n2 , o2 , p1 , p2 ) . Thus , expression of multiple copies of dau-1 completely inhibits dauer formation even after cholesterol depletion , indicating that dau-1 acts downstream of or in parallel to steroid-hormone signaling . Next , we used epistasis analysis between dau-1 and the nuclear hormone receptor daf-12 by generating a dau-1 . 1 ( tu490 ) ;dau-1 . 2 ( tu492 ) ;daf-12 ( tu389 ) triple mutant . Interestingly , triple mutant animals show no dauer formation after pheromone treatment and thus , mimic the daf-12 single mutant phenotype ( Fig 5; Table 1 y , z ) . This finding indicates that daf-12 is epistatic to dau-1 . Similarly , application of ∆7-DA does inhibit dauer formation in the dau-1 . 1 ( tu490 ) ;dau-1 . 2 ( tu492 ) double mutant ( Fig 5; Table 1 x2 ) . Together with the cholesterol depletion experiments , these results indicate that dau-1 acts downstream or in parallel of steroid-hormone signaling but upstream of daf-12 . One intriguing hypothesis would be that DAU-1 represents a novel inhibitor of the DAF-12 dauer-inducing function and acts independently of steroid-hormones . Finally , we studied the evolutionary history of dau-1 in the Pristionchus genus . Interestingly , genome-wide analysis revealed that dau-1 has two additional paralogs in RS2333/California; first , Contig1-snap . 329 on chromosome I with 95% amino acid sequence similarity and second , Contig24-snap . 126 on chromosome II with 32% amino acid sequence similarity ( Fig 4B , S2 Fig ) . We called these genes dauerless-like ( dal ) . Phylogenetic analysis indicates that the dau-1 locus evolved from a P . pacificus-specific duplication after the split from its sister species P . exspectatus ( Fig 4B ) . Specifically , P . exspectatus and P . arcanus have only one dal gene with extremely high sequence similarity to dau-1 that we named dal-1 ( Fig 4B ) . All P . pacificus strains also contain a dal-1-like gene . However , the close sequence similarity between dal-1 of P . exspectatus and P . arcanus and dal-1 and dau-1 of P . pacificus make it impossible to determine which gene resulted from the duplication in the P . pacifcus lineage . The other paralog of dau-1 , Contig24-snap . 126 is conserved throughout the genus Pristionchus and we named this gene dal-2 . dal-2 genes were observed in all tested Pristionchus species , but they show only limited sequence similarity to dal-1 and dau-1 . Furthermore , the only dal gene found outside of the genus Pristionchus is in the sister genus Parapristionchus , whereas dal genes were not observed in genomes and genome drafts of 10 more distantly related nematodes , including C . elegans . Taken together , dau-1 is an orphan gene that is not found outside Pristionchus but has a complex history involving several gene duplications over short evolutionary time scales .
Our work shows that genes that lack high sequence similarity to genes in other species can be of importance for the development , ecology and evolution of an organism . While the conservation of developmental control genes has become a general truism of the modern life sciences [25] , only a fraction of all genes in an animal is highly conserved . Genome sequencing projects revealed that a substantial part of genes show limited or no sequence similarity to genes in other organisms [26] . In some cases , such as the sequencing projects of nematodes of 11 different genera , more than 20% of all gene predictions are orphan genes [19] . While transcriptomics and proteomics studies do provide evidence for the expression of orphan genes , little is known about their exact function . This study on P . pacificus dau-1 reveals that orphan genes can indeed be integral parts of more complex regulatory networks , an observation that results in several interesting evolutionary questions . First , molecular phylogeny of dau-1 strongly suggests that dau-1 and related genes evolve rapidly both , with regard to copy numbers and sequence divergence . Therefore , it remains unknown if dau-1 represents a novel or a fast evolving gene . The fact that the distant paralog Contig24-snap . 126 shares only 32% sequence similarity is in line with findings in Drosophila that many genes are fast evolving [27 , 28] . Second , the absence of sequence similarity to genes in other species excludes the usage of gene ontology to obtain first indications for the biochemical function of dau-1 . Therefore , despite the genetic and molecular evidence for the role of dau-1 in dauer development as provided in this study , its exact target as inhibitor of dauer development remains currently unknown . We speculate however , that the developmental switch gene daf-12 is the direct target of dau-1 , a hypothesis that would be consistent with our epistasis analysis . No matter , if dau-1 is a novel or fast-evolving gene , its potential interaction with the daf-12 gene or protein would be intriguing . If dau-1 is indeed a novel gene , its interaction with DAF-12 would represent a novel inhibitory loop of the dauer regulatory network . Instead , a fast-evolving dau-1 locus would suggest that parts of the network can evolve rapidly . This would be consistent with the observation that P . pacificus and C . elegans DAF-12 show little to no sequence similarity outside of the steroid-ligand and DNA-binding sites although both genes encode for large proteins [24] . While future studies will have to reveal the biochemical function of DAU-1 , such work is complicated by the fact that all previous attempts to crystalize the protein have failed so far . As a second major conclusion , our work supports the general notion , also coming from several studies in medicine , that CNV can affect important developmetal decisions . In the example of dau-1 , several lines of evidence support the CNV hypothesis ( Fig 6 ) : one copy of dau-1 in the Ohioan strain cause high dauer formation and two copies in the Californian strain causes low dauer formation . In contrast , multiple copies suppress dauer formation altogether , whereas physical and genetic ablation of the CAN neurons and both copies of dau-1 in RS2333/California result in extremely high dauer formation . Thus , CNV is not only relevant in human disease , but also of importance for invertebrate development . Very recent genomic studies further support the notion that CNV represents a widespread evolutionary phenomenon . The comparison of young genes across multiple stickleback populations revealed extensive CNV again linking CNV with fast evolving genes [29] . In retrospect , we were surprised that CNV of dau-1 resulted in a mappable QTL on chromosome I , which harbors a copy in both strains , RS2333/California and RS5134/Ohio . Whereas the lower recombinantion frequency in the concerned area of chromosome III prevented us from identifying the associated gene , which might well be dau-1 . 2 , we assume that differences in the expression and/or activity of dau-1 . 1 between RS2333/California and RS5134/Ohio are responsible for the QTL associated with ME25944 . Indeed , dau-1 . 1 has 5 SNPs resulting in amino acid differences between DAU-1 . 1 of RS2333/California and RS5134/Ohio and a total of 13 SNPs in the duplicated area . However , the strong effects observed after dau-1 overexpression in transgenic animals in any combination ( Table 1 c-g ) prevents us from using swapping experiments to investigate the role of individual SNPs . The presence of additional dal genes in the P . pacificus genome results in specific questions that will be addressed in future studies . CRISPR-Cas9 induced gene inactivation can be used to study the function of Contig1-snap . 329 and Contig24-snap . 126 , but first attempts to produce such mutants failed ( M . G . M . , J . deVriend , M . Atzhigi & R . J . S . ) . One intriguing hypothesis would be that dal-1 , the gene with the highest sequence similarity to dau-1 . 1 and dau-1 . 2 , is also involved in the inhibition of dauer development . In this context it is important to note that even the double knockout phenotype of dau-1 . 1 and dau-1 . 2 ( Table 1 x ) does still not result in 100% dauer formation , a function that might well be attributed to dal-1 . If so , it is interesting to note that such a potential function of dal-1 in P . pacificus would be independent of the CAN neurons , as dau-1 . 1; dau-1 . 2 double mutants completely mimic the CAN ablation phenotype . Our results provide novel aspects of the cellular and genetic mechanisms of nematode dauer development . Besides the identification of dau-1 , we show a role of the CAN neuron in the regulation of P . pacificus dauer formation . The P . pacificus CAN neuron is similar to the corresponding cell in C . elegans , both by position and form . However , the function of both cells differ: in C . elegans the CAN neuron is essential for viability and CAN ablation results in the death of the larvae suggesting a function in osmoregulation . In P . pacificus , CAN ablation is viable facilitating the identification of its role in dauer regulation . Thus , our work reveals new functions for a new gene , dau-1 , and a new function for a previously unconsidered cell . Additional studies are necessary to link the dau-1 expression pattern to other genetic components of the regulatory network . Unfortunately , the P . pacificus transgenic system only allows the introduction of DNA fragments smaller than 20 kb [30] , preventing us from studying the expression pattern of the Ppa-daf-12 locus , which is more than 40 kb in size . Our work on dau-1 provides a molecular mechanism for cross-preference and intraspecific competition among P . pacificus strains . While previous studies already indicated strong natural variation in pheromone production and sensing supporting intraspecific competition as a new role in nematode ecology , these studies did not investigate the molecular mechanisms involved in the integration of upstream pheromone variation into the dauer regulatory network [11 , 12] . The results of our epistasis analysis clearly indicate that dau-1 acts downstream of small-molecule pheromones at the level of endocrine hormone signaling , which is known as convergence point of various signaling inputs in C . elegans dauer regulation . Therefore , the regulation of daf-12 emerges as a key principle in dauer development and evolution . In this context it is important to note that daf-12 is the key developmental switch gene for the dauer vs . direct developmental decision and such developmental switch genes have long been predicted to represent major features of phenotypic plasticity in animals and plants [31] . Finally , while our work starts to identify the molecular mechanisms associated with natural variation and intraspecific competition between P . pacificus strains , this work does not touch on the ecological implications . We have previously shown that multiple , distinct P . pacificus haplotypes can be found on the same living beetle [4] and that even closely related strains can differ tremendously in dauer longevity and fitness [13] . The majority of the available more than 800 wild isolates of P . pacificus have been sampled as dauer stages from scarab beetles and all tested strains can still form dauers indicating that no strains represents a Daf-d phenotype . RS5134/Ohio represents an extreme example with high dauer induction , high dauer survival and relatively high fitness after dauer recovery , whereas RS2333/California shows much lower dauer induction , survival and recovery ( Fig 1 ) . In light of our molecular findings on dau-1 , these life history traits result in several ecological conclusions and questions . dau-1 cannot be the only molecular player involved in the observed natural variation pattern between strains , as only RS2333/California and RS106/Poland have the additional dau-1 copy . Also , the majority of strains have only one dau-1 copy but still differ in dauer induction and longevity indicating that additional factors must exist . In particular , molecular differences are to be expected in dauer physiology , a phenomenon little studied in both C . elegans and P . pacificus [20] . Future studies can adress differences in dauer physiology using transcriptomic approaches between strains and might identify mechanisms involved in the different survival pattern of nematode strains .
Nematodes were grown on nematode growth medium ( NGM ) agar plates with the E . coli strain OP50 as food source [32] . Crosses were performed using one J4 hermaphrodite and two males . Dauer pheromone was purified from nematode liquid cultures as described previously [13] . The dauer pheromone assay was modified for three different types of experiments . First , for assays with transgenic animals ( Table 1 a-h ) , a mixture of 190 μl water and 10 μl pheromone was distributed evenly over the surface of NGM agar plates . We spotted each plate with 20 μl kanamycin-treated OP50 [11] . Per plate , four young adult hermaphrodites were allowed to lay eggs overnight , producing approximately 100 progeny . After three days , dauer formation was calculated as the percentage of progeny that entered the dauer stage . Second , epistasis experiments with transgenic animals ( Table 1 i-p ) were performed as described above but using NGM agar plates lacking cholesterol . Third , to combine ablation experiments with dauer pheromone assays , the assay was modified to enable the use of J2 larvae instead of adults . The dauer formation of mutant animals was also tested using J2 larvae to obtain comparable results . For assays with ablated and mutant animals ( Table 1 q-z ) , a mixture of 180 μl water and 20 μl pheromone was distributed evenly over the surface of NGM agar plates lacking cholesterol . We spotted 15 μl kanamycin-treated OP50 and picked 50 J2 larvae onto each plate . After two days , dauer formation was calculated as the percentage of J2 larvae that entered the dauer stage . To test if the dau-1 . 1;dau-1 . 2 double mutant responds to DA , 15 μl of ∆7-DA were added to each plate , resulting in a final concentration of 15 μM . The mean dauer formation of three independent biological replicates was calculated for all experiments . No dauers were observed on control plates without pheromone . Competition assays were performed in Ussing chambers without adding external dauer pheromone [11] . For the competition experiment , RS2333/California was grown in liquid culture in one compartment of an Ussing chamber , while RS5134/Ohio was grown in the other compartment of the same chamber . As controls , a second chamber contained only RS2333/California in both compartments , and a third chamber contained only RS5134/Ohio . At the beginning of the assay , the nematodes from one fully- grown NGM agar plate , initially containing 10 J4 larvae , were washed off into one compartment . For each time point ( 7 , 8 , 9 , 10 , 11 , 12 , 13 , and 14 days ) , three replicates of a sample volume of 30 μl were taken from each compartment , and dauer formation was calculated as the percentage of dauers in the sample volume . The mean dauer formation of the three replicates was calculated for each time point . We repeated all experiments multiple times and obtained similar results . After crossing RS2333/California males with RS5134/Ohio , heterozygous F1 animals were inbred for 10 generations to generate 911 RILs ( S1A Fig ) . For each RIL , we determined the dauer formation phenotype in response to both parental pheromones . We selected 136 RILs with low , medium , and high dauer formation phenotypes for genotyping with simple sequence length and conformation polymorphism markers ( S1 Table ) . QTL mapping was performed using the qtl package of the program R [33] ( S1B Fig ) . Transgenic lines were generated as described previously [30] . A 9 . 5 kb genomic RS2333/California or RS5134/Ohio construct , consisting of a 3 . 3 kb upstream region , the 1 . 7 kb dau-1 gene , and a 4 . 5 kb downstream region , was injected into RS2333/California or RS5134/Ohio animals . To generate reporter lines , an RS2333/California translational reporter construct , containing a 4 . 7 kb upstream region and the dau-1 coding region driving RFP expression , was injected into RS2333/California animals . We obtained three independent reporter lines . Non- transgenic nematodes that had lost the transgenic array were used as controls in dauer pheromone assays . A RS2333/California line expressing only the RFP injection marker was generated as an additional control . Cell ablation was performed as described previously [34] . The CAN neurons were ablated in freshly hatched J2 larvae , which were then immediately transferred to dauer pheromone assay plates . After determining the dauer formation phenotype , animals were checked to confirm the absence of CAN neurons . Nematodes , in which the CAN neurons were still present , were excluded from the calculation of the mean dauer formation of the three biological replicates . CRISPR/Cas9 induced gene inactivation was performed as described previously [23] . After injection , 576 F1 animals were screened for deletions by Sanger sequencing . We obtained two mutant lines ( tu490 and tu491 ) with a 7 bp and a 5 bp deletion in dau-1 . 1 and one mutant line ( tu492 ) with a 10 bp deletion in dau-1 . 2 . Deletion mutants were backcrossed multiple times , and double and triple mutants were generated by Mendelian genetics . To analyze the genomic region associated with the QTL ME25944 for SNPs , deletions , and duplications , we used the genomic resequencing data from 104 P . pacificus strains [22] . The dau-1 locus was identified as being duplicated in RS2333/California when compared to RS5134/Ohio by the software cnv-seq based on a significant difference in read coverage ( P<10–39 ) [35] . For comparisons of gene expression levels , we prepared and analyzed RNA-seq libraries as described previously [36] . Quantification of expression levels as fragments per kilobase transcripts per million fragments sequenced showed strong similarity between the two strains ( Spearman's ρ = 0 . 75 , P<10−16 ) . 1298 genes were identified as being differentially expressed ( FDR<0 . 05 ) [37] , among which we found the gene prediction corresponding to dau-1 . 1 ( Contig44- snap . 18 , version Hybrid1 ) . To investigate the evolutionary history of dau-1 , we reconstructed maximum likelihood trees using homologous sequences from other diplogastrid nematodes: gene predictions for P . exspectatus [22] , gene predictions for P . arcanus , and a transcriptome assembly for Parapristionchus giblindavisi . Protein sequences were aligned using the MUSCLE aligner ( version v3 . 8 . 31 ) [38] , and maximum likelihood trees were calculated using the phangorn R package [39] after selection of the best model using the ProtTest3 webserver [40] . 95% confidence intervals were calculated for the mean dauer formation values obtained in dauer pheromone assays and competition assays . To compare dauer formation of two strains , Fisher's exact test was performed using the program R ( www . r-project . org ) , and p values less than 0 . 05 were considered statistically significant .
|
The nematode dauer stage represents the major survival and dispersal strategy , and had a remarkable impact in the evolutionary and ecological success of nematodes . Our recent work in Pristionchus pacificus revealed substantial natural variation in various aspects of dauer development , i . e . pheromone production and sensing and dauer longevity and fitness , including a strain from Ohio with extremely long-lived dauers , very high fitness and high dauer formation in response to other strains´ pheromones . However , the molecular mechanisms associated with natural variation in dauer development are currently unknown . Using quantitative-trait-loci analysis , we discovered that the orphan gene dauerless controls dauer formation by copy number variation . Strains with one dauerless copy show high dauer formation , whereas strains with two copies have strongly reduced dauer formation . Transgenic animals expressing multiple copies do not form dauers . dauerless is exclusively expressed in CAN neurons , and both CAN ablation and dauerless mutations increase dauer formation . Strikingly , dauerless underwent several duplications and acts in parallel or downstream of steroid-hormone signaling but upstream of the nuclear-hormone-receptor daf-12 . Our findings reveal the importance of gene duplications and copy number variations for orphan gene function and suggest daf-12 as major target for dauer regulation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
The Orphan Gene dauerless Regulates Dauer Development and Intraspecific Competition in Nematodes by Copy Number Variation
|
Trypanosoma cruzi has been subdivided into seven Discrete Typing Units ( DTUs ) , TcI-TcVI and Tcbat . Two major evolutionary models have been proposed to explain the origin of hybrid lineages , but while it is widely accepted that TcV and TcVI are the result of genetic exchange between TcII and TcIII strains , the origin of TcIII and TcIV is still a matter of debate . T . cruzi satellite DNA ( SatDNA ) , comprised of 195 bp units organized in tandem repeats , from both TcV and TcVI stocks were found to have SatDNA copies type TcI and TcII; whereas contradictory results were observed for TcIII stocks and no TcIV sequence has been analyzed yet . Herein , we have gone deeper into this matter analyzing 335 distinct SatDNA sequences from 19 T . cruzi stocks representative of DTUs TcI-TcVI for phylogenetic inference . Bayesian phylogenetic tree showed that all sequences were grouped in three major clusters , which corresponded to sequences from DTUs TcI/III , TcII and TcIV; whereas TcV and TcVI stocks had two sets of sequences distributed into TcI/III and TcII clusters . As expected , the lowest genetic distances were found between TcI and TcIII , and between TcV and TcVI sequences; whereas the highest ones were observed between TcII and TcI/III , and among TcIV sequences and those from the remaining DTUs . In addition , signature patterns associated to specific T . cruzi lineages were identified and new primers that improved SatDNA-based qPCR sensitivity were designed . Our findings support the theory that TcIII is not the result of a hybridization event between TcI and TcII , and that TcIV had an independent origin from the other DTUs , contributing to clarifying the evolutionary history of T . cruzi lineages . Moreover , this work opens the possibility of typing samples from Chagas disease patients with low parasitic loads and improving molecular diagnostic methods of T . cruzi infection based on SatDNA sequence amplification .
Trypanosoma cruzi , the causative agent of Chagas disease , has been subdivided into seven Discrete Typing Units ( DTUs ) , TcI-TcVI and Tcbat , which have been associated with different geographic distribution and transmission cycles [1–4] . Historically , the genetic diversity displayed by T . cruzi was attributed to predominant clonal evolution [5–7] . However , an increasing number of evidences indicates that natural genetic exchange may be frequent and has had a fundamental role in the evolution of T . cruzi DTUs [8–11] . Thereby , while the theory of clonal evolution may explain the usual mode of T . cruzi population expansion , it is widely accepted that hybridization events had an important impact on the current population structure of this parasite , including the existence of hybrid lineages [9 , 11–14] . Two major evolutionary models have been proposed to explain the origin of hybrid lineages , the "Two Hybridization" [12] and the "Three Ancestor" [13] models . After analyzing nine nuclear loci from 26 isolates representative of T . cruzi DTUs TcI-TcVI , Westenberger et al . ( 2005 ) [12] postulated the hypothesis that an ancient fusion between TcI and TcII strains led , through a loss of TcI/II hybrid heterozygosity and independent clonal evolution , to the origin of TcIII and TcIV; later on a more recent hybridization event between TcII and TcIII strains generated TcV and TcVI by independent clonal evolution . On the other hand , following the analysis of five microsatellite loci and three mitochondrial genes from 75 TcI-TcVI stocks , Freitas et al . ( 2006 ) [13] proposed the existence of at least three ancestral lineages ( TcI-TcIII ) and that two recent and independent genetic exchange events between TcII and TcIII strains resulted in TcV and TcVI; whereas the origin of TcIV could not be fully addressed due to the few isolates analyzed from this DTU . Recently , a third and more complex scenario has been postulated by Tomasini and Diosque ( 2015 ) [14] . Thirteen housekeeping genes from 25 isolates representing T . cruzi DTUs TcI-TcVI were analyzed , as well as data published by other authors , including those from Westenberger et al . ( 2005 ) [12] and Freitas et al . ( 2006 ) [13] . They proposed that a common T . cruzi ancestor diverged into two groups: TcII and TcI-TcIII-TcIV , followed by TcIV separation and diversification into South ( TcIVS ) and North ( TcIVN ) American populations . Subsequently , TcI and TcIII diverged from the previous ancestor and TcIII received TcIVS mitochondrial DNA by multiple introgression events; phenomenon also described by others [9 , 15] . Finally , as it was proposed for the Three Ancestor model , two recent and independent hybridization events between TcII and TcIII led to the origin of TcV and TcVI [14] . T . cruzi satellite DNA ( SatDNA ) , widely used as target for molecular diagnostics of Chagas disease [16–20] , comprises 195 bp units organized in tandem repeats of about 30 ± 10 kb in some chromosomes [21] and constitutes approximately 5% of parasite genome [22] . Phylogenetic inference from 100 SatDNA sequences from TcI-TcIII and TcVI stocks showed that TcIII and TcVI sequences were distributed into TcI and TcII clusters , supporting the Two Hybridization model [23] . In a more recent network genealogy analysis of 139 SatDNA sequences from TcI-TcIII and TcV-TcVI stocks it was found that all TcIII sequences , including those from Elias et al . ( 2005 ) [23] , were grouped together with TcI sequences [24] . However , in the light of the Two Hybridization model the authors suggested that TcII SatDNA fingerprints were present in the ancestral TcIII but have been smudged in current TcIII strains , as proposed for other TcII genes [8 , 12 , 25] . Herein , we have gone deeper into this matter analyzing SatDNA sequences from T . cruzi DTUs TcI-TcVI , including TcIVS and TcIVN isolates , for phylogenetic inference . In addition , we performed a signature pattern analysis to identify polymorphic sites associated to specific T . cruzi lineages and designed new primers for molecular diagnostic purposes .
The T . cruzi stocks used in this work ( Table 1 ) came from already-existing collections ( see Acknowledgments section for details ) . The DTU classification of these isolates has been previously reported [1 , 26] and was confirmed using a Multiplex qPCR assay with TaqMan probes targeted to nuclear and mitochondrial genomic markers , as previously described [26] . Epimastigote forms of T . cruzi stocks were cultured in liver infusion-tryptose medium with 10% fetal calf serum ( NATOCOR , Cordoba , Argentina ) at 28°C , as previously described [27] . Parasite genomic DNA was purified using the High Pure PCR Template Preparation Kit ( Roche Diagnostics , Indianapolis , IN ) according to manufacturer instructions for cultured cells . Ten μg of genomic DNA was digested with FastDigest SacI restriction enzyme ( Thermo Scientific , Waltham , MA ) . The 195 bp fragments were purified from agarose gels using Wizard SV Gel and PCR Clean-Up System ( Promega , Madison , WI ) and cloned in the pBluescript SK ( - ) plasmid ( Stratagene , La Jolla , CA ) at the SacI site . After SatDNA qPCR confirmation [20] , the positive clones were sequenced using M13 forward primer ( MACROGEN , Seoul , Korea ) . In addition to the 201 sequences obtained in this work , 134 SatDNA sequences from eight T . cruzi stocks were downloaded from GenBank ( Table 1 ) . Sequences were aligned using ClustalX v2 . 1 [28] and edited with BioEdit v7 . 0 [29] . Phylogenetic tree was built using Bayesian inference with MrBayes v3 . 2 [30] . Analysis was performed using an appropriate substitution model according to the Akaike Information Criterion , estimated with jModelTest v2 . 1 [31] . Analysis was run for 100 million generations and sampled every 50000 generations , in the CIPRES Science Gateway server [32] . Convergence was assessed by effective sample size values higher than 200 using Tracer v1 . 6 [33] , and the initial 10% sampling was discarded as burn-in . In addition , genetic distances within and between phylogenetic clusters and T . cruzi DTUs sequences were estimated using an appropriate substitution model according to the Akaike Information Criterion with MEGA v7 . 0 [34]; standard error was estimated by bootstrap analysis ( 1000 replicates ) . Sequences were analyzed using VESPA v1 . 0 [35] to identify SatDNA type signature patterns associated to phylogenetic clusters . To validate our findings , the consensus sequence of each T . cruzi stock was obtained with BioEdit v7 . 0 [29] , considering all nucleotides present in at least 20% sequences for each polymorphic site , and classified into SatDNA types according to its similarity to a particular signature pattern . Finally , the SatDNA type of each consensus sequence was compared with the corresponding DTU of each T . cruzi stock . The graphic representation of the SatDNA consensus sequence of all T . cruzi stocks was obtained using WebLogo [36] . Based on it , specific SatDNA primers cruzi1c ( 5'-TGAATGGYGGGAGTCAGAG-3' ) and cruzi2c ( 5'-ATTCCTCCAAGMAGCGGAT-3' ) were designed for being used in a real-time PCR assay together with cruzi3 TaqMan probe [17] . Primer properties and specificity were verified using OligoAnalyzer v3 . 1 ( available at http://www . idtdna . com ) and BLAST search against non-redundant database [37] , respectively . The novel cruzi1c/cruzi3/cruzi2c SatDNA qPCR assay was compared with the validated SatDNA qPCR method [20] . Both qPCR assays were performed as previously described [20] , in simplex format , and challenged against genomic DNA from 12 T . cruzi stocks [TcI ( Sylvio X10 cl1 and K-98 stocks ) , TcII ( Y and JG stocks ) , TcIII ( M5631 cl5 and LL051-P24-Ro stocks ) , TcIV ( 4167 and Am64 stocks ) , TcV ( MN cl2 and PAH179 stocks ) , and TcVI ( CL Brener and RA stocks ) ] , in concentrations that ranged from 1 . 0 to 0 . 0625 fg/μL . Both amplifications were carried out simultaneously in an ABI7500 real-time PCR device ( Applied Biosystems , Foster City , CA ) .
Bayesian phylogenetic analysis showed that all 335 SatDNA sequences were grouped in three major clusters named TcI/III , TcII and TcIV ( Fig 1 ) , which corresponded to sequences from TcI and TcIII , TcII , and TcIV , respectively; whereas TcV and TcVI stocks had two sets of sequences distributed into TcI/III and TcII clusters . No monophyletic subgroup was found within TcIV cluster for sequences from TcIVS ( 4167 and Am64 ) and TcIVN ( Dog Theis ) stocks . TcI/III was the cluster with more sequences ( 179 ) , more than TcII ( 105 ) and TcIV ( 51 ) together , and also the one with the most complex topology; containing two main subgroups of 18 and 66 sequences . The first one was highly supported [posterior probability ( PP ) = 1] and comprised SatDNA sequences from all TcIII , two TcV ( B147 and NR cl3 ) and two TcVI ( RA and VD ) stocks . The second one ( PP = 0 . 66 ) was the most diverse and included sequences from all TcI , TcIII and TcVI , and two TcV ( 115 and B147 ) stocks . Table 2 details the TcI/III and TcII clusters distribution of SatDNA sequences from TcV and TcVI hybrid T . cruzi stocks . As shown , while TcV stocks 115 and NR cl3 had the highest rates of SatDNA type TcI/III and TcII sequences , respectively , B147 showed a similar distribution of sequences into both clusters . In the case of TcVI stocks , RA and VD ranged between 60 and 40% of SatDNA type TcI/III and TcII sequences , respectively , Tulahuen showed the opposite distribution and CL Brener had twice more SatDNA type TcII than TcI/III sequences . Estimation of genetic distances [Mean of nucleotide substitutions per site ( ± Standard Error ) ] within clusters gave values of 5 . 6 ( ± 1 . 1 ) , 4 . 0 ( ± 0 . 8 ) and 6 . 3 ( ± 1 . 1 ) % for TcI/III , TcII and TcIV clusters , respectively . On the other hand , the comparison between clusters gave higher distances between TcI/III and TcIV clusters [9 . 5 ( ± 2 . 9 ) %] than between TcI/III and TcII [7 . 3 ( ± 2 . 2 ) %] , and TcII and TcIV clusters [6 . 9 ( ± 2 . 1 ) %] . When genetic distances were analyzed grouping sequences by T . cruzi DTUs , similar estimates were obtained within TcI [5 . 3 ( ± 1 . 1 ) %] , TcII [4 . 0 ( ± 0 . 8 ) %] , and TcIII [5 . 2 ( ± 1 . 1 ) %] compared to their corresponding clusters; whereas both TcV and TcVI had the highest value [8 . 6 ( ± 1 . 6 ) %] . As expected , a low genetic distance was found between sequences from TcI and TcIII stocks , and higher values were observed when they were compared with those from TcII stocks ( Table 3 ) . In general , TcV and TcVI sequences were very similar between them and closer to TcII than to TcI and TcIII sequences; whereas TcIV sequences were more distant to those from TcI and TcIII stocks than to TcII , TcV and TcVI sequences . The signature patterns identified for the three clusters of SatDNA sequences are shown in Table 4 . Fifteen polymorphic sites associated to one or two types of SatDNA sequence were found . All together , they defined a specific signature pattern for each type of SatDNA sequence ( TcI/III , TcII and TcIV ) . Only for two sites ( 16 and 129 ) the predominant nucleotide was able to resolve among the three types of SatDNA sequences , but even in these sites it was possible to find sequences from one cluster with a less frequent nucleotide that was predominant in sequences from another cluster . To validate these findings , the consensus sequence of each T . cruzi stock was obtained and the result of SatDNA type classification was compared with its corresponding DTU ( Table 5 ) . For each consensus sequence , the polymorphic sites associated to a particular SatDNA type were counted and the sequences classified on the basis of the SatDNA type that reached the highest score . For homozygous lineages ( TcI-TcIV ) , SatDNA type classification completely matched with the reported DTU of each T . cruzi stock; although , as expected , for TcI and TcIII stocks it was not possible to resolve between both DTUs and , in consequence , they were classified as SatDNA type TcI/III . Similarly , T . cruzi stocks of heterozygous lineages ( TcV and TcVI ) , harboring both types of SatDNA sequences TcI/III and TcII , could not be assigned to a specific DTU and were classified as SatDNA type Hybrid . Fig 2 shows the graphic representation of the consensus sequence of the 335 SatDNA sequences analyzed in this work . Based on the conserved regions and polymorphic sites associated to SatDNA types , cruzi1c and cruzi2c primers were designed to amplify a segment of 98 bp , aimed to be used in a SatDNA qPCR assay , with cruzi3 TaqMan probe , for molecular diagnostic purposes . The novel cruzi1c/cruzi3/cruzi2c SatDNA qPCR assay was compared with the validated cruzi1/cruzi3/cruzi2 qPCR method , against a set of dilutions of genomic DNA from 12 T . cruzi stocks representing DTUs TcI-TcVI ( Table 6 ) . As shown , the new qPCR assay was more sensitive than the previous method for all T . cruzi DTUs , except for TcII stocks , for which both qPCRs gave similar sensitivity . In particular , the most remarkable differences between the results of both qPCR assays were found for Sylvio X10 cl1 ( TcI ) , 4167 ( TcIV ) , Am64 ( TcIV ) , and MN cl2 ( TcV ) stocks .
Two major evolutionary models have been proposed to explain the origin of T . cruzi hybrid lineages , but while it is widely accepted that TcV and TcVI are the result of genetic exchange between TcII and TcIII strains , the origin of TcIII and TcIV is still a matter of debate . Thereby , the main difference between the Two Hybridization [12] and the Three Ancestor [13] models is whether or not TcIII and TcIV were originated from a hybridization event between TcI and TcII strains , respectively . Accordingly , the acquisition of TcI alleles via TcIII by TcV and TcVI hybrid lineages is supported by some authors [8 , 12 , 25] , whereas others have found no evidence of it [13 , 14 , 38] . Likewise , previous analysis of SatDNA sequences have also shown contradictory findings concerning to the existence of SatDNA type TcII sequences in TcIII isolates [23 , 24]; whereas no TcIV sequence has been analyzed yet . In the present work , we have gone deeper into this matter analyzing 335 distinct SatDNA sequences from 19 T . cruzi stocks representing DTUs TcI-TcVI , including TcIVS and TcIVN isolates ( Table 1 ) . All TcI and TcIII sequences were grouped in the same cluster , whereas TcV and TcVI sequences were distributed into TcI/III and TcII clusters ( Fig 1 ) , as previously described [24] . The fact that none TcIII sequence clustered with TcII sequences and that all TcIV sequences were grouped in a unique and independent cluster support the theory that TcIII and TcIV are not the result of a hybridization event between TcI and TcII , in concordance with the Three Ancestor model [13] and other authors [14 , 38] . The highly supported monophyletic subgroup including TcIII but not TcI sequences found in TcI/III cluster may reflect a remnant of an ancestral hybridization event during the origin of TcIII; however , no evidence was found that it could involve TcII strains , as previously proposed [24] . Although we cannot deny this possibility , it seems unlikely that TcII SatDNA fingerprints have been smudged in current TcIII strains , whereas the remnant of the ancestral hybridization event observed in TcIII have also been found in TcV and TcVI stocks . The closeness between TcI and TcIII sequences found by phylogenetic inference and genetic distance estimation ( Table 3 ) is in concordance with previous analysis of nuclear genome data indicating that both DTUs share a common ancestor [14] or the participation of TcI in an ancestral hybridization event yielding TcIII [24] . On the other hand , the fact that TcIV does not share any SatDNA sequence with TcI and TcIII , and vice versa , and the highest genetic distances found between them , contrast with the theory that these three DTUs emerged from the same ancestor [14] . It is worth noting that TcIV sequences were not split up into TcIVS and TcIVN subgroups , as it has been described for other genetic markers [9 , 14]; although this DTU was found as the homozygous lineage with the highest intra-DTU diversity . Concerning to TcI , despite its well-known genetic diversity [39 , 40] and the fact that four isolates from different geographic regions were analyzed , this lineage showed an intermediate intra-DTU distance . Finally , the lowest genetic diversity of TcII is in concordance with previous analysis of other nuclear sequences [9] , but it could also be due to the fact that only two TcII isolates were included in this work and both came from the same geographic region . The presence of TcII and TcIII SatDNA fingerprints in TcV and TcVI stocks , supports the accepted theory that both DTUs are the result of genetic exchange between TcII and TcIII strains [9 , 12 , 13] . The heterozygosity and the common parental ancestors of TcV and TcVI explain the fact that both lineages had the highest intra-DTU diversity and showed the lowest genetic distance between T . cruzi DTUs ( Table 3 ) . The high diversity of SatDNA sequences within TcV and TcVI contrasts with the homogeneity found in both DTUs analyzing other nuclear markers [9] , possibly due to that SatDNA belongs to the fast-evolving portion of eukaryotic genomes [41] . The different rates of SatDNA type TcI/III and TcII sequences among TcV and TcVI stocks ( Table 2 ) , may reflect: i ) independent clonal evolution since both DTUs were originated , ii ) several independent hybridization events between different TcII and TcIII strains that led to TcV and TcVI strains with distinct SatDNA content , iii ) genetic exchange between hybrid progeny and parental lineages , or iv ) a combination of these and other possible scenarios . The remarkable genetic diversity as well as the different geographic distribution and transmission cycles of T . cruzi DTUs make their identification a matter of great interest for ecological , epidemiological and clinical studies [2 , 3] . Several strategies have been proposed to genotype T . cruzi isolates but , due to sensitivity constraints , most of these methods have been applied only to cultured stocks and biological or clinical samples with high parasitic loads [26 , 42–47] . The low sensitivity of these strategies resides in the single or low copy number of their target sequences , therefore the significant impact that the use of molecular targets with high copy number like SatDNA sequence may have on clinical and epidemiological genotyping studies . No consensus motifs of SatDNA sequence have been found for any T . cruzi DTU [24] . Following a different approach , we have identified specific SatDNA TcI/III , TcII and TcIV signature patterns ( Table 4 ) . Although further validation will be necessary to implement SatDNA typing , the perfect coincidence between SatDNA classification and the reported DTU for 19 T . cruzi stocks ( Table 5 ) supports its application in genotyping studies; principally when the usual methods fail . The major limitation of this approach is that it cannot distinguish between the presence of hybrid lineages TcV and TcVI , and the existence of mixed infections with TcI or TcIII and TcII strains . However , in these cases the epidemiological characteristics of the region from where the sample is taken may help to fill this gap . This typing strategy could be particularly useful for chronic Chagas disease patients with low parasitic loads , whose samples are usually very difficult to genotype [26 , 48] . Indeed , a first version of this approach allowed the characterization of samples from chronic patients that gave non-detectable results using traditional genotyping methods [49] . Based on the polymorphism of minicircle hypervariable regions , a highly repetitive sequence from kinetoplastid DNA ( kDNA ) , a minicircle lineage-specific PCR assay has been developed to detect the presence of single or mixed infections of TcI , TcII , TcV and TcVI in clinical samples [50 , 51] . Considering that the analysis of SatDNA or kDNA sequences does not allow the identification of all T . cruzi DTUs , as for most molecular markers , a combined strategy using both repetitive sequences might help to fully resolve the genotyping of clinical samples with low parasitic loads . Due to its high copy number , SatDNA sequence has been one of the most used targets for molecular diagnostics of T . cruzi infection [16–20] . However , the most used primers for conventional PCR ( Tcz1/Tcz2 ) [52] , Sybr Green real-time PCR ( SatFw/SatRv ) [18] , and TaqMan real-time PCR ( cruzi1/cruzi2 ) [17] approaches based on SatDNA amplification were designed long time ago when few sequences from TcI , TcII and TcVI , and almost none from TcIII , TcIV and TcV isolates were available . Therefore , we were interested in revising the suitability of these primers and , in case of being necessary , designing new ones . Except for SatRv , all the target sequences of these primers include polymorphic sites that were not considered in their design ( Fig 2 ) . In particular , the target sequences of Tcz1 and cruzi1 include polymorphic sites associated to specific SatDNA types and , in consequence , could be leading to misdiagnose infections with some TcI , TcIII and TcIV strains; principally in patients with low parasitic loads . During the design of cruzi1c and cruzi2c primers , the amplicon size was reduced to improve TaqMan qPCR efficiency , as recommended [53] . Both considerations , avoiding polymorphic sites and reducing amplicon size , led to an improved sensitivity of SatDNA qPCR assay ( Table 6 ) . It is worth noting the higher sensitivity of the new qPCR assay against TcIV and some TcI and TcV stocks , compared to the validated qPCR method [20] . TcIV , traditionally associated with the sylvatic cycle and occasional oral outbreaks due to food contamination [2 , 3] , was recently found as the second more frequent DTU in Bolivian chronic Chagas disease patients living in Madrid , Spain [54]; indicating that the incidence of this DTU in clinical cases may have been underestimated . The lowest sensitivity of both qPCR methods against TcI stocks was probably due to the lowest copy number of SatDNA sequence observed in strains from this DTU [55] . Analytical and clinical validation studies will be necessary before using the new qPCR assay for the molecular diagnostics of T . cruzi infection . Summarizing , our findings support the theory that TcIII is not the result of a hybridization event between TcI and TcII , and that TcIV had an independent origin from the other DTUs , contributing to clarifying the evolutionary history of T . cruzi lineages . Moreover , this work opens the possibility of typing samples from Chagas disease patients with low parasitic loads and improving molecular diagnostic methods of T . cruzi infection based on SatDNA sequence amplification .
|
Trypanosoma cruzi , the causative agent of Chagas disease , is a genetically complex protozoan parasite . T . cruzi strains have been classified into seven Discrete Typing Units ( DTUs ) , TcI-TcVI and Tcbat , which have been associated with different geographic distribution and transmission cycles . Two major evolutionary models have been proposed to explain the origin of hybrid lineages , but the phylogenetic relationships between the different T . cruzi DTUs are not completely understood . On the other hand , the molecular markers usually used for DTU identification are found in a low copy number and do not allow the direct typing of biological samples with low parasite burden . However , this goal can be reached using molecular targets with highly repetitive sequences like T . cruzi satellite DNA ( SatDNA ) , widely used for molecular diagnostics of Chagas disease . This study has been focused on the phylogenetic analysis of SatDNA sequences from DTUs TcI-TcVI aimed to clarifying the origin of T . cruzi lineages . In addition , our findings open the possibility of typing clinical samples with low parasitic loads and improving SatDNA real-time PCR sensitivity for molecular diagnostics of T . cruzi infection .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"biogeography",
"ecology",
"and",
"environmental",
"sciences",
"medicine",
"and",
"health",
"sciences",
"parasite",
"evolution",
"population",
"genetics",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"parasitology",
"protozoans",
"forms",
"of",
"dna",
"dna",
"population",
"biology",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"geography",
"bioinformatics",
"phylogeography",
"evolutionary",
"genetics",
"trypanosoma",
"cruzi",
"biochemistry",
"trypanosoma",
"eukaryota",
"dna",
"sequence",
"analysis",
"nucleic",
"acids",
"database",
"and",
"informatics",
"methods",
"earth",
"sciences",
"genetics",
"biology",
"and",
"life",
"sciences",
"satellite",
"dna",
"evolutionary",
"biology",
"organisms"
] |
2017
|
New insights into Trypanosoma cruzi evolution, genotyping and molecular diagnostics from satellite DNA sequence analysis
|
This study investigated the transmission and prevalence of Leishmania parasite infection of humans in two foci of Visceral Leishmaniasis ( VL ) in Georgia , the well known focus in Tbilisi in the East , and in Kutaisi , a new focus in the West of the country . The seroprevalence of canine leishmaniasis was investigated in order to understand the zoonotic transmission . Blood samples of 1575 dogs ( stray and pet ) and 77 wild canids were tested for VL by Kalazar Detect rK39 rapid diagnostic tests . Three districts were investigated in Tbilisi and one in Kutaisi . The highest proportions of seropositive pet dogs were present in District #2 ( 28 . 1% , 82/292 ) and District #1 ( 26 . 9% , 24/89 ) in Tbilisi , compared to 17 . 3% ( 26/150 ) of pet dogs in Kutaisi . The percentage of seropositive stray dogs was also twice as high in Tbilisi ( 16 . 1% , n = 670 ) than in Kutaisi ( 8% , n = 50 ) ; only 2/58 wild animals screened were seropositive ( 2 . 6% ) . A total of 873 Phlebotomine sand flies were collected , with 5 different species identified in Tbilisi and 3 species in Kutaisi; 2 . 3% of the females were positive for Leishmania parasites . The Leishmanin Skin Test ( LST ) was performed on 981 human subjects in VL foci in urban areas in Tbilisi and Kutaisi . A particularly high prevalence of LST positives was observed in Tbilisi District #1 ( 22 . 2% , 37 . 5% and 19 . 5% for ages 5–9 , 15–24 and 25–59 , respectively ) ; lower prevalence was observed in Kutaisi ( 0% , 3 . 2% and 5 . 2% , respectively; P<0 . 05 ) . This study shows that Tbilisi is an active focus for leishmaniasis and that the infection prevalence is very high in dogs and in humans . Although exposure is as yet not as high in Kutaisi , this is a new VL focus . The overall situation in the country is alarming and new control measures are urgently needed .
In recent years reports have emerged of increased leishmaniasis transmission in Europe [1] , drug resistant leishmaniasis has spread further [2] , and the spread of HIV/leishmaniasis co-infection is a trend of particular concern [3] , [4] . Visceral Leishmaniasis ( VL ) is mainly caused by two species of parasites , the anthroponotic L . donovani and the zoonotic L . infantum , for which a variety of canids serve as the animal reservoir . Most infections are asymptomatic , although longitudinal follow-up has shown that some infected individuals eventually predispose to clinical disease . Malnutrition and immune suppression , notably HIV infection , predispose to clinical disease , and children are especially affected [5] . Zoonotic VL occurs in many former Soviet Union countries and presents one of the most serious public health concerns in Georgia [5] , [6] . Several natural VL foci have been identified in the country where various species of Phlebotomus and canine reservoirs facilitate the transmission ( see Figure 1 ) [7] . The first four cases of VL in Georgia were described in 1913 , which to the best of our knowledge was the first report about this disease in the entire Caucasus region . The first well-described outbreaks ( 540 cases ) of VL were reported in eastern Georgia in 1954 [6] , [8] . All cases were registered in 6 cities and 164 villages , mainly in the east of the country , in the Shida Kartly and Kakheti regions ( see Figure 1 ) . Malaria control efforts in Eastern Georgia in the sixties included massive spraying campaigns with the insecticide dichlorodiphenyltrichloroethane ( DDT ) [8] , which is believed to have also caused a significant reduction in the sand fly population , as during the next 40 years , until the 1990s , only sporadic VL cases were registered , and only in the extreme eastern part of the country . Since 2005 , 19 VL fatal cases have been registered in Georgia , usually the result of late diagnosis and/or misdiagnosis . VL traditionally affects children in the 1–5 years age group . However in the last 5 years a relatively high number of adults have been among the reported VL cases , indicating that the disease may be re-emerging as an epidemic instead of the previously low endemic situation [official disease records , National Centre for Disease Control ( NCDC ) , Tbilisi , Georgia] . The first VL case in Tbilisi was registered in 1990 [7] . Two cases of Leishmania/HIV co-infection were diagnosed in 2008; for both cases the outcome was fatal [official disease records , NCDC 2008] . Cutaneous leishmaniasis ( CL ) is less frequent: 125 CL cases were registered in the period 1928–1964 , of which 110 ( 88 . 0% ) occurred in Tbilisi and villages situated in the western part of the Mtkvari river valley –Shida Kartli region , 56 km west from Tbilisi ( see Figure 1 ) . After a long interval without registered cases , new cases of CL started to appear , and mandatory registration of CL started in 2001 . Between 2001 and 2007 , 1–5 cases of CL were reported in most years , which increased to 12 cases in 2008–2009 , 8 of them in Tbilisi [official disease records , NCDC] . Both CL and VL are underreported , due to their relatively recent re-emergence , a consequent lack of awareness in the population , and the lack of a leishmaniasis training program for medical doctors . In Kutaisi the first VL cases were only registered as recently as 2007 ( 3 cases ) [official disease records , NCDC 2007] in which year also the first sand flies of the Phlebotomus genus were identified in this area [G . Babuadze unpublished study] . During the last two decades , the annual number of clinical VL cases has remained persistently high , and varies between 122 and 189 . In the period 1995–2010 , 1919 VL cases were registered , including 1052 from Tbilisi [official disease records , NCDC] . Several natural VL foci have been identified , where several species of Leishmania vectors exist and canine reservoirs facilitate the transmission of this disease [7] . The situation is most urgent in the East Georgian city of Tbilisi , which has approximately 1 . 5 million inhabitants . Most of the natural VL foci are located in the very centre of the city , which are the oldest and most popular areas . Urban transmission is aided by the elongated shape of the city along the banks of the river Mtkvari ( Kura ) , especially in areas located close to the hills and forests from where wild animals , including jackals and foxes , often appear . This situation facilitates the synanthropy between wild canids and stray and domestic dogs as these animals are known as reservoirs for Leishmania [8] . Since 1997 all VL cases have been reported from the left ( Eastern ) bank of the Mtkvari ( see Figure 2 ) . In order to understand the spread of VL in Tbilisi and Kutaisi , three surveys were performed: ( 1 ) an infection screening of populations in selected districts of the two cities by Leishmanin Skin Test ( LST ) ; ( 2 ) a seroprevalence study of Leishmania infection in wild canines , stray and pets dogs; and ( 3 ) an entomologic identification of potential Leishmania vectors . We have identified alarmingly high prevalence rates in humans , vectors and dogs , especially in Tbilisi , although the emergence of VL in Western Georgia , where it has never been reported before , is almost equally serious .
Tbilisi is the capital and the largest city of Georgia ( 726 km2 , 1 , 480 , 000 inhabitants ) , situated in the South Caucasus at 41°43′ North Latitude and 44°47′ East Longitude , lying on the banks of the Mtkvari ( Kura ) River . Highest elevation is 770 m and the lowest is 380 m above the sea level . Kutaisi is Georgia's second largest city , in the western region of Imereti and has approximately 186 , 000 inhabitants . It is located along both banks of the Rioni River , 221 km west of Tbilisi . The city lies at an elevation of 125–300 meters . For this study , Tbilisi was divided in three districts ( see Figure 2 ) . The climate of the Tbilisi area can be classified as humid subtropical , with relatively cold winters and hot summers , which constitutes a good seasonal habitat for sand flies . Ethical clearance for conducting this study was secured by the Institutional Board of Review ( IRB ) of the National Center for Disease Control and Public Health ( IRB00002150 ) in compliance with Georgian legislation and international bioethical frameworks . All volunteers were interviewed and written informed consent was obtained for participation in the study . Leishmanin was obtained by WHO from the Pasteur Institute , Teheran , Iran , made with phenolized L . major promastigotes [9] . The LST detects CL [10] as well as asymptomatic infection and cured VL although VL patients with active disease are generally LST negative as a result of a strong humoral response [11] . As such the LST is generally employed to assess prevalence of infection in a population rather than disease levels [12] . The sensitivity of the LST for cutaneous leishmaniasis was estimated as 88% in a recent study [13] . For the LST survey ( during a 3 months period , June–August 2012 ) a cross-sectional study was carried out using a multi-stage cluster selection and probability proportional to size sampling [14] , [15] . District #1 consists of the central areas located in the central part of city situated on the right ( Western ) side of the Mtkvari River , and is bordered by green parks , hills and forest . In this district VL cases have been registered every year since 1997; District #2 includes the areas located in the eastern part of city ( left bank of the Mtkvari River ) , where VL cases have consistently been registered since 2000 ( incidence lower than in District #1 ) ; District #3 consists of the remaining parts of Tbilisi where VL cases also have been permanently registered since 2000 ( comparable less than in the first two districts ) [official disease records , National Center for Disease Control ( NCDC ) , Tbilisi , Georgia] . A total of 981 LST were performed within the thus-defined districts in Tbilisi ( 816 ) , and in Kutaisi ( 163 ) . After cleaning the skin over the flexor surface of the forearm with 70% alcohol , 0 . 1 ml of the antigen , containing standard amounts of Leishmania promastigotes , was administered intradermally employing single use insulin syringes . Skin test results ( indurations ) were read at 48 hours , although a small number was tracked for a 72 hours reading , according to the WHO recommendations [16] . Indurations were considered to be positive if they were at least 6 mm in diameter . In June–August 2011/2012 blood-serum specimens from 1571 asymptomatic dogs were sampled in Tbilisi ( 623 pet and 670 stray dogs ) and Kutaisi ( 151 pet and 50 stray dogs ) . The pet dog sampling was performed in the same areas where the LST screening was performed . Samples were taken from almost all pet dogs , with a few exceptions due to lack of consent from the owners . The blood from the stray dogs was collected thanks to the Municipal Service of Emergency and Urgent Situations of both Tbilisi and Kutaisi . Wild animals were captured alive and released after blood collection during a 2-month period ( September–October 2012 ) , with written permission from the Ministry of Environment of Georgia . Blood samples from 77 wild canines ( 38 foxes and 39 jackals ) were taken . Collected canine blood samples were placed into Vacutainer vials for serum and stored at 4°C . For the detection of VL antibodies in canine serum Kalazar Detect rK39 Rapid tests ( InBios International Inc , Seattle , USA ) were performed according to the manufacturer's instructions [17] , [18] . This test detects the circulating antibodies to recombinant K39 antigen of L . donovani-infantum complex and is highly specific ( 100% ) and sensitive ( 97% ) in diagnosing symptomatic and asymptomatic infections [19] , [20] . The test procedure involved adding 20 µl of serum to the absorbent pad on the bottom of the test strip . Test strips were placed into a well of a sterile 96 well plate , to which 2–3 drops of buffer solution ( provided with the test kit ) were added; results were read within 10 minutes . There is a variety of housing styles within the study area , including apartment blocks of 9 floors or more; however , some 90% are modest private homes constructed of brick or stone . Most of these are walled compounds , with courtyards and gardens containing a variety of trees including fruit trees . Within these areas , pens for animals including dogs , chickens and rabbits are frequent which offers a diversity of blood meal sources , resting and breeding sites for sand fly species . Windows and doors are unscreened , providing easy access for sand flies to residents . Vector surveys were implemented in Tbilisi ( June–October , 2011 ) and Kutaisi ( June–October , 2012 ) . The total number of collected sand flies was 873 ( 656 in Tbilisi and 217 in Kutaisi ) ; of these , 516 were female . Sandflies being phototrophic [21] , [22] Seven CDC miniature light traps ( John W . Hock Company , Gainesville , FL 32606 , USA ) were used to collect sand flies during three consecutive nights per month in each area [23] . Traps were placed outside houses , one per family compound , within fenced or protected habitats , especially near animal pens or in courtyards close to houses , between 7 pm and 7 am . Collected female sand flies were morphologically identified , dissected for detection of Leishmania parasites and scored according to various parameters: blood feed/unfed; gravid/non gravid [24] . Live female sand flies were removed from traps and transferred to 10–20% soapy water solution to clean and immobilize them; afterwards they were rinsed in clean distilled water and soaked for about 10 minutes in 1% sodium hypochlorite solution to disinfect them . Sand flies were dissected in sterile conditions ( sterile dissecting needles on a sterile microscope slide in a drop of sterile phosphate-buffered saline ( PBS ) ) according to the method of Lawyer et al . [25] . Two terminal segments of the sand fly abdomen containing the spermathecae and the guts were separated from the whole body . Midguts were transferred to a fresh drop of sterile PBS on another clean slide for identification . Each gut was covered with sterile , glass cover slip . In positive sand flies we observed a dense infection with many parasites attached to the microvillar lining of the midgut wall and to the cuticular intima of the stomodeal valve . Tightly packed Nectomonads and haptomonads were observed in the thoracic midgut , behind the stomodeal valve , which was forming a “plug” with a high proportion of easily motile metacyclic parasites that escaped into the sterile dissection buffer . They were distinguished by a small cell body with a long flagellum and fast movement . These motile cell forms were clearly visible and distinguishable from other elements under the 40× magnification . Data statistical analysis was performed using SPSS Statistics v20 software ( IBM ) . We analyzed variants to compare differences between the groups and determined statistical significance , with P values less than 0 . 05 .
The results with the Leishmanin Skin Test ( LST ) , summarised in Table 1 , revealed a high prevalence of LST positives in age groups 5–9 , 15–24 and 25–59 years in Tbilisi District #1 ( 22 . 2% , 37 . 5% and 19 . 5% , respectively ) . Prevalence for the same age groups was significantly lower in Kutaisi , at 0% ( P = 0 . 0062 ) , 3 . 2% ( P = 0 . 0018 ) and 5 . 2% ( P = 0 . 0017; all Χ2 test ) , respectively , and the overall difference in prevalence , 19 . 3% versus 7 . 3% , was also highly significant ( P = 0 . 00016 ) . Other notable results include high prevalence in Tbilisi District #2 for ages 10–14 ( 28 . 6%; n = 21 ) ; Tbilisi District #3 ages 1–4 ( 21 . 7%; n = 23 ) ; ages 60 and above in Kutaisi ( 17 . 24% ) ( Table 1 ) . While overall prevalence was not significantly different between Tbilisi districts ( P>0 . 05 ) , a clear difference was observed in total prevalence in Tbilisi compared with Kutaisi ( P = 0 . 0019 , Χ2 test; Table 1 ) . We further compared LST-positivity rates between male and female subjects in the same age groups ( Supporting Information Table S1 ) . Whereas prevalence in male subjects ( 15 . 2% overall ) and female subjects ( 11 . 7% ) was not significantly different , a significant difference was observed specifically in the 25–59 age group ( 16 . 2% males versus 7 . 8% females LST positive; P = 0 . 011 ) ; this may reflect a different behaviour pattern in working age subjects , with a larger percentage of men working outdoors in various occupations , in addition to males spending more social time outdoors; this is consistent with Georgian society , especially in an urban environment . Based on the results of the rK39 test , we found that the highest proportion of seropositive pet dogs is present in District #2 ( 82/292 pet dogs , 28 . 1% ) and District #1 ( 24/89; 27% ) in Tbilisi; in District #3 the percentage was 21 . 5% ( 52/242 ) . Surprisingly , the percentage was also quite high in Kutaisi , as 17 . 3% of pet dogs were positive ( 26/150 ) , even though the first few cases of human VL in this city were reported only a few years ago . Table 2 shows a breakdown of rK39-positivity rates in pets , stray dogs and wild canids showing a highly significant ( P = 5 . 3×10−7; 3-way X2 test ) difference between those groups , with almost a quarter of all pet dogs testing positive as opposed to only 2 . 6% in wild canids . A further stratification was applied for the pet dogs , by age and breed ( Table 2 ) , and a X2 test confirmed that age and breed had a significant ( P<0 . 001 ) influence on test outcome ( prevalence ) . The table further shows that in particular the age of the dog is a main determinant , with increased age increasing infection rate , rather than the size of the breed . We identified five Phlebotomus species of three subgenera in Tbilisi and 3 of these species were also found in Kutaisi . Sand fly infectivity rates , species composition and the primary vectors have previously been reported to be different on either side of the Mtkvari River [6] . Amongst the 5 Phlebotomus species in Tbilisi the most abundant were P . sergenti ( 43% of total ) and P . kandelakii ( 45% ) ; whereas no infected specimens of P . sergenti were found , P . kandelakii displayed an infection rate of 5 . 5% , with all the positives originating in District # 1 - ( Table 3 ) . In District #1 , situated on the right side of the Mtkvari River , the most abundant species was P . kandelakii , with 69 . 8% . In District #2 , only two sand flies species were identified and P . sergenti was by far the most abundant ( 91 . 8% ) . In the rest of Tbilisi ( Districts 3 ) P . sergenti was also the most abundant ( 84% ) Phlebotomus species but no infected sand flies were found in this area . The most abundant sand fly species collected in Kutaisi city were P . balcanicus ( 53 . 5% ) and P . halepensis ( 45 . 8% with infectivity rate of 1 . 3% ) - the two least prevalent Phlebotomus species in Tbilisi . These results show a very different sand fly population in the two cities .
We report here the first detailed survey of leishmaniasis prevalence in humans and canids , and of phlebotomine sand fly populations , in the two main cities of Georgia . Historical records as well as more recent clinical records show that leishmaniasis is more prevalent in the Eastern parts of the country . Consistent with this trend , we find highly significant differences in infection rates between the Eastern city of Tbilisi and the Western city of Kutaisi . Overall prevalence of infection , as measured by the standard LST test , was double the rate in Tbilisi ( P = 0 . 0019 ) ; this was significant for most age groups and where it did not reach significance , this was mostly due to smaller sample sizes in Kutaisi . Whereas LST does not detect 100% of all infections , especially during clinically active disease , potentially producing false negatives; however despite its long use , this test has never been associated with significant numbers of false positives . We thus contend that the numbers reported here are more likely to be an underestimation than an overestimation of true leishmaniasis prevalence in Georgia . The sand fly infection rate was quite high in Tbilisi and Kutaisi , 2 . 8% and 1 . 3% , respectively . The species of Phlebotomus were also different according to regions , with P . sergenti and P . kandelakii in Tbilisi and P . halepensis and P . balcanicus in Kutaisi . From this we surmise that the introduction of the disease to the Kutaisi region involved the adaptation of L . infantum to different vectors after the introduction of infected hosts to this region , with its distinct , endogenous sand fly population . There is no historical record of sandfly infestation in the Kutaisi area , but the last sandfly survey of Western Georgia we are aware of dates from 1956 [26] and this study reported only a very low number of P . chinensis , and only in the Zestafoni district ( approx 30 km South-east of Kutaisi ) . Apart from this , we have found no mention of leishmaniasis , or its vectors , in Kutaisi or the surrounding region in Western Georgia in the literature or in clinical records [8] , [26] . Here , we report specimens of both P . halepensis ( 43% of total sand fly population in Kutaisi ) and P . balcanicus ( 56% of total ) infected with Leishmania parasites in Western Georgia ( first time described in the country ) . Using the models of Pampiglione et al . for Mediterranean leishmaniasis [27] , [28] , the low seropositivity rate in the 15–24 and 25–59 age groups in Kutaisi , relative to age groups 1–4 and 10–14 ( Table 1 ) , suggests that this VL focus was developed relatively recently , as in old foci infection rates increase with age . Although we find the highest incidence in the 60-plus age group , this represents only 5 positives and we would hesitate to present this as evidence that Kutaisi is in fact an old focus . As the overall prevalence in Kutaisi is lower than in Tbilisi , the Kutaisi focus is also clearly less active but we need to emphasize that the results presented here constitute the first research on VL in Kutaisi since the first case of this disease was reported there , in 2007 [official statistical data of NCDC] , and constitute a relatively small dataset . The presence of infected P . halepensis in Western Georgia is potentially significant for a further reason: while this species is a suspected vector for L . infantum [29] it is also a suspected vector for L . major , and experimental infections with this species , at least , have been reported [30] While there is no proof of transmission of L . major in Georgia this possibility has never been investigated , and the presence of a potential vector in both eastern and western Georgia is of concern . Pratlong et al recently described the epidemiological features of Old World cutaneous leishmaniasis foci , based on the isoenzyme analysis of 1048 strains , and list a confirmed L . major strain [31] ( as well as an L . donovani strain [32] ) from Georgia . While it is not known whether this strain originated from local transmission or from a traveller infected elsewhere , 5 to 10 cases of cutaneous leishmaniasis are annually reported in Georgia [official disease records , National Center for Disease Control ( NCDC ) , Tbilisi , Georgia] . However it is not currently known whether this is caused by L . infantum , which can also be responsible for cutaneous leishmaniasis [33] , [34] , or by L . major as suggested in the publication by Pratlong [31] . Within Tbilisi , the highest prevalence of LST positive subjects was found in central districts of the city situated on the right side of Mtkvari River ( District 1 ) , which has the highest population density in Tbilisi . Although this failed to reach statistical significance from the other two city areas sampled it correlates with the high number of seropositive pet dogs in that area and the highest diversity of Phlebotomine sand fly species in this district ( Table 3 ) . The high LST prevalence among adult males compared to adult females in Tbilisi can be explained by their more frequent contacts with the vectors . The social pattern is that in summer time adult males spent a significant part of the evening with their neighbours and/or friends in either their own gardens or in nearby open spaces . At the same time most of adult females are doing housekeeping work and their social life is also far more indoors . A positive LST result is thought to indicate durable cell-mediated immunity after asymptomatic infection or clinical cure of VL and it persists in immunocompetent patients [35] . According to our observations during this study , for those dogs that were clinically suspected to have leishmaniasis , the rK39 test was weakly positive [19] . The seropositivity among pet dogs was found to be proportionate to their age as previously reported in other countries [18] . Seroprevalence was also high among stray dogs , however much less than in pet dogs ( P = 0 . 00017 ) , which means that the cycle is typically domestic . Indeed , most of the cases happen in houses with back gardens and chicken shelters , an appropriate environment for sand fly breeding , where dogs also cohabitate . A possible reason for the comparatively low prevalence in stray dogs may be the frequent movement of these animals within and outside the city , taking in areas of low population and/or vector density . Therefore , urgent control measures should focus on infected domestic dogs and vector control in the potential breeding sites around the house . We think that the roaming behaviour of stray dogs limits the contact with VL vectors but potentially contributes to the spread of the disease to new areas , developing new VL foci . The present study also reports for the first time that apparent VL being positive in the rK39 test was observed in wild foxes and jackals in Georgia . This appears to prove the hypothesis of Bardjadze , who suggested that these animals ( especially the fox ) are the reservoirs in non-permanent VL foci in this country [8] . However , it must equally be noted that we found only 2/77 wild canines ( 1 fox and 1 jackal ) to be positive for rK39 , a prevalence much below that of pet and stray dogs . This result seems consistent with the observations of Courtenay et al ( 2002 ) [36] , who also observed low infection rates in wild foxes and concluded these are not important for the spread of L . infantum infection . The transmission in active urban foci is thus from domestic dogs to human and , in the densely populated urban environment , appears to be much higher than in sylvatic environments . Thus , while we confirm that foxes and jackals do appear to carry leishmania parasites in non permanent foci of the Eastern part of the country they are not considered the main source of this disease in permanent foci such as Tbilisi and Kutaisi , in part due to their seasonal migration uplands in the summer , which is the disease transmission period . We determined 5 Phlebotomus species in the selected districts of Tbilisi . The sand fly season starts in the last weeks of June or early July and ends in the middle of September [8] . The sand fly population peaks in the middle of July and starts to come down after the middle of August . The number of these vectors is strongly dependent on climate and environmental conditions [6] , [7] . The data summarised in Table 3 show that vector composition was different on either side of the Mtkvari River , even though the overall infection rate was very similar ( P>0 . 05 ) . In District 1 , P . kandelakii was by far the most prevalent species ( 71 . 7% ) , whereas this species was not found in District 2 , where P . sergenti was dominant ( 94 . 1% ) ; P . sergenti was also the dominant sand fly species found in the rest of Tbilisi ( 77 . 3% ) . P . kandelakii was found to be infected with Leishmania parasites . No infected sand flies were found in District 3 ( ‘other Tbilisi’ ) , but this could reflect the comparatively low number of flies sampled in this area . This study shows that Tbilisi is an active focus for VL with very high infection prevalence in pets , in stray dogs and in humans as determined by the LST and rK39 tests . The infection rate in sand flies is also high , consistent with a recent report on the sand fly population in Tbilisi [37] . The microclimate of this city and the social behaviour of the population create conditions that are very favourable for the sand flies and for the spread of the infection . We demonstrate that almost the entire population of Tbilisi is at risk from VL , including all age groups , and in all districts , in large part because of the high percentage , and number , of seropositive dogs in the city . The outcome of the LST survey shows that a very significant percentage of the population has already been in contact with the parasite , although this does not imply that all will develop the clinical manifestations . The comprehensive survey described in this manuscript has for the first time documented the very significant risk to the Georgian population from visceral leishmaniasis , and that its transmission is spreading to the west of the country . The results obtained will allow the Georgian health authorities to initiate control measures to reduce the high urban transmission rates responsible for the current outbreak , which displays the largest dimensions in many years , and to formulate a national strategy for leishmaniasis prevention and for improving treatment efforts to protect its population and economy from this very severe disease , of which the capital is especially at risk . In order to arrive at a comprehensive national strategy , however , it will be necessary to expand this survey to other districts of Georgia and to map the environmental risk factors not only in known areas of transmission but also at the national level to anticipate the progression of the disease to other vulnerable areas .
|
Leishmaniasis is a disease complex of various clinical manifestations caused by infection with protozoan parasites ( Leishmania spp ) . It is transmitted through the bite of infected sand flies ( Phlebotomus or Lutzomyia spp ) and dogs are the main reservoir host for the Leishmania infantum species described previously in Georgia . It is prevalent in many tropical and subtropical regions of the world . In Georgia , visceral leishmaniasis has been known to occur in the East and in the capital , Tbilisi , but to date no investigation of vectors , and of prevalence in humans and in canines , has been conducted . Here , we report 5 different species of sand fly in Tbilisi and 3 in the West-Georgian city of Kutaisi , with infected vectors found in both places . In some districts of Tbilisi more than a quarter of pet dogs were seropositive for Leishmania parasites; prevalence in stray dogs was somewhat lower . Even in Kutaisi , where no leishmaniasis has previously been reported , 17 . 3% of pet dogs tested positive . This was reflected in high prevalence of infection in humans in the capital ( 14 . 5% overall ) , compared to 7 . 3% in Kutaisi . We conclude that the infection rate with visceral leishmaniasis in Georgia is alarmingly high and that its transmission has significantly spread west-wards in recent times .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"disease",
"epidemiology",
"epidemiology"
] |
2014
|
Epidemiology of Visceral Leishmaniasis in Georgia
|
Single genome sequencing of early HIV-1 genomes provides a sensitive , dynamic assessment of virus evolution and insight into the earliest anti-viral immune responses in vivo . By using this approach , together with deep sequencing , site-directed mutagenesis , antibody adsorptions and virus-entry assays , we found evidence in three subjects of neutralizing antibody ( Nab ) responses as early as 2 weeks post-seroconversion , with Nab titers as low as 1∶20 to 1∶50 ( IC50 ) selecting for virus escape . In each of the subjects , Nabs targeted different regions of the HIV-1 envelope ( Env ) in a strain-specific , conformationally sensitive manner . In subject CH40 , virus escape was first mediated by mutations in the V1 region of the Env , followed by V3 . HIV-1 specific monoclonal antibodies from this subject mapped to an immunodominant region at the base of V3 and exhibited neutralizing patterns indistinguishable from polyclonal antibody responses , indicating V1–V3 interactions within the Env trimer . In subject CH77 , escape mutations mapped to the V2 region of Env , several of which selected for alterations of glycosylation . And in subject CH58 , escape mutations mapped to the Env outer domain . In all three subjects , initial Nab recognition was followed by sequential rounds of virus escape and Nab elicitation , with Nab escape variants exhibiting variable costs to replication fitness . Although delayed in comparison with autologous CD8 T-cell responses , our findings show that Nabs appear earlier in HIV-1 infection than previously recognized , target diverse sites on HIV-1 Env , and impede virus replication at surprisingly low titers . The unexpected in vivo sensitivity of early transmitted/founder virus to Nabs raises the possibility that similarly low concentrations of vaccine-induced Nabs could impair virus acquisition in natural HIV-1 transmission , where the risk of infection is low and the number of viruses responsible for transmission and productive clinical infection is typically one .
Much of what is known about virus-host interactions underlying HIV-1 persistence and pathogenesis in humans has come from quantitative measurements and mathematical modeling of viral replication dynamics and virus evolution in response to selective pressures , including antiretroviral drug therapy and adaptive immune responses [1]–[9] . Comparable insights into HIV-1 transmission have been gleaned from analyses of acute infection viral sequences derived by single genome amplification ( SGA ) and interpreted in the context of a model of random virus evolution [8] , [10] , [11] . This latter approach makes possible an unambiguous molecular identification of actual transmitted/founder ( T/F ) viruses that are responsible for establishing productive clinical infection by HIV-1 in humans [8] , [10] , [12]–[15] and by SIV in rhesus macaques [16]–[18] . Importantly , because the SGA - direct amplicon sequencing strategy precludes Taq-polymerase mediated recombination and nucleotide misincorporation errors in finished sequences , it allows for the analysis of mutational linkage across complete viral genes and genomes [8] , [10] . Based on these considerations , we postulated that a precise molecular identification of T/F virus genomes and their evolving progeny could enable a comprehensive proteome-wide assessment of the earliest adaptive immune responses that shape and constrain the early replicating HIV-1 quasispecies . This hypothesis was affirmed for HIV-specific cytotoxic T-cell ( CTL ) responses in three acutely-infected subjects [7] , [8] . Here , we examined in the same three subjects whether this strategy could illuminate the earliest virus-specific neutralizing antibody ( Nab ) responses . Nabs constrain the replication of most viruses and are essential to the efficacy of most viral vaccines [19] , and this is presumed to be the case for HIV-1 [19]–[22] . There is clear evidence in Indian rhesus macaque models of SHIV infection that Nabs directed toward HIV-1 gp41 or gp120 can confer sterilizing immunity [23]–[25] . Further evidence in support of the protective potential of Nabs has come from heterologous low-dose mucosal SIV infection in Env-vaccinated rhesus macaques [26] . In humans , however , antibody correlates of protection from infection are still being identified [27]–[29] and the minimum titers of Nabs necessary to impede virus infection in vivo have not been determined , although it is clear that moderate and high titers of Nabs can lead to HIV-1 selection and escape [6] , [30]–[36] . The present study thus focused on four aspects of the Nab response in HIV-1 infected humans: ( i ) identification of genetic ‘footprints’ of the earliest detectable Nab responses to HIV-1; ( ii ) characterization of Env epitopes recognized by the earliest Nabs and molecular pathways of virus escape; ( iii ) determination of the titers of Nabs that are sufficient to select for virus escape in vivo; and ( iv ) viral replication fitness costs associated with Nab escape . Previous studies have addressed some of these same questions but with different experimental strategies that allowed for lesser degrees of molecular and dynamic resolution . Wei [6] and Richman [30] first used single-cycle Env trans-complementation assays to detect autologous strain-specific Nab responses , but neither study used SGA to identify T/F viruses or to look for genetic linkage of mutational escape pathways , nor did they use deep sequencing methods to detect the earliest escape mutations . Other investigators examined Nab responses against early viruses but without deep sequencing or a detailed kinetic analysis of low-titer antibody effects [32]–[37] . Here , we hypothesized that an in-depth kinetic analysis of the evolution of T/F viral env genes could provide for the most sensitive detection of Nab pressure on the replicating virus quasispecies — even before direct phenotypic detection of virus neutralization in vitro — and that such findings could be corroborated and extended by deep sequencing , site-directed mutagenesis , antibody adsorptions and in vitro testing of T/F and escape variant Env proteins for neutralization by autologous antibodies .
Subjects CH40 , CH77 and CH58 were each productively infected by single T/F viruses as demonstrated by SGA or 454 deep sequencing [8] , [38] . These T/F viruses exhibited a phenotype typical of primary HIV-1 strains including CD4 dependence , CCR5 tropism ( CH40 and CH58 ) or CCR5/CXCR4 dual tropism ( CH77 ) , resistance to CD4-induced and V3-specific antibodies , variable sensitivity to the broadly neutralizing antibodies b12 , 4E10 , 2F5 and HIVIG , and CD4+ T tropism [8] , [10] , [39] . We tested the three T/F viruses for neutralization sensitivity to autologous and heterologous plasma antibodies . Autologous neutralization titers ( IC50 ) at 3 , 6 and 12 month time points were 1∶1446/1∶2432/1∶1282 ( CH40 ) , 1∶38/1∶100/1∶239 ( CH77 ) and <1∶20/1∶48/1∶243 ( CH58 ) ( Figure S1; Table 1 ) . None of the plasma specimens exhibited heterologous neutralizing activity at dilutions as low as 1∶20 . The kinetics of appearance and magnitude of autologous Nab responses and the corresponding plasma viral load and CD4+ T cell measurements ( Figure S1 ) , were typical of HIV-1 subtype B infections [6] , [30] , [40] . In each subject , we examined the neutralization sensitivity of full-length infectious molecular clones ( IMCs ) of the T/F virus compared with IMCs of consensus 6 month sequences ( Figure 1 ) . The latter IMCs , like the T/F viruses , were all replication competent ( Figure S2 ) and contained phenotypically confirmed CTL escape mutations in addition to putative Nab escape mutations ( Figures 1A , 2–4 ) . We found neutralization of the T/F IMCs by 6 month plasma antibodies in titers comparable to those detected using Env pseudotyped viruses ( Table 1; Figure S1 ) . 6 month consensus IMCs containing putative Nab escape mutations in Env showed significant resistance to neutralization compared with T/F viruses and 6 month IMCs lacking putative Nab escape mutations ( p<0 . 05 for each , two-tailed paired t-test ) ( Figure 1B ) . Replication fitness costs resulting from Nab escape were analyzed in competitive replication assays by comparing T/F IMCs to 6 month IMCs with and without Nab escape mutations . These analyses suggested minimum costs to viral fitness from Nab escape mutations of between 0% and 24% ( Figure S2 and Dataset S1 ) . SGA sequencing of sequential plasma specimens was used to further characterize candidate Nab epitopes by looking for the earliest indications of Nab selection and escape across full-length gp160 env sequences . Figures 2–4 depict the temporal accumulation of nonsynonomous and synonomous env mutations in subjects CH40 ( Figure 2A ) , CH77 ( Figure 3A ) , and CH58 ( Figure 4A ) . For each subject , we determined the effect on neutralization sensitivity of early amino acid substitutions represented in sequences from the earliest sampled time points through 6 months post-seroconversion ( Figures 2–4 ) . This was done using site-directed mutagenesis to introduce mutations alone or in combination into T/F envs and testing the Envs for neutralization sensitivity . A total of 31 mutants were tested ( Table 1 ) . To further define the epitope specificities of the early Nab responses in subject CH40 , mAbs AbCH83 and AbCH84 were generated from day 132 B cell cultures by screening for neutralization of CH40 T/F virus . Both mAbs utilized VH 3–30 and Vκ 3–15 VH and VL families , respectively , and were clonally related ( Tsao , CY et al . manuscript in preparation ) . Both AbCH83 and AbCH84 bound well to CH40 T/F Env gp140 ( EC50 = 0 . 2 and 0 . 07 ug/ml , respectively ) and both neutralized the CH40 T/F Env pseudovirus potently ( IC50 = 0 . 075 and 0 . 034 ug/ml ) ( Table 1 ) . AbCH83 and AbCH84 were strictly strain-specific , failing to neutralize heterologous viruses including CH77 and CH58 at concentrations as high as 10 ug/ml ( not shown ) . When tested against the panel of CH40 site-directed mutants shown in Table 1 , AbCH83 and AbCH84 demonstrated patterns of neutralization that were strikingly similar to each other and to the day 111 CH40 plasma ( Table 1 ) . For example , AbCH83 and AbCH84 , like the day 111 plasma , potently neutralized the T/F virus , and V1 and V3 mutations that conferred resistance to day 111 plasma ( E146K , E146G , N139T/E146T/M147L , R327K/E332K , and N300H ) also conferred resistance to the two mAbs ( Table 1 ) . Mutations representing minor sequence variants that conferred partial escape from day 111 plasma also conferred similar degrees of partial escape from AbCH83 and AbCH84 ( G145E , N144K , T295N ) . A site-directed K160N mutant of the CH40 T/F virus did not alter neutralization sensitivity to plasma Nabs or to the two mAbs ( Table 1 ) . Interestingly , the C-terminal V3 mutations R327K and E332K together conferred complete escape to each mAb and to day 111 plasma antibodies , but neither mutation alone conferred resistance to the two mAbs and only partial resistance to polyclonal antibodies in plasma . Thus , the two mAbs AbCH83 and AbCH84 nearly recapitulated the polyclonal Nab reactivity in plasma at the day 111 time point , suggesting that the latter was monospecific and directed to a single epitope distinct from that recognized by PG9/PG16/2909 [44] . To distinguish Nab reactivity targeting linear versus conformational ( discontinuous ) epitopes , we performed competition and adsorption assays with linear Env peptides , full-length gp120 proteins , and full-length tethered gp140 proteins corresponding to the sequences of the autologous T/F viruses , and as controls , heterologous peptides and proteins and randomly shuffled peptide sequences . First , we used overlapping linear peptides spanning complete variable loop and constant region sequences where Nab escape mutations first arose ( V1 in CH40; V1 and V2 in CH77; and C2 and V4 in CH58 ) . At concentrations of 25 ug/ml , none of the linear peptides reduced the neutralizing activity in patient plasma ( data not shown ) . As a positive control , we showed that 25 ug/ml of HIV-1 V3 peptides could inhibit V3-targeted Nabs in the same TZM-bl assay [45] . To determine if Nabs in patient plasma recognized epitopes presented on gp120 , we performed adsorption assays using autologous gp120 Env monomers and tethered gp140 Env trimers . These Env proteins were attached to magnetic beads , incubated with patient plasma , and then removed before performing neutralization assay because Env alone can neutralize HIV-1 infectivity by binding cell surface CD4 . The mAb b12 , previously shown to neutralize all three subjects' T/F Envs ( IC50 = 0 . 7–1 . 5 ug/ml ) , served as a positive control to assess the conformational and antigenic integrity of the synthesized gp120 and gp140 glycoproteins . As shown in Figure 5 , b12 at a concentration of 10 ug/ml reduced viral infectivity to below 25% of control in each subject . This inhibitory effect was completely eliminated by preadsorption of b12 with either the Env monomer or Env trimer from each subject , indicating that the gp120 and gp140 proteins had intact b12 binding sites . For CH40 , the gp120 Env monomer was ineffective at adsorbing plasma Nab . The tethered gp140 Env trimer , however , adsorbed neutralizing activity allowing infectivity to rise to 57% against day 111 plasma and 45% against day 181 plasma ( Figure 5A ) . These data suggested that CH40's earliest plasma Nab recognized a conformational epitope that is best displayed on trimeric Env , either on contiguous components of neighboring monomers or within a single Env monomer that is dependent on trimeric Env for appropriate presentation . For CH77 and CH58 , weaker autologous Nab titers made the effects of adsorption more difficult to discern , but in both cases , the Env monomers and Env trimers were equally effective at adsorbing neutralizing activity ( Figure 5B , C ) . For CH77 , the baseline infectivity was high for the low titer day 102 plasma ( >60% ) , and increased with non-specific binding to the BSA-coated bead , but both the Env monomer and trimer further increased this to >100% . CH77's day 159 plasma demonstrated more clearly that the Env monomer and trimers could both effectively bind and adsorb plasma Nab , with greater differences between negative controls and the Env-coated beads . The adsorption experiment in CH58 demonstrated similar findings; both the gp120 monomer and the gp140 trimer adsorbed neutralizing activity equally ( Figure 5C ) . The ability of gp120/gp140 proteins , but not linear peptides , to adsorb neutralizing activity suggests that the earliest plasma Nabs from CH77 and CH58 targeted conformational epitopes that did not require quaternary structure for effective presentation . Escape mutations were analyzed in the context of a relatively complete model of gp120 that was assembled from crystal structures of core gp120 with V3 and core gp120 with N and C termini ( Figure 6A–C , left panels ) [46]–[48] . This model lacked only the V1/V2 region , and escape mutations in V1/V2 were thus modeled with the scaffolded structure of V1/V2 [49] ( Figure 6 , right panels ) . Despite the availability of atomic-level structures of each of these component portions of gp120 , the overall conformation of gp120 in the context of the functional viral spike is still unknown , and so the positions of escape mutations in the spike were inferred from lower resolution electron microscopy results ( Figure 6D ) [50]–[52] . For CH40 , Nab escape mutations were observed at the amino- and carboxy-termini of the V3 region , and included an additional PNLG site at residue 295; escape mutations in the V1 loop were also observed . Phenotypically-proven escape mutations in V1 predominated at the earliest time point ( day 111 ) ( Figure 2; Table 1 ) and evolved through 412 days of follow-up , indicating that these mutations made a major contribution to neutralization escape . A synergistic effect between V1/V2 and V3 neutralization escape mutations has been noted in other contexts [32]–[34] , [41] , and these two epitopes are spatially close on the low resolution viral spike ( Figure 6D ) . The ability of trimeric gp140 , but not monomeric gp120 , to absorb neutralizing activity from the CH40 sera ( Figure 5A ) , is consistent with this interpretation and implicates a conformational epitope involving the V1 and V3 regions that requires quaternary protomer interactions for its integrity . For CH77 , the escape mutations appeared predominantly in V2 , with a number of different V2 mutations selected , several of which alter a PNLG site . Escape mutations were also selected in the V1 and C2 regions and these involved glycan modifications as well . On the trimer structure ( Figure 6D ) , the C2 and V1/V2 regions are not spatially close , suggesting that that the mutations may have had conformational influence on distant sites or that they were selected to escape different antibody responses . For CH58 , escape mutations appeared primarily on the gp120 outer domain in the C2 , C3 and V4 regions . These mutations map to a glycosylated outer vertex of the viral spike , which is relatively restricted in space , and therefore likely represents a single epitope not expected to be quaternary in nature . This interpretation is supported by the finding that monomeric gp120 absorbs neutralizing activity from this serum ( Figure 5C ) . SGA-based sequencing provides a proportional representation of plasma viral populations but with limited sensitivity due to practical constraints of gene-wide sequencing [10] . With a sample size of 30 sequences , there is a 95% probability of detecting a variant that comprises ≥10% of the population [10] . To detect variants comprising substantially less than 10% of the circulating plasma virus , we used parallel allele-specific sequencing ( PASS ) and 454 pyrosequencing . PASS involves PCR amplification within a polyacrylamide gel using modified primers and fluorophore-labeled nucleotides to distinguish single nucleotide polymorphisms in each amplicon [53] . Using PASS , we characterized hundreds to thousands of sequences per time point over a six nucleotide span in CH40 and CH77 corresponding to known Nab escape mutations . Due to the low plasma viral load of CH58 , PASS analysis was not feasible in this subject . For CH40 , SGA sampling ( 14 sequences ) of day 45 plasma revealed only T/F sequences in V1 , whereas PASS detected 1 . 1% ( 5/492 ) of sequences with the E146K Nab escape mutation and 0 . 4% ( 2/492 ) of sequences with an M147L mutation ( Figure 2B; Table S1 ) . PASS yielded similar increases in sensitivity in detecting Nab escape mutations in subject CH77 , where SGA sequencing revealed 100% T/F virus at day 32 in V2 and PASS identified a small variant population ( 0 . 2% ) of the predominant ( T187aN ) Nab escape variant seen in the subsequent time points ( Figure 3B; Table S2 ) . 454 pyrosequencing extended these analyses in subject CH40 where the immunodominant V1 epitope was analyzed over time . Sequences from days 16 , 45 and 181 post-seroconversion were amplified and bi-directional reads from two amplicons spanning the V1 loop of the T/F sequence were codon aligned and analyzed . The number of high-quality , interpretable reads spanning V1 ranged from 10 , 275 to 22 , 344 ( Table 2 ) . These V1 sequences yielded 81 , 110 and 249 unique nucleic acid sequences and 63 , 77 and 246 unique amino acid translations , for days 16 , 45 and 181 , respectively ( Table 2 ) . Compared to SGA and PASS sequences , 454 sequencing substantially increased the sensitivity for detection of rare variants ( Figure 7 ) . Over a six amino acid span covering the V1 Nab epitope region NGEMME ( HXB2 positions 144–149 ) , SGA detected only the T/F among fourteen day 45 plasma viral genomes ( Table S1 ) and PASS detected two variant sequences in 7 of 459 viral genomes ( Table S1 ) , whereas 454 pyrosequencing detected 18 variants among 481 sequences that differed at day 45 from the T/F genome . To test whether the detection of rare Nab escape variants by 454 sequencing constituted statistically significant evidence of selection at this epitope , we compared this region of env with an adjacent control region . The Nab epitope region ( NSNGEMMEKGEV ) corresponded to Env codons 142–154 ( Figure 7B ) and contained the phenotypically-confirmed early Nab escape mutations ( Figure 3 and Table 1 ) . We compared this region with an adjacent control region composed of the remaining 12 amino acids falling between the two conserved cysteine residues that bound the V1 loop . These two regions had similar sizes , were both located in a variable region , and were covered within the same 454 reads , thus precluding issues of differential sequence coverage and error rates . We compared the two regions using three different statistical methods . Entropy comparisons: Using Shannon entropy , a simple measure of variation in DNA and protein sequence alignments that reflects both the number of variants and their distribution , we computed the amino acid entropies per site and then compared the entropies inside and outside of the Nab epitope region . Entropies increased with time ( Figure 7B ) and were significantly greater inside than outside of the Nab epitope region for each time point , including just day 16 post-seroconversion ( p = 0 . 0302 ) and day 45 ( p = 1 . 97×10−5 by one-sided Wilcoxon rank-sum test ) . This significant difference in entropies resulted from a rank-based statistic , which is insensitive to the extremes of values , as seen as large peaks at N143 and E145 ( Figure 7B ) . This indicates greater variability among nearly all sites inside compared to outside the Nab epitope region . Positive Selection: To detect evidence of positive selection pressure , we assessed the ratio of non-synonymous to synonymous substitution rates ( dN and dS , respectively ) . When dN<dS , negative selection is evident; conversely , dN>dS indicates positive selection . The large sample sizes that result from 454 pyrosequencing are computationally intractable for most established procedures that test for positive selection . We used SNAP ( Synonymous Non-synonymous Alignment Program ) , which corrects for alternative mutational pathways in a codon and performs efficiently for large sequence sets . Again , we compared the 12 amino acid Nab epitope region with the remaining 12 amino acids in V1 and computed the distance-corrected synonymous and non-synonymous substitution rates with SNAP . For each amplicon sampled , we summarized SNAP results as contingency tables , wherein columns indicate dN<dS or dN>dS and rows indicate sites inside or outside the epitope . We then populated the table with counts for each sequence relative to the T/F sequence ( excluding cases where dS = dN ) and used one-sided Fischer's exact tests to evaluate whether the epitope was enriched for non-synonymous substitutions ( Table 3 ) . These comparisons indicate significant increases in positive selection within the epitope at day 45 ( p = 0 . 013 ) and day 181 ( p = 0 . 002 ) , with a trend at day 16 . Poisson Model: Following the method of Giorgi et al . [54] , we used a simple model of sequence evolution to test for homogeneous infection . The model gives a null hypothesis of Poisson-distributed intersequence distances . For the single variant transmission of CH40 , rejecting the null model indicates that selection is present in the sequences sampled . Because APOBEC hypermutations violate model assumptions , the test is repeated without APOBEC mutated sites . With APOBEC mutations excluded , the Poisson model test results indicate simple evolution and no selection in reads sampled at day 16 and day 45 ( P>0 . 9 , Table 4 ) . When APOBEC sites are retained in the analysis , the Nab epitope fails to conform to the Poisson ( P<10−9 , Table 4 ) , consistent with selection facilitated by APOBEC . By 181 days post-screening , the Poisson model is rejected regardless of whether or not APOBEC hypermutations are excluded ( P<10−9 , Table 4 ) . All three methods indicated that variation within the putative V1 Nab epitope region was statistically significantly enriched over background mutations; selection for Nab escape mutations was unequivocal by all methods of analysis at day 45 and day 181 and supported at day 16 by an increase in entropy within the epitope region ( p = 0 . 03 , Figure 7 ) .
We used three increasingly sensitive DNA sequencing methods – SGA , PASS and 454 – to look for genetic evidence of Nab selection on the evolving HIV-1 quasispecies . By three to six months post-seroconversion , SGA sequencing identified a set of candidate Nab escape mutations , which in every subject was discontinuous and could be distinguished from CTL escape mutations [7] . Each of the candidate Nab escape mutations that we inferred from SGA sequencing was shown phenotypically to confer significant ( 2 to >70 fold ) resistance to early Nabs ( Table 1 ) . Remarkably , at the time of initial detection of Nab titers , regardless of titer , the virus quasispecies in each subject demonstrated complete or near complete replacement of the T/F sequence by escape mutants at their respective Nab epitopes . This indicated a pre-existent Nab response . PASS analysis corroborated this finding by revealing genetic evidence of Nab escape significantly earlier at just 45 days and 32 days post-antibody seroconversion in subjects CH40 and CH77 , respectively ( Tables S1 and S2 ) . This was at a time point when Nab titers to each T/F virus were undetectable at a 1∶20 plasma dilution in the TZM assay ( Table 1 ) . Nabs at this early time point were also below the level of phenotypic detection when tested in the sensitive A3R5 cell-based virus entry assay [55] ( D . C . M . , unpublished ) . In subject CH40 , where viral loads were highest and deeper sequencing could be done , 454 analysis identified a much larger number of variants in the V1 epitope region of the T/F virus sequence at a still earlier time point 16 days post-seroconversion as well as at 45 and 181 days post-seroconversion ( Table 2 ) . The 454 data further suggested a role for APOBEC mutations facilitating this escape , since at days 16 and 45 post-seroconversion the V1 Nab epitope region was enriched for mutations at APOBEC motifs ( Table 4 ) . APOBEC and Vif function have been implicated in virus escape from early CTL immune pressure [56] , and our results suggest that APOBEC may play an analogous role in the dynamics of early Nab escape at certain epitopes . It is possible that the increased genetic diversity in V1 arising from APOBEC mediated polymorphisms facilitated more rapid escape in this region than in other regions of the Nab epitope . Overall , a combination of SGA , PASS and 454 pyrosequencing enabled the genetic detection of Nab escape variants significantly before Nabs rose to titers detectable in the TZM assay . This enhancement in detection amounted to 95 days in subject CH40 , 70 days in subject CH77 , and 109 days in subject CH58 . Future studies in other subjects with more narrowly spaced sampling intervals may better define these windows . In each subject we found an early monospecific Nab response directed toward a single conformational epitope that was unique to each T/F virus strain . This was demonstrated most clearly in subject CH40 where polyclonal plasma antibodies and autologous mAbs targeted essentially the same epitope at the base of V3 ( depicted in Figure 6 ) . Previous studies have reported epitopes in this region of gp120 to be immunogenic and a target of both broadly and narrowly reactive neutralizing mAbs [37] , [57] . Interestingly , we observed that the binding of both AbCH83 and AbCH84 to autologous CH40 Env gp140 as assessed by Biocore analysis could be blocked by the potent and broadly neutralizing PGT 121 mAb , which in other contexts is dependent on N332 [57] ( B . F . H . , unpublished ) . The early Nab response in CH40 , unlike responses in CH77 and CH58 , targeted an epitope dependent on trimeric Env for structural integrity . Thus , structural modeling and empirical analyses suggested that in subject CH40 , virus escaped Nab pressure indirectly by early mutations in V1 and directly through mutations in the putative V3 epitope , indicating a close association between V1 and V3 in the context of the native functional Env trimer . In all three subjects , we identified Nab epitopes involving unique sites on the Env glycoprotein , with continuous virus evolution at the respective epitopes , without evidence of broadening of the Nab response to additional sites on the Env trimer over the first year of infection ( Figures 2–4 ) . In CH77 , escape occurred predominantly in V2 , where the addition of PNLG site conferred Nab escape at a likely protein epitope . In CH58 , modeling suggested that early Nabs targeted a single conformational epitope involving the Env outer domain , with escape arising through the loss of any of several component glycans . Thus , in each subject , virus employed glycan shifts as well as gain or loss of glycans to mediate escape from the sequential rounds of the Nab response . These findings , in conjunction with reports of monospecific early Nab responses in subtype C infection [58] , [59] , suggest that individual immunodominant regions of Env , specific to the unique conformation of each T/F Env , are targeted by early Nab responses . The observation that very low level Nab titers can impede virus entry and select for virus escape in vivo is consistent with recent findings of selection for SIVmac251 and SIVsmE660 Nab escape mutations in early-chronic infection of rhesus macaques by low titers of Nabs [60] , the association of low-titer Nabs with protection against SIVsmE660 challenge in the nonhuman primate ( NHP ) model [26] , and results from low dose mucosal NHP challenge models in which concentrations of Nabs corresponding to modest in vitro titers were able to effectively prevent SIV acquisition [24] , [25] . To our knowledge , however , this is the first demonstration in human HIV-1 infection that very low titers of Nabs in the range of 1∶50 to 1∶20 or even lower can impede virus replication and select for virus escape . To explore quantitatively the in vivo activity of early Nab responses , we employed a mathematical model to estimate the proportion of de novo infection events blocked by Nabs , or the Nab efficacy ( see Figures S3 , S4 and Dataset S1 ) . The results , which represent minimum estimates , ranged from a low of 19 . 6% to a high of 35 . 2% and represent a Nab response that is sufficiently potent to drive replacement of the T/F virus within several weeks ( Figure S4 and unpublished data ) . Our conservative modeling likely underestimated true Nab efficacy because we utilized minimum estimates for biological parameters with uncertain quantities and did not account for potential fluctuations in Nab efficacy . Future studies where sampling time points are better structured for evaluating dynamic changes in Nab titers and viral quasispecies composition would allow for greater precision in estimations of Nab efficacy in vivo and a better understanding of the kinetics of Nab development . These caveats notwithstanding , the data raise the possibility that in the setting of sexual transmission , where the risk of infection per coital act is low and the number of transmitted viruses responsible for productive clinical infection is typically one , a vaccine that elicited Nabs of sufficient breadth but at titers as low as 1∶50 to 1∶20 or possibly even lower could have a demonstrable protective effect . It is possible that such a low titer neutralizing activity in vaccinees from the Thai RV144 trial could have contributed to the observed 31% protective effect of the vaccine [27] , [61] , [62] . The rates of Nab-driven T/F sequence replacement are more rapid than previously reported [6] , [30] , [32] , [33] , [35] , [36] , [40] , [42] , [59] but substantially slower than rates of loss due to the initial CTL responses [4] , [63] A unique aspect of the present study is that we could directly compare the rate of Nab escape with the rate of CTL escape in the same three subjects [7] . Based on SGA analyses , we previously observed virtually complete replacement of the T/F virus population at defined CTL epitopes within 45 days ( CH40 ) , 14 days ( CH77 ) and 45 days ( CH58 ) of antibody seroconversion [7] , [8] . This contrasts with 111 , 102 and 154 day intervals shown in the present report for Nab escape . Similarly , in a 454 pyrosequencing analysis of CTL escape kinetics , we previously observed in subject CH40 a 1% replacement of T/F sequences just prior to antibody seroconversion ( corresponding to day 0 in the present study ) , a 52% replacement by day 16 , and a 99 . 4% replacement by day 45 [38] . Comparable numbers for Nab escape variant frequencies in subject CH40 in the present study were <1% , 2% and 3% , respectively , again highlighting the much faster rate of CTL escape compared with Nab escape . Furthermore , in the former study , we found that the average rate of HIV escape from CTL responses in acute infection to be 0 . 17 day−1 with a maximum of 0 . 42 day−1 [7] , [38] . The average rate drops to 0 . 03 day−1 by 100 days post seroconversion [64] . This slower rate of virus escape from chronic CTL responses is similar to that of contemporaneous Nab responses measured in the current study , suggesting that Nabs could contribute along with CTLs to virus containment during this later time period . In acute infection of unvaccinated subjects , however , Nab responses likely contribute negligibly to early virus containment . The costs to replication fitness associated with virus escape from autologous Nab responses have not been well characterized but could contribute to partial virus containment at setpoint viremia . Fitness costs of Nab escape mutations have frequently been considered to be minimal [35] , [65] , but Derdeyn and colleagues described a Nab escape mutation in V2 , which when placed in the autologous T/F virus backbone , conferred a measurable fitness cost [58] . Morris and colleagues [31] similarly noted transient decrements in plasma virus load coincident with the development of strain-specific Nabs . We studied Nab escape mutations within the context of a 6 month consensus IMC so the effects of compensatory mutations could be accounted for and so mutations resulting from escape from Nabs could be distinguished from those resulting from escape from CTLs . Our analyses suggested that Nab escape mutations conferred reductions to replication fitness ranging from 0 to 24% . This corresponds to an estimated average impairment to virus entry due to early strain-specific Nabs of as much as 31 . 3% to 48 . 8% . Finally , we note that the exquisite sensitivity and rapid adaption of HIV-1 Env to Nabs contrasts with recent observations for the HIV-2 Env , where high-titer Nabs register little effect on env evolution or Env Nab escape [66]–[68] . A biological explanation for these differences is not obvious . For HIV-1 , the enhanced sensitivity and rapid adaptation to Nab pressure in vivo provides an explanation for the HIV-1 Env's propensity to maintain a fully assembled glycan/conformational shield [6] . Paradoxically , it is this enhanced sensitivity of HIV-1 to Nabs in vivo that appears to be responsible for its vaunted ability to resist neutralization by all but the most broadly reactive and potent Nabs . Another provocative implication of the current study is that , in vivo , Nabs impede HIV-1 spread whether this is occurring by ‘cell-free’ or ‘cell-to-cell’ mechanisms . This is at odds with the suggestion that ‘cell-to-cell’ spread of HIV-1 provides a mechanism for replicating virus to escape Nab or antiretroviral drug pressure [69] , [70] . Further investigation is needed to resolve this question .
This study was conducted according to the principles expressed in the Declaration of Helsinki . It was approved by the Institutional Review Boards of the University of Pennsylvania , the University of Alabama at Birmingham , the University of North Carolina and Duke University . All subjects provided written informed consent for the collection of samples and subsequent analysis . The experimental strategy was first to define the kinetics of appearance of autologous and heterologous Nabs in each subject by a conventional single-cycle virus entry assay [6] , [71] . Next , we performed an in-depth SGA-based analysis of plasma viral env gp160 RNA sequences at serial time points beginning prior to antibody seroconversion and extending beyond the first year of infection . The SGA approach allowed us to look for amino acid selection across intact env gp160 genes not accounted for by CTL-driven virus escape [7] that might reflect Nab-mediated virus escape . Next , putative Nab epitopes were corroborated by cloning and analyzing full-length infectious molecular clones ( IMCs ) corresponding to T/F and consensus 6 month sequences and by performing site-directed mutagenesis on T/F env genes so as to introduce individual putative escape mutations arising in the first year of infection for phenotypic testing against sequential patient plasma specimens and monoclonal antibodies . Plasma samples containing neutralizing activity were adsorbed with autologous or heterologous Env peptides or polyproteins to distinguish linear from conformational Nab epitopes . This was followed by deeper sequence analyses using parallel allele-specific sequencing ( PASS ) and 454 pyrosequencing to identify the earliest genetic signatures of Nab escape at confirmed epitopes . T/F and consensus 6 month IMCs , with and without Nab escape mutations , were evaluated for in vitro replication kinetics to access fitness costs of Nab escape , and structural and mathematical models were used to interpret data within the context of viral Env structure and replication kinetics in vivo . Peripheral blood samples were obtained from subjects 700010040 ( CH40 ) , 700010058 ( CH58 ) , and 700010077 ( CH77 ) after obtaining informed consent under the Duke University and University of North Carolina human use review boards . All subjects were North American men who had sex with men ( MSM ) who denied injection drug use all were infected with HIV-1 subtype B , and all were antiretroviral drug naïve throughout the study course . At initial sampling , the three subjects were at peak viremia ( Fiebig stage II , plasma vRNA+ , and Ab− ) just prior to HIV antibody seroconversion [10] , [72] , [73] . Viral RNA from each time point was extracted and reverse transcribed to cDNA as previously described [8] . Approximately 20 , 000 viral RNA copies were extracted using the BioRobot EZ1 Workstation with EZ1 Virus Mini Kit ( version 2 . 0; QIAGEN ) , and 5 , 000 vRNA molecules were reverse transcribed using SuperScript III ( Invitrogen ) and the primer R2 . B3R 5′-ACTACTTGAAGCACTCAAGGCAAGCTTTATTG-3′ . SGA was performed as described previously [8] , [10] . Briefly , cDNA was serially diluted so as to identify a dilution where PCR positive wells constituted less than 30% of the total number of reactions . At this dilution , most wells contain amplicons derived from a single cDNA molecule . PCR reactions used Platinum Taq High Fidelity polymerase ( Invitrogen ) and nested primers OFM19 and Vif1 ( first-round ) and EnvA and EnvN ( second round ) to generate full-length gp160 env sequences . To obtain subgenomic sequences containing putative Nab epitopes in CH58 , nested primers CH58C2 . OutF , CH58C2 . OutR , CH58C2 . InF , and CH58C2 . InR were used to amplify a 554 nucleotide region spanning V1 through C3 , and nested primers CH58 . C3V4 . 5outA , CH58 . C3V4 . 3outA , CH58 . C3V4 . 5InA , and CH58 . C3V4 . 3InA were used to amplify a 377 nucleotide region spanning regions V3 through C4 . PCR parameters were as follows: 94°C for 2 min , followed by 35 cycles of 94°C for 15 s , 58°C for 30 s , and 68°C for 4 min , followed by a final extension of 68°C for 10 min . The product of the first-round PCR was used as a template in the second-round PCR reaction under the same conditions , but with a total of 45 cycles . The amplicons were inspected on precast 1% agarose E-gel 96 ( Invitrogen Life Technologies ) . All PCR procedures were carried out under PCR clean room conditions . SGA primer sequences: OFM19: 5′-GCACTCAAGGCAAGCTTTATTGAGGCTTA-3′ Vif1: 5′-GGGTTTATTACAGGGACAGCAGAG-3′ EnvA: 5′ GGCTTAGGCATCTCCTATGGCAGGAAGAA-3′ EnvN: 5′-CTGCCAATCAGGGAAGTAGCCTTGTGT-3′ . CH58C2 . OutF: 5′-CCATGTGTACAATTAACCCCACTCTGTGTC-3′ CH58C2 . OutR: 5′-CTGTTCTCTTAATTTTGTAACTATCTTC-3′ CH58C2 . InF: 5′-GTAGCGAGGGAAAGGAAATGAAGAACTG-3′ CH58C2 . InR: 5′-GTGTTATTCCATTGTTCTCTACTAAGGTTAC-3′ CH58 . C3V4 . 5outA: 5′-CTGCTGTTAAATGGCAGTCTAGCAGAAAAAGATATAG-3′ CH58 . C3V4 . 3outA: 5′-CTCATATCTCCCCCTGCAGGTCTGAAGGTC-3′ CH58 . C3V4 . 5InA: 5′-GTACAAGACCCAACAACAATACAAGAAAAAGTATAAC-3′ CH58 . C3V4 . 3InA: 5′-CCTTTGATGGGAGGGGCATACATTGCTTTTC-3′ Amplicons were directly sequenced by cycle-sequencing using BigDye Terminator chemistry ( Applied Biosystems ) . Sequencing reaction products were analyzed with an ABI 3730xl genetic analyzer ( Applied Biosystems ) . Both DNA strands were sequenced using overlapping fragments . Individual sequence fragments for each amplicon were assembled and edited using the Sequencher program 4 . 8 ( Gene Codes; Ann Arbor , MI ) . Chromatograms containing mixed bases ( double peaks ) were excluded . All sequences were manually inspected and aligned in MacClade 4 . 08 to optimize alignments . Consensus sequences were generated for each individual from the earliest sample ( pre-antibody seroconversion , Fiebig Stage II ) and longitudinal sequences aligned accordingly . All sequences were deposited in GenBank ( accession numbers: JQ957568–JQ957796 ) . Full-length gp160 env genes were amplified by nested PCR from acute infection plasma HIV-1 RNA , cloned , and sequenced to confirm their identity with T/F genomes [6] , [10] . Site-directed mutations corresponding to naturally-occurring mutations were introduced with QuickChange site-directed mutagenesis kit ( Strategene ) . IMCs of T/F genomes were previously described [8] , [39] . SGA-derived sequences from 6 months ( 159–181 days ) post-seroconversion were used to determine a consensus sequence . At polymorphic positions , the majority nucleotide was selected . At positions where there was no single nucleotide representing >50% of sequences , the most prevalent nucleotide was selected . Six month IMCs , with and without putative Nab escape mutations , were constructed by chemical synthesis ( Blue Heron ) and site-directed mutagenesis by methods previously described [8] , [39] . All IMCs were sequence confirmed . Plasma samples were assayed for Nab activity against IMC-derived virions or Env- pseudotyped virions using a single-round JC53BL-13/TZM-bl pseudotype reporter assay [6] ) , JC53BL-13 cells were plated and cultured overnight . A total of 2 , 000 infectious units of each pseudotyped virus were combined with fivefold dilutions of heat-inactivated test plasma or serum and incubated for 1 h at 37°C . Non-HIV-infected heat-inactivated human plasma was added as necessary to maintain a constant overall concentration . The virus-Ab mixture was then added to JC53BL-13 cells , and after 2 days , the cells were lysed , and the luciferase activity of each well was measured using a luciferase assay reagent ( Promega , Madison , WI ) and an ABI Tropix ( Applied Biosystems , Foster , CA ) . Background luminescence was determined in uninfected wells and subtracted from all experimental wells . Cell viability and toxicity were monitored by basal levels of luciferase expression and by visual inspection . Relative infectivity ( percentage of control ) was calculated by dividing the number of luciferase units at each plasma dilution by the values in wells containing no test plasma . The dilution of test plasma or serum that inhibited 50% of virus infectivity ( IC50 titer ) was determined using a linear regression-least squares fit method . mAbs were tested for neutralizing activity beginning at 10 ug/ml and proceeding with five-fold dilutions , as previously described [6] . IgG+ memory B cells were isolated from frozen peripheral blood mononuclear cells ( PBMCs ) from day 111 after enrollment and cultured at near clonal dilution as described [74] . Cells were obtained from CH40 at 132 days post-seroconversion by selecting CD2− , CD14− , CD16− , CD235a− , IgD− and IgG+ cells through two rounds of separation with magnetic beads ( Miltenyi Biotec , Auburn , CA ) . Cells were then resuspended in complete medium containing 2 . 5 µg/ml CpG ODN2006 ( tlrl-2006; InvivoGen , San Diego , CA ) , 5 µM CHK2 kinase inhibitor ( Calbiochem/EMD Chemicals , Gibbstown , NJ ) and EBV ( 200-µl supernatant of B95-8 cells/104 memory B cells ) . After overnight incubation at 37°C in 5% CO2 , 21 , 600 viable cells were seeded in 96-well round-bottom tissue culture plates at a cell density of 3 memory B cells/well in presence of ODN2006 , CHK2 kinase inhibitor and irradiated ( 7 , 500 cGy ) CD40 ligand-expressing L cells ( 5 , 000 cells/well ) . Cells were re-fed at day 7 and harvested at day 14 . The two 96 well supernatants that most effectively neutralized CH40 T/F Env pseudotyped virus in the TZM-bl assay , as previously described [74] , were selected for further analysis . RNA from positive cultures was extracted ( RNeasy minikit; Qiagen ) , and the genes encoding Ig V ( D ) J rearrangements were amplified by RT and nested PCR , and the mAbs expressed as recombinant IgG1 antibodies ( designated AbCH83 and AbCH84 ) as previously described [74] . Recombinant monoclonal antibodies AbCH83 and AbCH84 were assessed for neutralization in TZM-bl neutralization assays against autologous and heterologous pseudotyped viruses and for Env binding with enzyme-linked immunosorbent assays with a panel of autologous and heterologous tier 1 and 2 viruses and viral proteins [74] . For CH40 , a 26-mer peptide corresponding in sequence to the CH40 T/F virus V1 region and a control 26-mer scrambled peptide were synthesized ( New England Peptide , Gardner , MA ) . For CH77 and CH58 , 18-mer peptides overlapping by 10 amino acids were synthesized ( Sigma-Aldrich; Medical Research Council Human Immunology Unit , WIMM , Oxford , UK ) to match the T/F sequences of interest for CH77 and CH58 . For CH77 , peptides spanning V1: LTPLCVTLNCTDSNGDS ( 3284 ) , V2: PIDTKTNTSKYRLISCNT ( 3292 ) , DVVPIDTKTNTSKYRLIS ( 3291 ) , and C2: IPIHYCAPAGFAILKCKD ( 1181 ) , AGFAILKCKDKKFNGTGP ( 1467 ) , KDKKFNGTGPCKKVSTVQ ( 3297 ) were used . For CH58 , peptides matching the sequence of the C2: AGFAILKCNNKTFNGTGQ ( 3549 ) , NNKTFNGTGQCTNVSTVQ ( 3550 ) , and V4: KANGTTGNDTIILPCRIK ( 3570 ) were used . For both CH77 and CH58 , two control peptides matching regions other than Env were used [IVYIEYRKIVRQRKIDRL ( 3512 ) , MQSLYILGIVALVVAAIL ( 3509 ) ] . Autologous gp120 and gp140-tethered Env glycoproteins were generated as described [49] , [75] . HIV-1 gp140 trimeric envelope glycoproteins contained a mutated furin cleavage site and a C-terminal fibritin trimerization domain with 8xHisTag . Mammalian codon-optimized genes encoding the wild type and mutant gp120s and gp140s were synthesized and cloned into the mammalian expression vector pHLSec2 ( GENEART AG , Regensburg , Germany ) . For preparation of each envelope glycoprotein , 500 ug of the plasmid DNA was mixed with 1 ml of 293fectin ( Invitrogen , Carlsbad , CA ) for 20 minutes before the DNA-293fectin complex was added into 850 ml of FreeStyle 293F cells ( 1 . 4×106 cells/ml ) in a 2 L shaking flask . After transfection , the cells were returned to suspension incubation for 24 hours at 37°C , 8% CO2 and 125 rpm . The culture was fed with 50 ml of the enriched medium CellBoost-5 ( HyClone , Logan , UT ) and sodium butyrate at final concentration of 2 mM ( SIGMA , St . Louis , MO ) . After 5 days of suspension culture post transfection , supernatant was harvested by centrifugation and filtered through 0 . 22 µm filter . For the gp120 protein preparation , the supernantant was purified through an affinity column of 17b ( made by cross-linking 17b antibody with Protein A plus agarose ( Pierce , Thermo , Rockford , IL ) . For the gp140 protein preparation , the supernantnt was concentrated and buffer exchanged through a Tangential Flow Filtration system ( Pall , Ann Arbor , MI ) against Ni-binding buffer , and purified through a Ni-NTA resin column ( QIAGEN , Valencia , CA ) . The purified proteins were concentrated and dialyzed against PBS and characterized by SDS-PAGE and immune blotting with anti HIV-1 IgG ( HIVIG ) . The gp120 Env monomers and tethered gp140 trimers were coupled to a solid phase tosylactivated magnetic Dynabeads MyOne beads ( Invitrogen ) as previously described [76] . One mg of protein was coupled to 50 mg ( 0 . 5 ml volume ) of tosylactivated magnetic beads . Coupling was performed at 37°C in a total volume of 1 . 25 ml in coupling buffer ( 0 . 1 M sodium borate buffer ( pH 9 . 5 ) w 1 M ammonium sulfate ) with gentle rocking over 8 to 12 hours . The Dyna beads and bound protein were separated from the coupling buffer with a magnet and resuspended with 5 ml of blocking buffer ( PBS ( pH 7 . 4 ) with 0 . 1% ( wt/vol ) BSA and 0 . 05% Tween 20 ) . The beads were then resuspended in 0 . 5 ml of storage buffer ( pBS ( pH 7 . 4 ) supplemented with 0 . 1% ( wt/vol ) BSA , 0 . 05% Tween 20 , and 0 . 02% sodium azide and stored at 4°C . Stocks of beads coupled with BSA , to assess for non-specific binding , were prepared in the same manner , incubating in the blocking buffer ( PBS ( ph7 . 4 ) with 0 . 1% ( wt/vol ) BSA and 0 . 05% Tween 20 ) for the initial step . Competition assays were performed with linear peptides having sequences described above . Neutralization assays were performed with additional peptide and plasma incubation for 30 minutes at 37°C prior the addition of the 2 , 000 infectious units of each virus for 1 hour at 37°C . The neutralization assay was then completed as described above with a final concentration/well of each peptide of 25 ug/ml . For the polyprotein adsorption studies , the plasmas were incubated with 12 uls of the protein-bead complex ( or BSA-bound beads ) for 30 minutes . The beads were removed with a magnet and discarded . This process was repeated an additional 2 times , using a total of 36 uls ( 0 . 036 mg ) of bead slurry . In previous reports , three rounds of bead adsorption resulted in nearly complete removal of Env-specific antibodies from serum/plasma samples [76] . After the final incubation , the plasmas were centrifuged at 7000 rpm for 7 minutes . Neutralization assays were then performed as described above . The PASS assay was performed as previously described [53] . Briefly , 20 µl of a 6% acrylamide gel mix ( 1 µM acrydite-modified primer ( CH40-rev or CH77-rev ) , cDNA template , 0 . 3% diallyltartramide , 5% rhinohide , 0 . 1% APS , 0 . 1% TEMED and 0 . 2% BSA ) was cast on a bind-saline ( Amersham Biosciences , Piscataway , NJ ) treated glass slide . In-gel PCR amplification was then performed with a 300 µl PCR solution ( 1 µM primer ( CH40/77-for ) , 0 . 1% Tween-20 , 0 . 2% BSA , 1× PCR buffer , 100 µM dNTP mix and 3 . 3 units of Jumpstart Taq DNA polymerase ( Sigma , St . Louis , MO ) under a sealed SecureSeal chamber ( Grace Bio-Labs , Inc . , Bend , OR ) in a PTC-200 Thermal Cycler . PCR conditions were as follows: incubation at 94°C for 3 min , 65 cycles of a denaturing step at 94°C for 30 sec . , an annealing step at 56°C for 45 sec . , and an extension step at 72°C for 1 min; and one cycle of an additional extension step at 72°C for 3 min . After PCR amplification , single-base extension ( SBE ) was performed using the fluorescently labeled nucleotides dGTP-Cy3 ( PerkinElmer , Waltham , MA ) , dTTP-Alexa-568 ( Invitrogen , Carlsbad , CA ) , dATP-Cy5 ( PerkinElmer , Waltham , MA ) , and dUTP-Cy5 . 5 ( GE Healthcare , Piscataway , NJ ) . Sequencing primers ( CH40-seq or CH77-seq ) annealed just upstream of the mutation site as well as for the next five consecutive bases . The gel was scanned with an Axon GenePix 4300A Microarray Scanner ( Molecular Devices , Sunnyvale , CA ) and analyzed with Progenesis PG200 ( Nonlinear Dynamics , Durham , NC ) software . Sequenced nucleotides were determined by comparing each polony's normalized intensity in all four channels . PASS PCR primer sequences: CH40-rev: 5′Acr-TTTCCCTGGTCCCATGGGTATACTTTTTC-3′ or CH77-rev: 5′Acr-ATTATTGCCGGGTCTCATACATTTG-3′ CH40/77-for: 5′-CCACAGACCCCAACCCACAAGAAG-3′ PASS sequencing primer sequences , CH40: 5′-TACTAATACCACTAATAGTAACGGG-3′ ( nt 6635–6659 ) , 5′-TACTAATACCACTAATAGTAACGGGG-3′ , 5′-TACTAATACCACTAATAGTAACGGGA-3′ , 5′-TACTAATACCACTAATAGTAACGGGGAA-3′ , 5′-TACTAATACCACTAATAGTAACGGGACA-3′ , 5′-TACTAATACCACTAATAGTAACGGGAAA-3′ , 5′-TACTAATACCACTAATAGTAACGGGGGA-3′ CH77: 5′-TTATAAACTTGATGTAGTACCAATAGATACA-3′ ( nt 6752–6782 ) 5′-TTATAAACTTGATGTAGTACCAATAGATACAA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAG-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAC-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAAA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAGA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACACA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAAC-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAAC-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAAAA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAGAA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACACAA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAACA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAAAAA-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAAAAG-3′ 5′-TTATAAACTTGATGTAGTACCAATAGATACAGAAA-3′ RNA was extracted from pelleted virions containing at least 200 , 000 viral RNA copies using EZ-1 viral RNA kit ( Qiagen ) from CH40 plasma from days 16 , 45 , 181 , and 412 post-seroconversion . cDNA was synthesized using superscript III reverse transcriptase ( Invitrogen ) in 5 replicates with the antisense primer 1 . R3 . B3R . The cDNA was immediately subjected to nested PCR amplification using Platinum Taq DNA Polymerase High Fidelity ( Invitrogen ) . For each time point , 96 replicate PCR reactions ( 40 µl each ) were performed with 5 µl cDNA , using forward primer BKB3F2 and the same reverse primer used for cDNA synthesis ( 1 . R3 . B3R ) . All 96 first round reactions for each time point were pooled and used as template for three inner PCR reactions . Each inner PCR reaction was performed with 32 replicates ( 40 µl each ) using 5 µl of pooled first round template , and specific primers that incorporated a 4 base identification tag as well as a 19 base 454 adaptor sequence . Agarose gel-run PCR amplicons were visualized with crystal violet/white light and then subjected to 454 sequencing . Each sample was run on a separate picotiter plate with GS-FLX titanium reagents . All amplicons were agar-gel purified , eluted in EB buffer ( QIAquick gel extraction kit ) , and visualized with gentian violet . Directional reads were codon-aligned from two amplicons that span the V1 loop to the T/F sequence and reviewed for sequencing errors as described previously [38] . The forward and reverse reads had similar variant frequencies and entropies per site ( P>0 . 8 by Wilcoxon test ) and thus were pooled to increase the sensitivity of minor variant detection . Where single-site deletions resulted from sequencing errors , we edited the reads to match the T/F sequence . We excluded from analysis reads with multiple consecutive deletions or insertions from base calls out of phase with the flow order , and withheld from selection tests translations that contained premature stop codons . We used the R package ‘binom’ ( version 1 . 0−5 ) to compute exact 95% confidence intervals from the binomial distribution , which quantifies uncertainty of variant frequencies due to resampling . Bulk PCR primers: 1 . R3 . B3R ( 5′-ACTACTTGAAGCACTCAAGGCAAGCTTTATTG-3′; nt 9642-9611 HXB2 ) . BKB3F2 ( 5′ CGGGTTTATTACAGGGACAGCAG 3′; nt 4899–4921 HXB2 ) 454 patient-specific primer pairs for two amplicons ( A & B ) : A–F: GTGGGTCACAGTCTATTATGGG HXB2 nt 6326–6347 A–R: GGCTCAAAGGATACCTTTGGAC HXB2 nt 6859-6838 B–F: GGGATCAAAGCTTAAAACCATG HXB2 nt 6015–6036 B–R: GCATTGTCACTGAAATTGACTG HXB2 nt 6522-6501 Specific adapter sequences ligated to 5′ end of each primer for directional sequencing: F: CGTATCGCCTCCCTCGCGCCATCAG R: CTATGCGCCTTGCCAGCCCGCTCAG Viral replication was assessed in activated primary CD4+ cells from normal human donors as previously described [8] with modifications . Relative replication rates were evaluated in parallel cultures infected by single virus strains and in competition cultures where cells were inoculated with identical numbers of two or three genetically-distinct virus strains . Relative growth rates were distinguished by PASS analysis . Fresh or frozen cells were treated with either 50 ng/ml or 3 µg/ml of staphylococcal enterotoxin B ( Toxin Technology , Sarasota , FL ) for 72 hours at 37°C to activate lymphocytes . 5×105 cells were incubated with 50 , 000 IU of virus ( multiplicity of infection 0 . 1 ) overnight at 37°C in 250 µl RPMI 1640 with 15% FBS and 30 U/ml IL-2 . Cells were washed three times and plated in 24-well polystyrene tissue culture plates in a volume of 500 µl RPMI 1640 with 15% FBS and 30 U IL-2/ml . 50 µl of media was removed for day 1 p24 baseline analysis . Every 2 days , 50 µl lf media was removed and frozen for p24 analysis . For viral replication competition assays , cells were isolated and activated as described above . 1×106 cells were incubated with 50 , 000 IU of each virus ( for a combined multiplicity of infection of 0 . 1 ) overnight at 37°C in 250 µl RPMI 1640 with 15% FBS and 30 U/ml IL-2 . Cells were washed three times and plated in 24-well polystyrene tissue culture plates in a volume of 1 ml RPMI 1640 with 15% FBS and 30 U IL-2/ml . 50 µl of media was removed for day 1 p24 baseline analysis , and 80 µl for an estimate of the input stock for PASS ( see above ) . Every 2 days , 50 µl lf media was removed and frozen for p24 analysis and 80 µl was frozen for PASS analysis . To investigate quantitative relationship between virus replication , diversification , and antibody-mediated selection , we extended a previous model of virus dynamics in HIV-1 infection [77] . In the new model , wild-type ( WT ) virus , , is defined as virus that has the T/F sequence in the Nab epitope under consideration . It is produced from infected cells , , at rate per cell and is cleared at rate per virion . The virus infects uninfected target cells , , at rate , where is the efficacy of circulating antibodies at neutralizing the wild-type virus and is the rate constant characterizing infection by WT virus in the absence of Nabs . A Nab escape mutant , , is resistant to neutralization by these Nabs . It infects target cells at rate , it is produced from infected cells , , at rate and it is cleared at rate per virion . Cells producing virus die at rate . Uninfected target cells are produced at rate and die at rate . The dynamics of the virus and cell populations are thus given by the following equations: ( 1 ) where we assume that at the start of selection , both WT and escape variants are present in the population . Our results are quantitatively similar if we include generation of the escape variant from the T/F virus by mutation ( not shown ) . Modeling escape of HIV from CTL responses , we and others have previously shown that the model ( 1 ) can be simplified due to the rapid clearance of viral particles from circulation [64] so as to consider only the dynamics of cells productively infected with wild-type virus , , and escape mutant , . In this model the concentration of virus is directly proportional to the density of infected cells , so that the ratio is also the ratio of mutant to WT virus . The simplified model is ( 2 ) where is the rate of growth of the population of cells infected with WT virus in the absence of a Nab response , and is the fitness cost of the escape mutation . Because the density of infected cells is proportional to the density of free virus particles , the rate of expansion of infected cells and free virus are identical . It is important to note that if target cell levels vary will be a function of time . In the model ( 2 ) the dynamics of the ratio of the density of the escape variant to the WT virus in the population , , is given by ( 3 ) The change in frequency of the WT , or T/F virus in the population , over time since the Nab response began is given by ( 4 ) where is the frequency of the WT virus at the start of Nab response , assumed to occur at time , and is the average rate of escape of the virus from Nab response over the considered time period [64] . It is clear from this model that escape will only occur if the efficacy of the Nab response in vivo is larger than the fitness cost associated with escape , . The efficacy of the Nab response may change over time , for example , because of an increase in the level of Nabs . To describe the kinetics of the Nab response to the WT virus , we use a simple model where the level of Nabs begins increasing after a time delay , , and saturates over time: ( 5 ) where is the rate of increase of Nab levels over time . The efficacy of Nabs at blocking new infections is likely to be proportional to their concentration . We describe the change in Nab efficacy at blocking de novo infections by the Emax model commonly used in pharmacodynamic modeling , i . e . ( 6 ) where is the antibody concentration which is 50% neutralizing . By using eqns . ( 6 ) and ( 7 ) and numerically solving eqn . ( 3 ) , we find the change in the frequency of the founder virus due to escape from the Nab response . In our SGA data the frequency of the WT or T/F virus sequence changes from 100% at an early time point to 0% at the subsequent time points . To estimate the minimal escape rate we replaced the value for the frequency of WT virus at the early time point , 100% , with and at a later time point , 0% , with where is the number of SGA-derived sequences available . This generates a lower bound estimate of the escape rate . A model of HIV-1 gp120 for all regions except V1/V2 was constructed from crystal structures of HIV-1 gp120 core with complete N and C termini [48] and of HIV-1 gp120 core with V3 [46] . A model of V1/V2 , meanwhile , was utilized directly from the scaffold determined context with PG9 [49] . The GlyProt server was used to model basic glycans at accessible potential N-linked glycosylation sites . Residues of gp120 involved in viral escape were mapped onto these atomic-level models . For the oligomeric viral spike context , the approximate locations of these residues were mapped as determined by cryo-electron microscopy [50]–[52] .
|
Characterizing early adaptive immune responses to HIV-1 can inform studies of virus persistence , pathogenesis and natural history and can guide rational vaccine design . Previous studies examined the role of neutralizing antibodies ( Nab ) in acute and chronic HIV-1 infection but not against the precise envelope ( Env ) glycoproteins of transmitted/founder ( T/F ) viruses and not in direct comparison with autologous cellular immune responses in the same subjects . Here , we identified T/F HIV-1 env genes and their progeny in three subjects by single genome sequencing and performed a dynamic assessment of Nab responses based on env evolution and phenotypic changes in the Env glycoprotein over time . Surprisingly , we found genetic evidence of Nab activity as early as 2 weeks post-seroconversion , with Nab titers as low as 1∶20 to 1∶50 ( IC50 ) selecting for virus escape . Nabs targeted different regions of the HIV-1 envelope ( Env ) in a strain-specific , conformationally sensitive manner . Although delayed in comparison with autologous CD8 T-cell responses , Nabs appeared earlier in HIV-1 infection than previously recognized and impeded virus entry at low titers . This raises the possibility that similarly low concentrations of vaccine-induced Nabs could impair virus acquisition in natural HIV-1 transmission , where the risk of infection is low and the number of viruses responsible for transmission and productive clinical infection is typically one .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"viral",
"immune",
"evasion",
"virology",
"biology",
"microbiology"
] |
2012
|
Early Low-Titer Neutralizing Antibodies Impede HIV-1 Replication and Select for Virus Escape
|
Stereotyped sequences of neural activity are thought to underlie reproducible behaviors and cognitive processes ranging from memory recall to arm movement . One of the most prominent theoretical models of neural sequence generation is the synfire chain , in which pulses of synchronized spiking activity propagate robustly along a chain of cells connected by highly redundant feedforward excitation . But recent experimental observations in the avian song production pathway during song generation have shown excitatory activity interacting strongly with the firing patterns of inhibitory neurons , suggesting a process of sequence generation more complex than feedforward excitation . Here we propose a model of sequence generation inspired by these observations in which a pulse travels along a spatially recurrent excitatory chain , passing repeatedly through zones of local feedback inhibition . In this model , synchrony and robust timing are maintained not through redundant excitatory connections , but rather through the interaction between the pulse and the spatiotemporal pattern of inhibition that it creates as it circulates the network . These results suggest that spatially and temporally structured inhibition may play a key role in sequence generation .
From the kingfisher’s dive to the performance of a piano concerto , sequences of stereotyped actions are central to the everyday lives of humans and animals . One of the most well-studied behavioral sequences in nature is birdsong , and the physiology underlying HVC ( the songbird analogue of mammalian premotor cortex and presumed neural sequence generator ) has been a topic of intense interest . Principal neurons in HVC produce sparse , time-locked bursts of activity that are stereotyped from trial to trial [1 , 2] . Temporally-ordered neural activity has also been observed in other species in the context of various sequential behaviors [3–5] , but the extreme precision and sparsity of the songbird premotor projection cells in HVC are unmatched . How is this spike timing precision maintained in the presence of biological noise ? What makes neural sequence generation in this context robust ? A well-studied model that provides a possible answer is the synfire chain [6 , 7] . The synfire chain assumes that excitatory ( principal ) neurons are organized into pools arranged in a redundant feedforward geometry ( Fig 1A ) . Although synfire chains have not been directly observed in HVC , simulations have shown that the redundant synfire geometry combined with a threshold non-linearity can generate stereotyped , precisely timed sequential activity similar to that observed experimentally in HVC . Typically , models based on the synfire chain either exclude inhibition entirely [8] , or simply use global inhibition to prevent runaway excitation or select between competing chains [9 , 10] . But a series of recent findings in zebra finch suggest that inhibitory spiking plays a larger role in defining the timing of excitatory cell sequences HVC than has been previously assumed . Guitchounts et al . [11] find that inhibitory spiking in HVC , like excitatory spiking , is extremely stereotyped from trial to trial . Kosche et al . [12] find that principal cells that project to the vocal motor pathway ( HVCRA ) receive stereotyped excitatory inputs at multiple points during the song , but only burst when excitatory inputs align with pauses in inhibition . And Markowitz et al . [13] recently observed that in a given region of HVC , excitatory cells that project to the basal ganglia ( HVCX ) and inhibitory cells ( HVCI ) fired at distinct phases of a stereotyped 30Hz component of the local field potential . Local field potentials were not globally synchronized across HVC; instead , their phase varied over the spatial extent of HVC , suggesting that the phasic coordination of cell firing was local rather than global . Here , we describe a general spiking model of neural sequence generation inspired by the observations of Markowitz et al . [13] that shows a similar local alternation of excitatory and inhibitory activity . Like the standard synfire chain , the model is based on feedforward excitatory chains that define an ordered sequence of cell firing . However , in contrast to the synfire chain , the separate strands of the chain do not coordinate their simultaneous activity through crossing excitatory connections . Rather , the strands interact and synchronize through common feedback inhibition ( Fig 1B ) , which produces local inhibitory cycles that act as spatiotemporal “scaffolding” for the feedforward excitatory activity of the chain . We demonstrate in simulation that this locally patterned feedback inhibition synchronizes pools of cells like a synfire chain , and we present analysis that quantitatively describes this effect in terms of the decay rate of inhibition and other system parameters . Our simulations also show that our proposed network and its local inhibitory dynamics help to control the drift of spike timing from one trial to the next . Unlike many previous HVC modeling efforts , [14–16] , our model is not intended to describe HVC in biological detail; instead , we use it to illustrate and investigate the possible contribution of local feedback inhibition to sequence generation . We do , however , discuss intriguing correspondences between the model’s behavior and observations in HVC .
In this work , a model of sequence generation is presented in which a spiraling excitatory chain conducts a pulse of excitatory activity repeatedly through multiple “zones” of inhibition ( Fig 2 ) . In each zone , the arrival of the pulse causes a pool of principal cells to fire; these cells then excite both a pool of principal cells in the next zone and a local source of feedback inhibition . The inhibition then decays until the pulse returns to that inhibitory zone . We show that the presence of this decaying inhibition acts to synchronize the firing of other local excitatory cells when the pulse returns and helps to establish spike timing consistency from one trial to the next . In order to demonstrate that synchronization in our model is independent of the synchronizing effect of synfire connectivity , we intentionally structure our excitatory chain without any fan-in or fan-out connectivity between consecutive pools of excitatory cells . Instead , each cell in a pool sends an excitatory projection to exactly one cell in the next pool . This connectivity pattern creates multiple parallel excitatory strands that do not interact through excitation—the only interaction between these strands is moderated by the local feedback inhibition activated when the pulse passes through each inhibitory zone ( Figs 1B and 2 ) . The conceptual model presented here can also be implemented with excitatory pools connected in a synfire chain , but such an implementation would not serve our purpose of demonstrating that synchronization in our model is independent of the synfire mechanism . The model described here was implemented in MATLAB . All code is available in a figshare digital repository at http://dx . doi . org/10 . 6084/m9 . figshare . 1570957 . All excitatory and inhibitory cells are modeled by quadratic integrate-and-fire ( QIF ) neurons with white noise [17] . The voltage V of a QIF neuron evolves according to the stochastic differential equation C d V = ( V 2 R + I ( t ) ) d t + D d W ( t ) ( 1 ) where C is the neuron’s membrane capacitance , R is the resistance associated with the leak current , W ( t ) is a white noise process with variance 1 , D is the amplitude of voltage noise , and I ( t ) is a source of time-varying external drive to the neuron . ( All quantities are without units . ) In our model , I ( t ) is a sum of excitatory and inhibitory post-synaptic currents ( EPSCs and IPSCs , respectively ) and a constant level of tonic drive . When V reaches a specific spiking voltage VS , it resets to a reset voltage VR < VS . In our model , VS = 1 . For the inhibitory cells , VR = −1 . We are interested in the synchronization of a single pulse that activates each excitatory cell only once ( emulating the sparse firing of RA-projecting cells in HVC ) , so once the voltage of an excitatory cell reaches VS , it is no longer recorded . All cells are divided between N inhibitory zones . Each zone contains P pools of Me principal cells . Pools are ordered and numbered 0 , … , NP − 1 in a chain that spirals through the zones P times so that pool p belongs to zone ( p mod N ) . The principal cells in pool p are numbered m = 0 , … , Me − 1 . Cell m in pool p projects one-to-one to cell m in pool p + 1 , forming a spiraling chain composed of Me parallel strands . When the system is initialized , spike times are chosen for cells in pool zero . For p > 0 , cell m in pool p is modeled by a QIF neuron with voltage V p m . We let t p m denote the firing time of excitatory cell m in pool p . At this time , an EPSC is initialized in cell m in pool p + 1 . The temporal profile of an EPSC in an excitatory cell is gee E ( t ) , where gee is the strength of excitatory-to-excitatory connections and E ( t ) is the evolution of a gating variable over time . We require that E ( t ) be a positive , continuous function with E ( t ) = 0 for t ≤ 0 and E ( t ) differentiable for t > 0 . Since the EPSC in cell m in pool p is initialized at time t p m , its height at time t is g e e E ( t - t p m ) . An additional tonic drive of magnitude IE is applied to each principal cell . Every zone n contains a collection of Mi inhibitory QIF neurons with voltages U n m for m = 0 , … , Mi − 1 . These neurons are excited by the firing of any excitatory cell in any pool in zone n: when an excitatory neuron fires and initializes an EPSC in its downstream excitatory cell , it also delivers an EPSC to all inhibitory cells in its zone . This EPSC is described by the function g e i M e E ( t - t p m ) , where gei is the strength of excitatory-to-inhibitory connections . The excitatory drive to each inhibitory cell is g e i M e ∑ p in zone n ∑ m = 0 M e - 1 E ( t - t p m ) , the sum of the EPSCs it has received from all excitatory cells in its zone . When an inhibitory cell fires in zone n , a sustained IPSC is delivered to all excitatory and inhibitory cells in that zone . A single inhibitory synaptic gating variable ϕn is incremented by k M i ( for some k > 0 ) each time a local inhibitory cell spikes , and decays exponentially with time constant Ti between spikes . Inhibition affecting excitatory and inhibitory cells in zone n is ϕn scaled by conductances gie and gii , respectively . Substituting a sum of excitation and inhibition for I ( t ) in Eq ( 1 ) , we have the following equations for excitatory and inhibitory neuron membrane potentials ( V p m and U n m , respectively ) : CedVpm= ( ( Vpm ) 2Re+geeE ( t−tp−1m ) −gieϕ ( p mod N ) +IE ) dt+DedXpm ( t ) CidUnm= ( ( Unm ) 2Ri+geiMe∑p in zone n∑m=0Me−1E ( t−tpm ) −giiϕn ) dt+DidWnm ( t ) dϕn= ( −ϕnTi+kMi∑sδ ( t−tns ) ) dt ( 2 ) where d X p m ( t ) and d W n m ( t ) are white noise processes with variance 1; { t n s } is the set of local inhibitory spike times , i . e . , times that U n m = 1 for any m , indexed by s; and δ ( t - t n s ) is a Dirac delta function that integrates to 1 at any local inhibitory spike time . This system must be initialized in simulation from a set of initial voltages V p m and U n m , a set of Me initial excitatory spike times t 0 m , and a set of N gating variables ϕn that determine the initial level of inhibition in each zone at time t = 0 . All voltages were initialized from zero at t = 0 . Initial excitatory spike times were set by drawing t 0 m from a Gaussian distribution with mean 0ms and variance 2ms2 . Inhibitory gating variables ϕn were initialized at a constant ϕ0 . Parameters were chosen for this model in order to produce the desired dynamics when local feedback was activated . Specifically , the model was built and tuned with the following objectives in mind: Most or all of the pool had to fire before most or all of the local inhibitory response . Thus , the firing of a pool of principal cells had to evoke local feedback inhibition with a sufficiently large delay . In order to implement this delay , we chose inhibitory cell membrane capacitance and resistance relatively large , slowing the response ( and , in particular , the membrane potential rise time ) of the inhibitory cells . A swifter inhibitory response , as might have been produced by cell with shorter membrane time constants or , e . g . , leaky integrate-and-fire neurons , would have produced competition between principal cells within the pool ( see , e . g . , [18] or [19] ) , which could play an important role in a sequence generating circuit of this type but was outside the scope of our study . At each spike volley , the decay of the inhibition produced by the previous local volley had to still be in progress . Thus , we had to choose an inhibitory decay time constant Ti that agreed roughly with the amount of time it took for an excitatory pulse to circle the loop once . If Ti was too large , inhibitory decay was too slow to produce noticeable synchronizing effects; if it was too small , the inhibition would be almost entirely gone by the time the pulse returned , with similar results . Synaptic conductances had to be tuned such that excitatory volleys consistently evoked responses in downstream excitatory and inhibitory cells , even when those cells were partially inhibited . As we note in the Discussion , propagation failure due to inhibition could help control for relaxed architectural constraints; however , pulse propagation failure was also outside the scope of our study . For feedback inhibition to improve the consistency of spike volley timing across trials , the response of the inhibitory populations to local spike volleys had to be consistent across trials . Running the simulation with a large number of inhibitory cells helped average out the effects of noise on the inhibitory population: when Mi was set to 1 , we did not observe improved timing across trials . As long as these four conditions were met , the effects of local feedback inhibition on spike volley synchronization and timing described below were robust to variation in model parameters . We simulated this system with two different sets of parameters . In simulation 1 , we set Me = 20 , Mi = 50 and N = 5 , and simulated the system with and without feedback inhibition . As we discuss below , this simulation run with feedback inhibition produced short periodic volleys of inhibitory spikes in each zone , an activity pattern considerably tidier than that of inhibitory cells observed in HVC . These dynamics obeyed conditions that made the system analytically tractable , allowing us to provide a quantitatively accurate theory explaining our simulation results . However , we also wanted to show that a more complex pattern of inhibitory activity could produce qualitatively similar results . For this purpose , we performed simulation 2 , for which we set Me = 50 and adjusted various parameters related to the inhibitory cells . All parameter values for both simulations are listed in Table 1 . For our simulations , we chose a simple but biologically-motivated function E ( t ) : E ( t ) = { 0 for t < 0 ( 1 - e - t τ r ) for 0 ≤ t < r ( 1 - e - r τ r ) e - t - r τ d for r ≤ t ( 3 ) where τr is the time constant for the rise of the EPSC , r is the duration of its rise , and τd is the time constant of its decay ( Fig 3 ) . We set τr = 9ms , τd = 5ms , and r = 8ms . We chose long rise times to mimic the ≈ 10ms duration of principal cell bursts in HVC [1] and the ≈ 10ms depolarizations observed in these cells and attributed to principal-to-principal cell excitatory potentials [2] . When we ran simulation 1 with no feedback inhibition , we set gie = 0 and IE = −0 . 3 in Eq ( 2 ) such that the level of drive to the excitatory cells was only the sum of a constant background and any incoming EPSC . We ran this simulation from the initial conditions described above . When we ran simulation 1 with feedback inhibition and simulation 2 , we set gie = 0 . 3 . This change increased the average level of inhibition , so we offset its effect by raising IE to −0 . 15 .
Our model produced a spatiotemporal pattern of coordinated excitation and inhibition . In each region , the periodic arrival of the excitatory pulse triggered a periodic local inhibitory feedback response; thus , a wave of blanket inhibition circulated among the regions following the excitatory pulse . This pattern created a local alternation between excitatory and inhibitory firing ( Fig 4 ) , reminiscent of the observation of cell-type specific phase preferences in the 30Hz component of the local field potential in HVC [13] . In simulation 1 , sharp volleys of principal cell spikes alternated with sharp volleys of inhibitory cell spikes ( Fig 4B and 4C ) . In simulation 2 , more inhibitory cells were included in each zone , inhibition between inhibitory cells was eliminated , and additional noise was added to inhibitory cell membrane potentials . In this simulation , inhibitory cells fired at all phases of the local cycle , but their firing rates increased after each local excitatory pool spiked and decreased again before the pulse returned to the same zone ( Fig 4D ) . Instead of excitatory and inhibitory spikes occurring in short , discrete , alternating volleys , excitatory spikes occurred during phases of reduced inhibitory spiking , resembling the excitatory spiking during pauses in inhibition observed by Kosche et al . [12] . Our simulations also agreed with the observation of Markowitz et al . [13] that the phasic coordination of principal cells and inhibitory cells was not global: locally , excitatory spiking was locked to an inhibitory cycle , but globally , excitatory spiking continued throughout that cycle ( Fig 5A ) . Upon examining individual excitatory spike rasters from our simulations ( Fig 5 ) , it was clear that local feedback inhibition promoted synchrony within pools of principal cells . Without feedback , there was no interaction between the m parallel strands of the chain , so the independent sources of noise caused the spike times within pools to drift apart; with feedback , the distribution of spike times instead remained tight as spiking propagated along the chain and around the ring of inhibitory zones . As a measure of within-pool synchrony and its evolution over time , we calculated the mean μp and variance vp of the spike times t p m in each pool p for each of 100 trials of simulation 1 ( Fig 6 ) . Without feedback , the trial-averaged variance vp of within-pool spike times increased without apparent bound . ( This is to be expected , since the spike timing along each strand of the chain is effectively a random walk independent of the other strands . ) When we introduced feedback inhibition , the trial-averaged vp instead decreased slightly and appeared to stabilize . Thus , feedback prevented the progressive desynchronization of pools and instead stabilized a tight distribution of spikes about the mean . To quantify the dependence of the synchronizing effect of local feedback on the model topology , we gradually introduced non-local E-to-I and I-to-E connections into the network in simulation 2 . For each trial , we instantiated a fraction F of the possible global E-to-I and I-to-E connections and then normalized connection strengths to ensure that the excitatory pulse still reached the last pool . We normalized connection strengths by modifying Eq ( 2 ) to read C i d U n m = ( ( U n m ) 2 R i + g e i M e + F P M e ∑ p in zone n ∑ m = 0 M e - 1 E ( t - t p m ) - g i i ϕ n ) d t + D i d W n m ( t ) d ϕ n = ( - ϕ n T i + k M i + F M i ∑ sδ ( t - t n s ) ) d t The synchronizing effect of local feedback inhibition persisted until the number of added global connections reached 50% of the number of local E-to-I and I-to-E connections ( Fig 7 ) . The synchronizing effect of local feedback inhibition on chained pools of cells can be understood as a specific instance of the more general phenomenon of synchronization by a slow-decaying pulse of shared inhibition described by Börgers and Kopell in [20] . A cell in pool p that receives its excitatory pulse earlier than the others in its pool also receives its pulse under a heavier blanket of inhibition , so its latency to spike is greater , whereas a cell that receives its pulse late can fire with reduced latency . Thus , decaying inhibition reigns in outlier spike times and forces spike times within a volley towards a shared mean . This intuition is explored more thoroughly in the Analysis section below . Observations of spike rasters across 100 trials of simulation 1 suggested that in addition to synchronizing local spike volleys , local feedback inhibition also made volley timing more consistent across trials . For each simulation trial i and each pool p , we calculated the mean spike time μ p i , and then took the variance v p t r i a l s of these means across trials and plotted it against p ( Fig 8 ) . Both with and without feedback , v p t r i a l s increased roughly linearly . However , we found that when pools were given local feedback inhibition , the rate of increase of v p t r i a l s was reduced . This finding is similar to the observation in [21] that , as a pulse in a synfire chain reaches a state of steady near-synchronous propagation , its propagation velocity shows less inter-trial variability . However , there are important differences between our result and theirs . In particular , they found that as the pulse approached its steady state , the time between successive spike volleys became more regular . In our data , feedback inhibition did not strongly affect the regularity of the time between the firing of pool p and pool p + 1 , but had a much more noticeable effect on the regularity of the time between the firing of pool p and the next pool in the same zone ( pool p + N ) . ( In simulation 1 , feedback decreased the cross-trial variance of the interval between mean spike times in pool 99 and pool 100 by 6 . 5% , as compared to a 28% decrease in variance of the interval between pools 95 and 100 . ) Thus , the reduction in inter-trial variability created by feedback inhibition was due to the stabilizing effect of lingering local feedback inhibition described above , which most directly influenced not the time of pulse propagation along excitatory connections , but rather the time for the pulse to circle the network and return . In order to more fully understand the cause of progressive synchronization in our model with feedback , we introduced additional assumptions and approximations to make the model analytically tractable . These assumptions and approximations made it possible to linearize the dynamics around a set of excitatory spike times and allowed us to express the relationship between model parameters and the stabilization of synchrony in terms of the solution to a first passage time problem . Combining this analysis with a computational investigation of the first passage problem , we found that we could quantitatively describe the effects of IPSC decay rate , EPSC rise rate , and noise on pool synchronization . Simulation 1 met our additional assumptions , and we found that the synchronizing behavior of simulation 1 indeed agreed with our analytical predictions . Our analysis demonstrates that progressive synchronization by feedback inhibition is not a special property of a finely tuned computational model but a generic property of spatially recurrent feedforward chains with local feedback inhibition .
We have put forward a model of sequence generation based on recent experimental findings in the songbird [12 , 13] . In this model , a feedforward chain of excitatory neurons passes repeatedly through multiple zones of inhibition , triggering local feedback inhibition in each . We have shown that this model can generate stereotyped neural sequences , creating synchrony among pools of cells through shared inhibition and stabilizing inter-trial spike timing . These effects can operate in place of ( or , presumably , in cooperation with ) the similar effects of the redundant feedforward excitatory connectivity that characterizes the synfire chain . Though previous models of neural sequence generation have used inhibition in a variety of ways [14 , 22–25] , they are all fundamentally distinct in structure and dynamics from our model . In “winnerless competition” models inspired by the dynamics of insect olfaction , Rabinovich et al . [22] generate sequences through competitive inhibitory interactions in a randomly connected network . Verduzco-Flores et al . [24] and Assisi et al . [23] use a network of excitatory and inhibitory units to learn and generate sequences that propagate using a combination of disynaptic inhibition and adaptation currents . A series of modeling papers have proposed that sequences are generated using strong global ( not local ) inhibition to select between multiple possible synfire chains [9 , 10 , 14 , 15 , 26] . Some of these have incorporated precise spatial constraints on otherwise global inhibitory connectivity in order to disinhibit principal cells at the appropriate times [14 , 15] . Like us , Gibb et al . [14] and Bertram et al . [25] explore models of feedforward chains of excitatory cells with local inhibition , but they give their excitatory chains no manner of spatial recurrence , so the local inhibition evoked by a pool of excitatory cells cannot affect the spiking of cells at a later point along the chain . Our model is unique in the use of spatially recurrent excitatory chains , and in the use of local feedback inhibition to stabilize synchrony ( rather than , e . g . , to propagate spiking by inducing rebounds as in [25] ) . We have shown in simulation that our model of an excitatory chain spiraling through inhibitory zones creates local alternation between excitatory and inhibitory cells , consistent with the observation of cell-type specific phase preferences in the 30Hz component of the local field potential in HVC [13] . Moreover , it reproduces the observation that spiking is not globally phase-coordinated , but occurs continuously throughout song . None of the models discussed above produce such a firing pattern—it is a natural consequence of localized inhibition and spatially recurrent excitatory activity , the same factors that differentiate our model from previous work and produce its synchronizing and temporally stabilizing dynamics . Our model is also consistent with paired recordings in slice , which have shown that excitatory neurons in HVC contact each other primarily through disynaptic inhibition [12 , 27] as would be expected in a network dominated by local inhibitory feedback and with only sparse , specific monosynaptic connections between principal cells . Finally , there is small but growing evidence that HVC activity is correlated over space [28 , 29] and that HVC connectivity is spatially structured [12 , 30–32] , consistent with a model in which spatial regions of HVC act as inhibitory zones . However , our model deviates from what is known experimentally in two important respects . First , interneurons in our model fire with strong periodicity , yet HVC interneurons have dense firing patterns with intermittent periodicity during singing [13 , 33] ( see S2 Fig ) . Additionally , the stereotyped ≈ 30 Hz LFP , which is correlated with interneuron firing , is not a perfectly periodic signal [13] . The periodicity in the model is a result of its highly simplified structure . If the topological structure of the chain were more complex than a simple spiral , inhibitory activity might more closely resemble what has been observed experimentally; however , this is beyond the scope of the current study . Second , our model implies that inhibitory interactions between HVCRA neurons should be primarily localized to a subregion of HVC , but recent evidence suggests that HVCRA pairs inhibit each other over relatively large distances ( hundreds of μm ) [12] . Our simulations with added global connectivity ( S2 Fig ) show that our model is robust to between-zone disynaptic inhibition up to a 2:1 local-to-global connection ratio . Moreover , future experiments will be needed to directly observe whether disynaptic inhibition between HVCRA neurons is in fact local or global . As constructed for this study , our model has the capacity to play back only one sequence . In the case of the zebra finch , which learns only one song , this is an appropriate constraint , but this limitation would have to be addressed for broader applications . Storage of and selection between multiple sequences has been explored by other authors , both in HVC [9 , 14 , 15] and more generally [10 , 26 , 34 , 35] . We note that the spatial recurrence which is central to our model could be exploited for this purpose: multiple disconnected chains passing through the same sequence of regions could activate the same cycle of local feedback inhibition and benefit from the same stabilization of timing . We have demonstrated in simulation and through proof that the presence of shared decaying inhibition progressively synchronizes the firing of pools of excitatory cells . We have also derived a specific asymptotic upper bound for the expected variance that decreases with increasing bmin , where bmin represents an upper bound on the magnitude of the decay rate of inhibition during local excitatory spiking . The more sharply the local inhibition is decaying during the spiking of a pool , the larger a value we can choose for bmin . Consequently , the more sharply local feedback inhibition is decaying when a pool of cells spikes , the tighter the resulting synchrony guaranteed by our analysis . ( We note that there is a non-trivial relationship between the exponential time constant Ti of inhibitory decay and the instantaneous rate of inhibitory decay—since inhibitory decay is exponential , the latter also depends on the level of inhibition and thus on the recent inhibitory spiking history . ) We have also shown in simulation that local feedback inhibition creates sequences with timing that is more stereotyped across trials . One possible intuitive explanation for this effect is that the inhibitory state of each zone stores information about the timing of the most recent local volley; thus , the drift in volley timing due to noise can be partially corrected as the excitatory pulse reaches each zone . In other words , the information about the timing of the previous volley delivered by E-to-E connections and the information about the timing of the most recent local volley delivered by I-to-E connections are both incorporated to determine the timing of each spike volley . Our model is closely related to the mechanism of “communication through resonance” developed by Hahn et al . [36] . In their model , pools of cells are synchronized by cycles of inhibition evoked by the arrival of external periodic excitatory pulses , whereas in our model pools of cells are synchronized by cycles of inhibition evoked by a single excitatory pulse as it returns periodically to each inhibitory zone . A similar model developed by Jahnke et al . [37] supports synchronous propagation of pulses through a network by imposing global oscillations resonant with transmission delays . Their model requires net-excitatory oscillating input or nonlinear coupling in order to keep the pulse from dying while the network is inhibited . In our model , these global oscillations are replaced by multiple local oscillations at different phases . Although the oscillations involve periods of inhibition , our network is never globally inhibited , circumventing this possible cause of propagation failure . It is important to note the relationship between our model and the ideas discussed by Long et al . in [2] . They show that principal cells in HVC routinely receive strong depolarizations immediately before spiking . They use this observation to support an excitatory chain model and reject a model in which cells fire as a ramp of excitatory drive pushes them sequentially past their firing thresholds . Our model effectively combines these two mechanisms: cells receive strong depolarizations due to the excitatory chain structure embedded in the network , but the timing of spikes is also influenced by more gradual downward ramps of inhibition that affect ( but do not fully determine ) the time that the drive to each cell crosses threshold . Since the timing of sequential activity is influenced by the local network state as well as the arrival of an excitatory pulse , a model of this type would be better suited than a synfire chain to produce sequential activity that is non-uniform in time [38] . It could be argued that a chain of excitatory connections with periodic spatial recurrence is biophysically implausible . However , our model only requires that excitatory activity pass through inhibitory zones sequentially and return to them regularly; preliminary simulations suggest that such a pattern of activity may be achievable on a random excitatory network grouped into zones with localized inhibitory feedback . As an excitatory pulse propagates through the network , it is followed by a wave of local inhibition . This inhibition prevents excitation from returning to a zone until it has decayed sufficiently . Once it has decayed , random percolation ensures that activity does return . The result is a pulse that passes through the zones in an order determined by connectivity and initial levels of inhibition , and that meets the decaying inhibition from its previous pass as it reaches each zone . This pulse may activate different cells on each pass , generating a sequence significantly longer than a single pass through the network . The extraneous excitatory connections in such a random network would trigger EPSCs in some cells while they remained under strong inhibition , potentially explaining the observation of multiple stereotyped depolarizations in HVCRA cells during zebra finch song [12] . Furthermore , as regions are repeatedly activated in the same order , the network might use spike-timing-dependent plasticity to learn the excitatory connections necessary to reliably reproduce this pattern , ultimately creating the circulating chain architecture assumed in our model . In Fig 7 , we show that synchronization through local feedback is somewhat robust to connection noise—once network connectivity became sufficiently localized , feedback inhibition would begin to contribute to pulse synchronization . For this mechanism to function , excitatory activity returning to a region of HVC must arrive during the decaying slope of the inhibition , and must therefore circulate through the inhibitory zones on a time scale matching the decay of inhibition . In our model , this is achieved through manual tuning . However , a learning process in HVC like the one described above might help to match these timescales by ensuring that excitatory activity returns to a zone as soon as inhibition decays sufficiently to permit it . Alternatively , the timescale of recurrence may relate to the cortical-thalamic loop cycle time , which does appear to match the time-constant of inhibitory decay [12] . In this view , the spiral does not exist entirely in HVC , but instead passes through the cortical-thalamic loop . Our model also may throw a new light on certain dynamics in the mammalian brain . Two brain regions in which precisely-timed sequential activity is thought to be essential are the motor cortex [39] and hippocampus [40] . Both of these regions have been shown to support rhythmic traveling waves , at beta frequencies ( 15–30Hz ) [41] and theta frequencies ( 4–10Hz ) [42] respectively . We suggest that these waves may be a manifestation of the locally-coordinated , globally out-of-phase inhibitory cycles that characterize our model and help synchronize neuronal pools and stabilize timing within a firing sequence .
|
Sequences of stereotyped actions are central to the everyday lives of humans and animals . It was hypothesized over half a century ago that these behaviors were enabled by linking together groups of neurons ( or “cell assemblies” ) into a feedforward chain using correlation-based learning rules . These chains could then be activated to generate particular behavioral sequences . However , recent data from HVC ( the songbird analogue of premotor cortex ) paint a more complicated picture: inhibitory and excitatory cells lock to different phases of a rhythm , with inhibitory cells providing windows of opportunity for the excitatory cells to fire . This study puts forward a mathematical model that uses both a feedforward chain geometry and local feedback inhibition to generate stereotyped neural sequences . The chain conducts an excitatory pulse through multiple spatial regions , arriving at each as local inhibition dips . Our simulations and analysis demonstrate that such patterned local inhibition can synchronize the firing of pools of neurons and stabilize spike timing along the chain . Our model provides a new way of thinking about sequence generation in the songbird and in neural circuits more generally .
|
[
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] |
[] |
2015
|
Neural Sequence Generation Using Spatiotemporal Patterns of Inhibition
|
Identification of early mechanisms that may lead from obesity towards complications such as metabolic syndrome is of great interest . Here we performed lipidomic analyses of adipose tissue in twin pairs discordant for obesity but still metabolically compensated . In parallel we studied more evolved states of obesity by investigating a separated set of individuals considered to be morbidly obese . Despite lower dietary polyunsaturated fatty acid intake , the obese twin individuals had increased proportions of palmitoleic and arachidonic acids in their adipose tissue , including increased levels of ethanolamine plasmalogens containing arachidonic acid . Information gathered from these experimental groups was used for molecular dynamics simulations of lipid bilayers combined with dependency network analysis of combined clinical , lipidomics , and gene expression data . The simulations suggested that the observed lipid remodeling maintains the biophysical properties of lipid membranes , at the price , however , of increasing their vulnerability to inflammation . Conversely , in morbidly obese subjects , the proportion of plasmalogens containing arachidonic acid in the adipose tissue was markedly decreased . We also show by in vitro Elovl6 knockdown that the lipid network regulating the observed remodeling may be amenable to genetic modulation . Together , our novel approach suggests a physiological mechanism by which adaptation of adipocyte membranes to adipose tissue expansion associates with positive energy balance , potentially leading to higher vulnerability to inflammation in acquired obesity . Further studies will be needed to determine the cause of this effect .
Obesity is characterized by excess body fat , which is predominantly stored in the adipose tissue . Obesity is considered one of the pathological features of metabolic syndrome ( MetS ) , which also includes insulin resistance , hypertension , and dyslipidemia [1] . Although not all obese individuals develop metabolic and cardiovascular complications , the clustering of these conditions of MetS suggests there may be pathogenic mechanisms common to all these phenotypes [2] , [3] . The specific mechanisms that may lead from obesity towards the higher risk of metabolic complications such as insulin resistance and type 2 diabetes ( T2D ) remain elusive . Our preferred nonexclusive hypothesis is the “adipose tissue expandability” hypothesis [2] , [4] , which states that obesity-associated metabolic complications such as insulin resistance are due to the finite capacity of adipose tissue to expand and therefore to store energy . In fact , once this limit of expansion is reached and the storage capacity of adipose tissue is exceeded , the lipids become deposited ectopically , leading to potentially toxic effects in peripheral tissues via the excessive accumulation of reactive lipid species . In support of the pathogenic role of limited adipose tissue expandability and functionality we have previously shown in a genetically obese leptin-deficient mouse ( ob/ob ) model that ablation of adipogenic PPARγ2 ( the POKO mouse ) leads to a 65% decrease in adipose tissue mass as compared to ob/ob mouse , accumulation of reactive lipids in pancreatic islets , liver , and muscle , and severe insulin resistance and diabetes [5] . Interestingly , adipose tissue of the POKO mouse is also characterized by an altered membrane phospholipid profile , characterized by diminished levels of plasmalogens , the most abundant ether lipids [5] . The opposite experimental paradigm , increasing the capacity of adipose tissue expansion by overexpressing adiponectin in white adipose tissue of ob/ob mice ( the AdTG-ob/ob mouse ) , led to a considerably improved metabolic profile as compared to ob/ob mouse [6] . Based on this evidence we hypothesized that adaption to the demands posed by positive-energy-balance-induced adipose tissue expansion may challenge the homeostatic mechanisms controlling phospholipid composition and their associated biophysical properties . However , given the importance of membrane lipid composition and fluidity to maintain the topology , mobility , or activity of membrane-bound proteins , and to ensure normal cellular physiology , we also speculated that the process of adipose tissue expansion should also incorporate allostatic adaptations aiming to maintain membrane functionality for as long as possible [7] , particularly under metabolically challenging conditions . In this respect we speculated that exhaustion of these allostatic adaptations may determine the maximal limit of expansion and may coincide with metabolic perturbations . Investigations of adipose tissue “lipidome , ” covering a global profile of structurally and functionally diverse lipids , provide a unique opportunity to pursue accurately and sensitively studies profiling hundreds of molecular lipids in parallel [8] . The spatial complexity of lipid metabolism presents a challenge for the study of the global lipidomic profiles . Studies of lipidomic profiles in the context of biochemical pathways may shed light on processes behind the synthesis of or regulation by specific lipids , but cannot address how these changes translate into lipid membrane properties and affect cellular physiology . However , molecular modeling tools and large scale computing capacities are becoming available that facilitate modeling of lipid bilayers in the context of their composition and function [9] , thus opening new opportunities to interpret regulation and changes of global lipidome also in the spatial and physiological context . Given the recent increase in the prevalence of obesity and MetS , it is clear that on top of their unquestionable polygenic component , other environmental factors such as lack of physical activity or high caloric diets are likely contributors to their progressive acceleration . However , individual heterogeneity in genetic and environmental profiles makes the search for specific clues and mechanisms facilitating adipose tissue expansion and its related metabolic complications in a general population a daunting task . One suitable clinical study design setting to at least homogenize and eliminate the genetic component of this challenge is the twin design . Study of monozygotic ( MZ ) twin pairs , discordant for body weight , provides an opportunity to explore the initial effects of acquired obesity and related complications since these individuals share not only an identical genetic background at the DNA sequence level , but also early life events and family environment . Identification of mechanisms behind the traits related to acquired obesity and MetS is also relevant from the therapeutic point of view , as these mechanisms may point to targets for specific disease phenotypes that are not related to specific genetic makeups . To obtain a global view of human adipose tissue lipidome in different degrees of acquired obesity , here we perform lipidomic analyses of adipose tissue in twin pairs discordant for obesity but still metabolically compensated . In parallel , we studied more evolved states of obesity by investigating a separate set of morbidly obese subjects . Information gathered from these experimental groups was used for molecular dynamics ( MD ) simulations of lipid bilayers using bioinformatics approaches , and the conclusions were further supported by in vitro adipocyte confirmatory studies . This approach uncovered the potential physiological mechanisms by which adipocyte membranes adapt to adipose tissue expansion associated with positive energy balance , ultimately leading to obesity . Furthermore , we demonstrate how in extreme obesity , failure of this adaptation is associated with the pathological metabolic manifestation of obesity .
We first investigated acquired obesity independent of genetic influences in 13 MZ twin pairs discordant for body mass index ( BMI ) , and nine BMI-concordant MZ twin pairs , of which five pairs were overweight ( Table S1 ) . The obese individuals were young adults and were considered healthy obese , i . e . , with no clinical comorbidities , and thus in the early stages of the development of obesity . In this cohort , we have previously found that obese individuals , as compared to their healthy lean twins , already show signs of insulin resistance [10] , exhibit a pro-inflammatory serum lipidomic profile [11] , have diminished mitochondrial DNA copy number and dysregulated expression of mitochondrial pathways such as branched chain amino acid catabolism , and have elevated inflammatory and immune response pathways in the adipose tissue [12] . Analysis of the weight-discordant twin pairs showed that the obese twins were on average 15 . 2 kg ( 20% , 5 . 3 kg/m2 ) heavier than the non-obese twins and had more fat subcutaneously , intra-abdominally , and in the liver ( Table S1 ) . Average adipocyte diameter was 17% larger in the obese than in the non-obese twins , and , interestingly , there was also a very high degree of intra-pair similarity of fat cell size ( FCS ) ( Figure S1 ) . As expected , adiponectin levels were lower , and leptin and hsCRP levels higher , in the obese twins . The obese twins were also more insulin resistant , as evidenced by the lower M-value and higher fasting plasma glucose and serum insulin levels . Among the dietary intake variables , the proportional intake of polyunsaturated fatty acids ( PUFAs ) was 26% lower in heavy twins as compared to their lean counterparts ( p<0 . 05 ) . The analysis of weight-concordant twin pairs showed no intra-pair differences in FCS , adipokines , or insulin sensitivity ( Table S1 ) . To characterize the adipose tissue lipidome in acquired obesity , we applied the global lipidomics approach using ultra performance liquid chromatography coupled to mass spectrometry ( UPLC-MS ) . The analysis was performed in positive ion mode ( ESI+ ) , which is sensitive to neutral lipids ( such as triacylglycerols ) and major phospholipid classes such as phosphatidylcholines ( PCs ) , phosphatidylethanolamines ( PEs ) , and sphingomyelins . A total of 313 lipids were detected and quantified in each of the 44 samples analyzed . Additionally , free cholesterol was determined by gas chromatography coupled to mass spectrometry ( GC-MS ) , and the data were included in the lipidomic dataset . Since we considered that adipocyte size might be a factor affecting membrane and cell lipid composition , we investigated the lipidome redistribution in relation to the cell size . Thus , following the calibration with internal standards we further normalized the data by the total amount of detected phospholipids in each sample . This allowed us to investigate the relative compositional changes of adipose tissue lipids . When comparing the weight-discordant twins , lipidomic analysis revealed characteristic differences in cellular phospholipids ( Figure 1A ) . The dominating characteristic of adipose tissue in the obese twins was the elevation of PUFA-containing phospholipids , which were predominantly ether lipids , and proportional diminishment of phospholipids containing shorter and more saturated fatty acids . These lipids were clustered according to BMI in both the discordant and concordant twin pairs , suggesting that these observed changes are characteristic of adipose tissue expansion per se , irrespective of genetic makeup . Figure 1B further illustrates this by showing the normalized concentrations of the most abundant significantly altered lipids . The most abundant PUFA-containing ether lipids were confirmed to be plasmalogens ( Figure S2 ) . Partial least squares regression of lipidomic data related to FCS showed that the changes in the top-ranking lipids associate with the increase of adipocyte cell size in obesity ( Figure S3 ) . In line with this , triacylglycerols were also elevated in the adipose tissue of obese twins at the marginal significance level false discovery rate ( FDR ) q<0 . 1 ( Figure S4 ) . Free cholesterol in adipose tissue did not differ significantly between the obese and lean twins ( Figure S5 ) . The observed changes in phospholipids in acquired obesity appear to be highly selective , showing both functional group as well as fatty acid specificity . It is known that PUFAs are selectively targeted to plasmalogens [13] , [14] , which could explain enrichment of adipose tissue ether lipids in acquired obesity ( Figure 1 ) . Notably , among the elevated plasmalogens observed in the obese individuals , the fatty acid dominantly esterified in the sn-2 position is arachidonic acid and not docosahexanoic acid ( DHA ) , which is also commonly found in plasmalogens [13] . To investigate whether specific global fatty acid compositional changes occur in the adipose tissue in response to weight gain , we analyzed fatty acid profiles in weight-discordant twin pairs . We found marked differences in several fatty acids ( Table S2 ) , which included diminishment of stearic ( C18:0 ) , linoleic ( C18∶2n6 ) , and α-linolenic ( C18∶3n3 ) acids and elevation of palmitoleic ( C16∶1n7 ) and arachidonic ( C20∶4n6 ) acids ( Figure 2A ) . In relation to known dietary intake , the data are consistent with diminished PUFA intake ( decreased linoleic and α-linolenic acids ) , while other changes appear to reflect an induced specific program of fatty acid desaturation and elongation ( Figure 2B ) , which leads to elevated palmitoleic acid via desaturation of palmitic acid , and to elevated long-chain PUFAs up to arachidonic acid , but not to DHA . The specific changes in PUFA composition are consistent with the observed changes in phospholipids . However , the elevation of esterified palmitoleic acid in the adipose tissue of the obese twins was not reflected in significantly elevated free palmitoleate levels in serum ( Table S3 and Figure S6 ) . The observed remodeling of membrane phospholipids in expanding adipocytes of obese twins suggests that these changes may also have an effect on membrane properties such as membrane fluidity ( packing ) , order , and thickness , which are known to affect cellular physiology [15] , [16] . There is evidence from in vitro model systems and from atomic-scale MD simulations of lipid bilayers that an increase in PUFA content will increase membrane fluidity [17]–[19] . The lateral diffusion of lipids [20] and the permeability of the membrane to small molecules [21] are dependent on the fluidity through packing and degree of order in the bilayer . However , the effect of the vinyl-ether bond in plasmalogens on membrane fluidity is poorly understood and has not yet been modeled using MD simulations . To study the consequences of altered lipidomic profiles on membrane biophysical properties , we performed atomic-scale MD simulations of eight different membrane systems . Based on measured abundance of differentially regulated lipids ( Figure 1 ) , the following lipid types were used in simulations: PC ( 16∶1/18∶0 ) , PC ( P-16∶1/18∶0 ) , PC ( 16∶0/20∶4 ) , PC ( P-16∶0/20∶4 ) , PE ( 16∶0/20∶4 ) , and PE ( P-16∶0/20∶4 ) . Structures of PC ( 16∶0/20∶4 ) and PC ( P-16∶0/20∶4 ) are shown in Figure 3A . In addition to six one-component bilayers composed of these lipids individually , we also studied two additional mixtures of PC ( 16∶1/18∶0 ) and PE ( P-16∶0/20∶4 ) with PE concentration of 59 mole percent ( low BMI mix ) and 70 mole percent ( high BMI mix ) . The mixtures were designed to mimic the compositional difference between the obese and non-obese twins ( Figure 1B ) . The selection of the specific six lipid molecules was motivated by the need to characterize specific effects due to three different types of structural changes observed in the lipids: ( 1 ) functional group ( PE versus PC ) , ( 2 ) vinyl-ether versus ester bond in sn-1 position , and ( 3 ) degree of saturation in the sn-2 chain . There is reason to stress that considering all possible lipid combinations in membranes is not feasible because of the major computational cost . Therefore , the above choice of the eight model systems aims to clarify the overall effects on membrane fluidity arising from the three different structural changes observed in lipids , but with a reasonable cost . Yet the total simulation time is major , about 1 µs . Figure 3B displays the average surface area per lipid , describing the overall packing of the lipids within the membrane plane . The larger the area per lipid , the more fluid the membrane is . Figure 3C shows that when plasmalogens are replaced with the corresponding ester lipids , the area per lipid increases by about 0 . 01–0 . 02 nm2 . An increase of 0 . 07–0 . 09 nm2 is found when PE is replaced by PC headgroup . Finally , increasing the unsaturation of PC acyl chains increases the average surface area per lipid by 0 . 02–0 . 03 nm2 . Together , membrane fluidity is promoted by decreasing plasmalogen concentration , using PC instead of PE headgroup , and increasing PUFA content . The snapshots in Figure 3D illustrate different packings and thus different degrees of fluidity in four of the studied bilayers . As expected , membrane thickness is negatively correlated with area for all studied systems . Furthermore , increased lateral packing correlates with a higher conformational order of the acyl chains . To illustrate this point , Figure 3C shows the inverse of molecular order parameter ( Smol ) averaged over the saturated segments of the sn-1 acyl chains . It is clear that the sn-1 chains of plasmalogens are more ordered than the corresponding chains in the ester-bonded lipids . To our surprise , no marked differences in the surface area and thus fluidity are observed when high and low BMI lipid mixtures are compared ( Figure 3B ) . The data therefore suggest that the increased fluidity due to elevated PUFA content in membrane phospholipids is compensated for by decreased fluidity of the elevated PE plasmalogen lipids , with the final result that there is no change in membrane fluidity and thickness . The membrane clearly has compensatory mechanisms to maintain its fluid nature . Next , we investigated the regulatory mechanisms that may be behind the observed lipid remodeling . Because of the intrinsic complexity of lipid metabolism , interactions of multiple components are likely involved in the regulation of lipid changes [22] . To capture such functional interconnections between the biological entities , we considered a network-based approach to be more suitable than studies of differential expression changes at the individual gene level . We selected ten clinical variables , 31 lipid-metabolism-related genes from the published dataset [12] obtained from the same samples used in the present study , two pathway profiles reflecting major changes observed in pathway analysis [12] , and ten lipid variables representative of major changes observed in adipose tissue lipid profiles ( Table S4 ) . As the only “input” variable , we used dietary PUFA intake ( PUFA percent ) . To distinguish direct and indirect interactions of these variables , we utilized the QPGRAPH method , which has been previously applied to study gene regulatory networks based on microarray data [23] . The key idea of QPGRAPH is to use partial correlations as a measure of dependency and build an undirected Gaussian graphical model where the variables are connected if and only if their partial correlation is significantly non-zero . Unlike the pairwise measure of associations , e . g . , Pearson correlation coefficients , partial correlation provides a stronger criterion for dependency by adjusting for confounding effects , and thus removes spurious associations to a large extent . This is particularly favorable for such an integration of multiple layers of information , as it inherently filters out false positives by discovering only those direct interactions with high confidence . The data-driven dependency network in twins discordant for obesity is shown in Figure 4 . Notably , PUFA percent was connected to CD36 , an important fatty acid transporter [24] , which was further connected to fatty acid elongase Elovl6 . In such a network context , identification of genes connected to many other genes or variables of interest , i . e . , so-called network hubs , is of particular interest . For example , despite not being differentially regulated itself , Elovl6 appears to be an important hub , with seven connections , including desaturase SCD1 . Among the other important regulators of lipogenesis , PPARγ was significantly down-regulated in obese twins , and , along with fatty acid elongase Elovl4 , which was not differentially regulated , was associated with the decreased fatty acid ratio of C22:5 versus C20:5 , a step which appears to affect the balance between the amounts of arachidonic acid and DHA ( Figure 2 ) . In agreement with earlier findings [25] , the decreased expression of PPARγ was also associated with decreased expression levels of insulin receptor substrate 2 ( IRS2 ) . We then investigated whether the observed network might be amenable to modulation of the lipid profiles observed in the weight-discordant twins . Given its position as a network hub ( Figure 5 ) and its known regulatory role in the control of cellular fatty acid composition [26] , we hypothesized that ablation of Elovl6 might be an upstream regulator of the lipid remodeling observed in obese twins , sensitive to dietary stimulus such as relative decrease in PUFA intake . To gain further insights we hypothesized that ablation of Elovl6 in the 3T3-L1 adipocyte cell line might reveal a mirroring lipid pattern . In fact , lipidomic analysis of preadipocytes and mature adipocytes revealed that ablation of Elovl6 leads to a lipid profile opposite to the one observed in obese twins , thus supporting our model ( Figure 5 and Table S5 ) . Specifically , knock-down of Elovl6 leads to a reduction in PUFA-containing phospholipids , which are predominantly ether-bonded , and to a proportional elevation of shorter and more saturated phospholipid species . To investigate the lipid profile in pathogenic stages of obesity , we compared the adipose tissue lipid compositional changes in obesity-discordant twin pairs ( Figure 1B ) with lipidomic profiles from our recent study of eight morbidly obese subjects recruited among patients undergoing laparoscopic surgery for the treatment of obesity [27] . The BMI range was 47 . 0–60 . 4 kg/m2; four subjects had elevated fasting serum insulin ( >10 mU/l ) , and among these two subjects were diagnosed with T2D . The levels of shorter and more saturated phospholipids were similar in the morbidly obese subjects and healthy twin pairs ( Figure 6A ) . However , the proportion of PUFA-containing ether lipids was markedly lower ( Figure 6B ) , indicating that the mechanism of lipid remodeling observed in healthy obese subjects had broken down or that it was unable to compensate at that level of stress . In contrast to the obese twins , the linear dependence of fasting serum insulin on FCS appears to have broken down in the morbidly obese subjects ( Figure 6C and 6D ) . The major outliers were the two subjects diagnosed with T2D , who had the smallest FCS and the highest fasting serum insulin concentrations ( Figure 6D ) .
Adipose tissue is a highly active endocrinometabolic organ whose main role is to provide efficient storage and mobilization of lipids to fulfill bioenergetic demands . Adipocyte function depends on the homeostasis of important cellular lipid mediators and lipid structural components of biological membranes required for accurate functional responses . These lipid functions are interconnected , e . g . , the membrane lipids also serve as precursors of lipid mediators . Modern lipidomics technologies enable characterization of cellular lipidomes across multiple structural and functional groups . However , the analysis and interpretation of data from such datasets are commonly limited to the level of biochemical and signaling pathways . In this study , we presented a new approach to study global lipidomes in tissues and cells by combining cellular lipid networks with lipid membrane modeling . This strategy enabled us to identify adaptive mechanisms that may lay behind the characteristic remodeling of the adipose tissue lipidome in response to positive-energy-balance-induced adipose tissue expansion during the evolution of obesity . There is evidence suggesting that the total number of adipocytes in humans tends to remain constant during a lifetime [28] under physiological conditions . Consequently the excess lipid load in acquired obesity poses an extra layer of stress on the adipose tissue , which can be compensated by increasing the size of the adipocytes and/or overstimulating the normal processes of adipocyte turnover in the adipose tissue . This relative stability of the adipose tissue cellular composition requires active regulatory mechanisms . In agreement with this perspective , our data show that FCS is increased in obese individuals relative to their lean twins in weight-discordant MZ twin pairs ( Figures 6C and S1 ) . We also observed a high degree of intra-pair similarity of FCS in weight-concordant MZ twin pairs ( Figure S1 ) , which suggests a strong genetic component tightly regulating FCS . These changes in cell size may lead to and/or be affected by minor variations in membrane composition , resulting in changes in membrane fluidity and lateral pressure [9] . It would be expected that these changes would consequently affect the membrane protein function and thus cellular physiology . In this context we consider adipose tissue expansion in response to positive energy balance to be a challenge for the maintenance of membrane integrity and function , requiring potent adaptive allostatic responses . With adipose cell expansion , more phospholipids have to be incorporated into the cellular membranes . However , this process also requires a very accurate quality control system that ensures that irrespective of the available lipid pool , the composition of the membrane and its functionality is appropriate in expanding adipocytes and newly recruited adipocytes . We found that the expansion of adipose tissue is accompanied by a proportional increase of PUFA-containing ether lipids and a decrease of more saturated and shorter-chain ester-bound lipids ( Figure 1 ) . These changes were also reflected in the altered fatty acid composition ( Figure 2 ) . Given that no increase was observed in saturated short-chain fatty acids ( Figure 2 ) , it is unlikely that de novo fatty acid synthesis could explain the observed fatty acid profile . Of interest , proportional dietary intake of PUFA was lower in obese twins than in their lean counterparts , and this was reflected in the diminished linoleic and α-linolenic acids ( Figure 2 ) . However , paradoxically , the concentration of arachidonic acid was increased in the obese twins , as was also reflected in the altered profile of the phospholipids . Notably , this increase was specific to arachidonic acid since there was no difference between the levels of DHA of the obese and lean twins ( Figure 2 ) . The observed lipid changes are not systemic since we have previously shown that the ether lipids are down-regulated in the serum of the obese twins [11] . Our findings support the view that positive energy balance leading to obesity initiates a program of membrane lipid remodeling in the adipose tissue , involving increased biosynthesis of unsaturated fatty acids including arachidonic acid ( Figure 7 ) . In agreement with earlier findings [29]–[31] , arachidonic acid is actively incorporated into the membrane phospholipids and selectively targeted to ethanolamine plasmalogens ( Figure 1 ) . However , if the only membrane compositional changes were these specific changes in the fatty acid composition , the membrane would become more fluid [19] . Instead , our membrane simulations showed that the targeting of PUFAs to ether lipids , and preferentially to the most abundant ethanolamine plasmalogen , helps to maintain membrane biophysical properties . In fact , it is known that a relative increase in the PE/PC ratio contributes to increasing the rigidity of the membrane [32] and therefore compensating for increased fluidity mediated by unsaturated fatty acids . Here we also provided evidence based on biophysical modeling that supports the view that the presence of a vinyl-ether bond in the sn-1 position of phospholipids contributes to further stiffening of the membrane . The adaptation of adipose tissue membranes in obesity may carry a price . In fact , plasmalogens can serve as antioxidants against reactive oxygen species and might thus protect the cells from oxidative stress [33] , [34] . However , when arachidonic-acid-containing plasmalogens are under oxidative stress , they become precursors of arachidonic-derived lipid mediators such as leukotrienes and hydroxyeicosatetraenoic acids [29] , [31] , [35] , [36] . These reactive lipids are important mediators of inflammatory response [37] . Thus , this remodeling process occurring in the membranes may make the adipocytes more vulnerable and prone to inflammatory responses ( Figure 7 ) , in agreement with the elevation of inflammatory pathways in the obese twins of weight-discordant twin pairs [12] . In the context of inflammation , a crucial step in the observed lipid remodeling is the relative decrease of C22:6 ( DHA ) and C22:5 as compared to C20:5 , leading to a relative increase of arachidonic acid content relative to DHA in acquired obesity ( Figure 2 ) . In contrast to arachidonic-acid-derived lipid mediators , the DHA-derived lipid mediators tend to have anti-inflammatory properties [38] . We propose that stimulation of fatty acid elongation from C20:5 to C22:5 and desaturation from C22:5 to C22:6 could be one strategy to maintain the membrane function in acquired obesity , while also avoiding the collateral damage of increasing the vulnerability of adipose tissue to inflammation . We observed a high degree of linearity in FCS variation in weight-discordant twin pairs ( Figure S1 ) . This suggests that adipocytes of these relatively modestly obese twins are still within the range of normal expansion and meeting the demands imposed by their specific nutritional demands . However , this physiological process appears to be disrupted in states of excessive adipose expansion beyond homeostatic capacity in morbidly obese subjects , characterized by decreased concentrations of ethanolamine plasmalogens compared to the “healthy obese” twins ( Figure 6B ) . In morbidly obese individuals , fasting serum insulin does not correlate with FCS ( Figure 6D ) . The two diabetic subjects in this group had , in fact , the smallest FCS , indicating that the limit of adipocyte expandability may have been reached . Identification of pathways involved in regulation of early adaptation to excess lipid load is important , since these pathways might provide clues about prediction , prevention , or treatment of obesity-related metabolic complications . We used the network approach , rather than a simpler gene expression level analysis , to determine critical nodes that could control adipose tissue lipidome remodeling . Doing so , we detected Elovl6 , which , while not changed on an mRNA expression level itself , appeared to be connected to multiple different factors involved in controlling adipose tissue lipidome ( Figure 4 ) . To validate that perturbations in Elovl6 function were capable of mediating at least some of the changes observed in the obese twins , we used a cell line to validate our network . Consistent with our hypothesis , knockdown of Elovl6 in 3T3-L1 adipocytes mirrored the results found in the obese twins , therefore suggesting that the observed lipid remodeling may be amenable to genetic modulation . As a potential limitation of the study , lipidomic analysis was performed in adipose tissue biopsies and not in isolated adipocytes . It is known that obesity is associated with the infiltration of inflammatory cells in the adipose tissue [39] , [40] . We have in fact observed elevation of multiple inflammatory and immune pathways in the obese twins in our sample [12] . However , adipocytes and their lipids constitute most of the adipose tissue pool per unit volume , and it is therefore unlikely that the observed large change in phospholipid composition is due to changes in percent cellular composition of the adipose tissue . Furthermore , if that were the case , the lipidome changes observed in morbidly obese subjects ( Figure 6 ) would have followed the same trend , but to an even greater extent , as observed in the obese twins . One of the advantages of the twin study setting in the context of understanding human obesity is the possibility to explore the initial effects of acquired obesity and related complications independent of genetic makeup . The weight differences between the twins in the discordant pairs began to emerge at 18 y of age [41] , and diet appears to be the major factor behind these differences [42] . The discordant pairs were of the same age ( mean ± standard deviation = 25 . 6±1 . 2 y ) as the concordant pairs ( 25 . 7±1 . 2 y ) ; therefore , age cannot be considered a confounding factor in our study . The fact that it was extremely challenging to find even this number of MZ twin pairs discordant for weight ( only 14 pairs out of the cohort of 2 , 453 twins born in Finland in the years 1975–1979 , among which adipose tissue was available from 13 pairs ) and the fact that the obesity phenotypes found were within a narrow range , suggest a strong genetic and early environmental basis for susceptibility to obesity . Therefore , our study design offers considerable advantages over other types of studies in humans where groups with different obesity phenotypes also differ for their genotypes . In summary , we have used a novel approach to study cellular lipidomes , and our study proposes an allostatic mechanism by which normal membrane function is maintained in the expanding adipose tissue at the expense of increasing its vulnerability to inflammation . The lipid remodeling seems to be triggered by the proportional decrease of PUFA content and is controlled by a complex network involving fatty acid desaturation and elongation . If small molecules could modulate this network for both membrane functional maintenance and vulnerability to inflammation , new opportunities may arise for the prevention or treatment of obesity-related metabolic complications .
Twin pairs included in the current study were recruited from a population-based longitudinal study of five consecutive birth cohorts in Finland ( 1975–1979 ) of twins ( n = 2 , 453 ) [43] , based on their responses to questions on weight and height at age 23–27 y . From this cohort , we searched for the top 5% most obesity-discordant MZ twin pairs ( one twin non-obese [BMI ∼25 kg/m2] and the other one obese [BMI ∼30 kg/m2] ) , with no significant height differences ( <3 cm ) . After screening all MZ twin pairs ( n = 658 ) , we identified 18 pairs above the 95th percentile of BMI differences ( 3 . 1 kg/m2 ) [10] , [41] , [44]–[47] . Fourteen of these pairs ( eight male and six female pairs ) were willing to participate , and thirteen pairs ( eight male and five female pairs; BMI differences 3 . 3–10 . 0 kg/m2 ) had adipose tissue samples available for the present study . We also studied nine randomly selected weight-concordant MZ pairs ( two male and three female overweight pairs and two male and two female normal-weight pairs , BMI differences 0 . 0–2 . 3 kg/m2 ) . All pairs were Caucasian , and the discordant pairs were of the same age ( mean ± standard deviation 25 . 6±1 . 2 y ) as the concordant pairs ( 25 . 7±1 . 2 y ) . The subjects were healthy ( based on medical history , clinical examination , and structured psychiatric interview ) , normotensive , and did not use any medications except contraceptives . Their weight had been stable for at least 3 mo prior to the study . Females were scheduled to attend during the follicular phase of their menstrual cycle . Monozygosity was confirmed by the genotyping of ten informative genetic markers [41] . Dietary information was collected based on 3-d food diary as described previously [10] . Further details on experimental methods are provided in Text S1 . The subjects provided written informed consent . A small amount of subcutaneous fat tissue aspirated by needle biopsy was used for the metabolic studies of adipose tissue , as described in the consent form . The protocol was designed and performed according to the principles of the Helsinki Declaration and was approved by the Ethical Committee of the Helsinki University Central Hospital . 3T3-L1 cells were cultured , and differentiated into adipocytes using the following protocol . 3T3-L1 preadipocytes were passaged in 25 mM glucose DMEM supplemented with 1% Pen-Strep ( Sigma Aldrich P0781 ) and 1% glutamine ( Sigma Aldrich G7513 ) . Differentiation cells were plated into six-well culture dishes and allowed to grow to confluence . Two days post confluence 3T3-L1 preadipocytes were induced to become adipocytes . The induction medium was 25 mM DMEM with 10% fetal bovine serum ( Sigma Aldrich ) , 1% L-glutamine , and 1% Pen-Strep ( FBS medium ) supplemented with 100 nM insulin , 1 µM dexamethasone , and 0 . 5 mM IBMX . At day 2 the medium was replaced with FBS medium with insulin alone . From day 4 until day 8 cells were grown in FBS medium alone until adipogenesis was complete . Differentiated cells were used only when at least 95% of the cells showed an adipocyte phenotype by accumulation of lipid droplets by day 8 . Cells were analyzed by phase contrast microscopy . Elovl6 knockdown cells ( or controls ) were generated with the pSiren-RetroQ retroviral vector system . All retrovirally transfected 3T3-L1 cell lines were kept in puromycin-containing medium throughout culture and differentiation procedures . Stable knockdown of Elovl6 was achieved by expression of short hairpin RNA from the pSiren-RetroQ vector ( Clontech ) . Target sequences for knockdown of Elovl6 ( GenBank Accession number NM_130450 ) were identified using the Dharmacon siDesign center , and control cells were infected with virus for a scrambled short hairpin RNA . Custom oligonucleotides ( sequences available on request ) were designed to incorporate these sequences into short hairpin RNA expression sequence , and cloned into pSiren-RetroQ according to the manufacturer's instructions . The knockdown was confirmed at mRNA and enzyme activity levels ( Figure S7 ) using the methods described in Text S2 . Global lipidomic profiles of adipose tissue biopsies were determined by using a UPLC-MS platform [48] . Data processing was performed using the MZmine software [49] , [50] . Slightly different analytical methods were applied for the tissue biopsies and 3T3-T1 adipocytes ( Text S3 ) . Serum-free fatty acids as well as adipose tissue esterified fatty acids were measured by gas chromatography . Free cholesterol in adipose tissue biopsies was determined using GC-MS . Further details on lipidomic analysis are provided in Text S3 . Statistical analyses were performed using the freely available R statistical software ( http://www . r-project . org ) . FDR q-values [51] were computed using statistical methods from R package “qvalue . ” Chemometric modeling using partial least squares [52] regression was performed using Matlab version 7 . 0 ( Mathworks ) and PLS Toolbox version 4 . 0 of the Matlab package ( Eigenvector Research ) . The model-based clustering was performed using the MCLUST method [53] , implemented in R ( Text S4 ) . Construction of the adipose tissue network for selected variables was performed using undirected Gaussian graphical Markov networks that represent q-order partial correlations between variables , implemented in the R package “qpgraph” [23] that forms part of the Bioconductor project ( http://www . bioconductor . org ) . In these networks , missing edges denote zero partial correlations between pairs of variables , and thus imply the conditional independence relationships in the Gaussian case ( Text S4 ) . The network was visualized using Cytoscape [54] and yED graphical editor [55] . All simulated bilayer systems consisted of 128 lipid molecules and about 3 , 500 water molecules . The initial structures of all bilayers were obtained by modification of a dipalmitoylphosphatidylcholine ( DLPC ) bilayer simulated for 130 ns as described in our previous study [56] . All the simulations were performed using GROMACS software package version 4 . 0 . 4 [57] over a time scale of 100 ns . The first 40 ns were considered an equilibration period , and the remaining period of 60 ns of each trajectory was analyzed . Further detail on experimental methods is provided in Text S4 .
|
Obesity is characterized by excess body fat , which is predominantly stored in the adipose tissue . When adipose tissue expands too much it stops storing lipid appropriately . The excess lipid accumulates in organs such as muscle , liver , and pancreas , causing metabolic disease . In this study , we aim to identify factors that cause adipose tissue to malfunction when it reaches its limit of expansion . We performed lipidomic analyses of human adipose tissue in twin pairs discordant for obesity—that is , one of the twins was lean and one was obese—but still metabolically healthy . We identified multiple changes in membrane phospholipids . Using computer modeling , we show that “lean” and “obese” membrane lipid compositions have the same physical properties despite their different compositions . We hypothesize that this represents allostasis—changes in lipid membrane composition in obesity occur to protect the physical properties of the membranes . However , protective changes cannot occur without a cost , and accordingly we demonstrate that switching to the “obese” lipid composition is associated with higher levels of adipose tissue inflammation . In a separate group of metabolically unhealthy obese individuals we investigated how the processes that regulate the “lean” and “obese” lipid profiles are changed . To determine how these lipid membrane changes are regulated we constructed an in silico network model that identified key control points and potential molecular players . We validated this network by performing genetic manipulations in cell models . Therapeutic targeting of this network may open new opportunities for the prevention or treatment of obesity-related metabolic complications .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biophysics/theory",
"and",
"simulation",
"cell",
"biology/membranes",
"and",
"sorting",
"diabetes",
"and",
"endocrinology/obesity",
"computational",
"biology/metabolic",
"networks",
"computational",
"biology/molecular",
"dynamics",
"computational",
"biology/systems",
"biology"
] |
2011
|
Association of Lipidome Remodeling in the Adipocyte Membrane with
Acquired Obesity in Humans
|
Pseudomonas aeruginosa strain PA14 is an opportunistic human pathogen capable of infecting a wide range of organisms including the nematode Caenorhabditis elegans . We used a non-redundant transposon mutant library consisting of 5 , 850 clones corresponding to 75% of the total and approximately 80% of the non-essential PA14 ORFs to carry out a genome-wide screen for attenuation of PA14 virulence in C . elegans . We defined a functionally diverse 180 mutant set ( representing 170 unique genes ) necessary for normal levels of virulence that included both known and novel virulence factors . Seven previously uncharacterized virulence genes ( ABC transporters PchH and PchI , aminopeptidase PepP , ATPase/molecular chaperone ClpA , cold shock domain protein PA0456 , putative enoyl-CoA hydratase/isomerase PA0745 , and putative transcriptional regulator PA14_27700 ) were characterized with respect to pigment production and motility and all but one of these mutants exhibited pleiotropic defects in addition to their avirulent phenotype . We examined the collection of genes required for normal levels of PA14 virulence with respect to occurrence in P . aeruginosa strain-specific genomic regions , location on putative and known genomic islands , and phylogenetic distribution across prokaryotes . Genes predominantly contributing to virulence in C . elegans showed neither a bias for strain-specific regions of the P . aeruginosa genome nor for putatively horizontally transferred genomic islands . Instead , within the collection of virulence-related PA14 genes , there was an overrepresentation of genes with a broad phylogenetic distribution that also occur with high frequency in many prokaryotic clades , suggesting that in aggregate the genes required for PA14 virulence in C . elegans are biased towards evolutionarily conserved genes .
Pseudomonas aeruginosa , an opportunistic Gram-negative human pathogen , is one of the leading causes of hospital-acquired infections . In the context of a breakdown in host defenses , it is capable of infecting a plethora of tissue types , causing both acute and chronic infections . Burn victims as well as immunocompromised , mechanically ventilated , and cystic fibrosis ( CF ) patients are particularly susceptible to P . aeruginosa infection [1] . Over the last few decades , a steady increase in drug resistant P . aeruginosa strains has made antibiotic treatment more difficult [2] . In part because no new antibiotics effective against P . aeruginosa are imminently available as treatment options , the pressing need for drugs to fight this pathogen has focused study on its virulence factors as potential drug targets , and more generally energized a search for novel anti-infectives [3]–[5] . One likely reason that P . aeruginosa is a common nosocomial pathogen is because it is capable of thriving in a wide variety of environmental niches , including surfaces in hospital rooms , water , soil and plants [6] . Consistent with its ecological ubiquity , P . aeruginosa has a relatively large genome that presumably promotes survival in diverse habitats . In addition to inhabiting a wide variety of ecological niches , P . aeruginosa is also a multi-host pathogen , capable of infecting hosts as divergent as amoebae , plants , insects , flies , nematodes , and mice [7]–[13] . Progress in fighting P . aeruginosa infections will be aided by a fundamental understanding of the myriad ways that P . aeruginosa can survive in different environments and cause disease in diverse hosts . Our laboratory has developed a Pseudomonas aeruginosa - Caenorhabditis elegans infection-based model for studying host-pathogen interactions that is genetically tractable from both the perspectives of the host and the pathogen . This model ( referred to as “slow-killing” or SK ) , which primarily utilizes P . aeruginosa strain PA14 [8] , requires live bacteria and a set of bacterial virulence factors that distinguish it from rapid toxin-mediated PA14 killing of C . elegans ( “fast killing” or FK ) that occurs on high osmolarity media [12] , [14] . Under standard laboratory conditions [15] , C . elegans animals have a normal lifespan of approximately two weeks when feeding on non-pathogenic Escherichia coli strain OP50 , and OP50 does not accumulate in the C . elegans intestine during the first few days of life . In contrast , when C . elegans are transferred at the L4 larval stage from E . coli OP50 to a lawn of P . aeruginosa strain PA14 , the animals die in two-three days [12] . A few PA14 cells initially accumulate in the anterior and posterior portions of the nematode intestine , then over the course of one to two days bacteria spread throughout the intestine and the intestinal lumen becomes severely distended with a corresponding reduction in volume of the intestinal epithelial cells . Ultimately , PA14 cells move across the intestinal epithelial barrier destroying tissue , but it is not known whether tissue invasion is required for killing [16] . C . elegans rapidly responds to the presence of pathogenic PA14 by enhancing the transcription of hundreds of genes including a number of predicted secreted proteins ( C-type lectins , CUB-domain containing proteins , ShK toxins ) that may have antimicrobial or detoxifying activity [17] , [18] . Two of the major classes of PA14 response genes , C-type lectins and CUB-like domain containing proteins , also play a role in mammalian innate immunity [19]–[21] . In C . elegans , many immune response genes are regulated by the PMK-1 p38 mitogen-activated protein kinase ( MAPK ) , the terminal MAPK in an evolutionary conserved signaling cassette required for defense against pathogens in both nematodes and mammals . Approximately 25% of the C . elegans genes regulated by PMK-1 are also induced in response to P . aeruginosa PA14 [18] and C . elegans p38 MAPK pathway mutants exhibit enhanced sensitivity to PA14 as well as a variety of other bacterial and fungal pathogens [22]–[24] . Several hundred genes have been implicated in P . aeruginosa virulence based on data obtained from a wide variety of host infection models ( Pseudomonas . com ) . Many of the well-studied P . aeruginosa virulence-related factors participate directly or indirectly in physical interactions with the host cell and/or host proteins , including secretion systems ( type II , type III , type VI ) and associated effectors ( including ExoT , ExoU , ExoS , ToxA , phospholipase C , and alkaline protease ) , flagella , and structures involved in attachment to host cells such as type IV pili . Other recognized virulence factors include those involved in quorum sensing ( AHL and PQS systems ) , iron acquisition ( pyochelin , pyoverdine ) , small molecule/toxin synthesis ( phenazines , hydrogen cyanide ) , alginate , LPS , and biofilm . Not all of these classes of virulence-related factors play a significant role in P . aeruginosa strain PA14 virulence in C . elegans; for example , the Type III secretion system and its associated virulence effectors have been shown to play no detectable role in nematode killing , in contrast to playing key roles in pathogenesis in mammals and insects [25] . However , a variety of P . aeruginosa PA14 virulence factors required for killing C . elegans in the SK infection model are also required for full pathogenesis in mammalian models , including the quorum sensing regulators LasR and RhlR , the two component regulator GacA , the alternate sigma factor RpoN , the periplasmic protease MucD , and the phosphoenolpyruvate-protein phosphotransferase and transcriptional regulator PtsP [13] , [26]–[29] . Additionally P . aeruginosa virulence-related factors involved in LPS biogenesis and type IV pilus assembly and function also play a role in both mammalian and C . elegans hosts [30]–[33] . A common theme that has emerged from the study of bacterial virulence in a wide variety of pathogens and hosts is an association linking virulence-related genes with regions of genomic plasticity , including genomic pathogenicity islands ( PAIs ) [34] , so-called “replacement islands” harboring the pyoverdine [35] and O-antigen biosynthetic loci [36] , and plasmids [37] . These findings indicate that horizontal gene transfer has played an important role in the evolution of virulence . For example , phylogenetic analysis of three sub-families of the type III effector HopZ in the plant pathogen Pseudomonas syringae , suggested that at least two were acquired by P . syringae from disparate donors [38] . Analysis of the occurrence of virulence factors across many pathogen genomes has suggested that there is an overrepresentation of virulence factors on genomic islands [39] , and two virulence factor- containing pathogenicity islands , PAPI-I and PAPI-II , have been identified in P . aeruginosa [40] . Although there are many published examples linking virulence-related factors to putative pathogenicity islands , a preliminary study from our laboratory showed that the presence of genes occurring in the highly virulent strain PA14 , but not in the less virulent strain PAO1 , could not be correlated with increased virulence across a wider sampling of strains , suggesting that virulence is a combinatorial and multifactorial product of the interactions of many potential virulence factors [32] . These data were seemingly at odds with the expected over-representation of virulence factors in strain-specific regions such as genomic islands , but were not definitive because only a limited set of virulence factors were available for analysis . Further , comparison of the sequences of five pathogenic P . aeruginosa strains suggested that virulence was primarily encoded by a core P . aeruginosa genome [41] , a set of genes shared by all strains , and not the auxiliary genome defined by regions of genomic plasticity that are strain-specific . An unbiased comprehensive list of P . aeruginosa virulence factors required to cause disease in C . elegans would allow us to better understand what genes are the major contributors to virulence and whether these genes are primarily located in regions of genome plasticity or not . We considered this question worthy of investigation because it seemed likely to us that the virulence factors of an opportunistic multi-host pathogen might as a group be distinct from the virulence factors of host-specific pathogens . We report here the results of a genome-wide screen using a previously constructed non-redundant PA14 transposon library consisting of 5850 members that represents insertions in approximately 80% of the non-essential ORFs in P . aeruginosa strain PA14 [42] . Previous studies to identify P . aeruginosa virulence factors in vivo using a number of different technologies and infection models have been limited by the complexity and redundancy of mutant collections or screening procedures [13] , [26] , [27] , [43]–[51] . We examined the genes identified in this genome-wide screen for their functions , presence on putative and characterized genomic islands , and their phylogenetic distribution across prokaryotes . We demonstrate that the major genes contributing to PA14 virulence in C . elegans are not enriched on genomic islands , are not PA14 or P . aeruginosa specific genes , and may in fact be biased for ancient genes common to many other prokaryotic species .
Seven putative virulence-related factors , cold shock domain protein PA0456 , ABC transporters PchH and PchI , aminopeptidase PepP , putative enoyl-CoA hydratase/isomerase PA0745 , ATPase/molecular chaperone ClpA , and putative transcriptional regulator PA14_27700 were chosen for further study . Mutants corresponding to these factors ( clpA , Figure 4B; pchH and pchI , Figure 6B; pepP , Figure S5B; PA0456 , Figure S6B; PA14_27700 , Figure S8B; and PA0745 , Figure S9B , C ) all have a strong avirulent phenotype in C . elegans , exhibit normal growth kinetics in vitro ( Figure S12 ) , and represent genes whose role in P . aeruginosa virulence has not been previously characterized . The avirulent phenotype of all these mutants was confirmed with multiple transposon alleles except for pchH for which there is only a single allele available . In addition , in the case of PA0745 , an in-frame deletion mutant was generated that was severely impaired in virulence , similar to the transposon allele #37629 isolated in the screen ( Figure S9C ) . Many of the genes previously identified as necessary for virulence of PA14 in C . elegans , for example those coding for the quorum sensing regulators RhlR and LasR , are known regulators of multiple virulence factors or virulence associated pathways [58] , [59] . Both lasR and rhlR mutants have a spectrum of pigment and motility defects . lasR and rhlR mutants produce reduced levels of the blue-green pigment pyocyanin , rhlR produces no pyocyanin , and lasR mutants produce varying amounts dependent on conditions and growth phase [60] . Under certain growth conditions , rhlR and lasR mutants have been reported to produce less of the fluorescent siderophore pyoverdine [61] . lasR and rhlR mutants also exhibit dramatically reduced swarming motility [62] , which is dependent on both the type IV pilus and the flagella and regulated by quorum sensing and a host of transcription factors [63] . We tested the seven selected virulence-related mutants for defects in motility ( twitching , swimming and swarming assays ) as well as for pyocyanin and pyoverdine production in comparison to lasR and rhlR mutants to determine whether they had a similar spectrum of defects and/or could be classified into groups based on common pigment or motility phenotypes ( Table 2 ) . It should be noted that of these phenotypes , only mutants in which type IV pilus function is affected have been shown to exhibit reduced virulence in the C . elegans infection model; pyocyanin does not appear to be necessary for virulence in the SK model and the roles of pyoverdine production , swimming and swarming have not been directly tested [12] , [14] . Significantly , with the exception of PA14_27700 , all of the mutants exhibited defects in some aspect of motility or pigment production . Mutation of putative cold shock protein PA0456 diminished pyocyanin production as did the quorum sensing regulators lasR and rhlR , whereas a pepP mutant had elevated pyocyanin levels . The putative enoyl-CoA hydratase/isomerase PA0745 produced reduced levels of pyoverdine . Among the tested mutants , 4/7 had clear swarming defects , but exhibited normal levels of swimming and twitching motility , suggesting that neither flagella nor type IV pili function was compromised . The clpA mutant was slightly attenuated for swarming and twitching , implying that there might be a type IV pilus defect in this mutant . The list of PA14 genes identified as being required for full virulence in C . elegans from the genome-wide screen provided the opportunity to examine the distribution within a species and conservation across bacterial species of a large set of genes required for virulence in a single host . We determined whether this set of virulence associated genes was biased towards Pseudomonas core or strain-specific ( auxiliary ) regions of the P . aeruginosa genome ( as defined by Mathee et al . [41] and/or whether these virulence genes were preferentially located on genomic islands , as previously suggested for Pseudomonas virulence factors [39] , [40] . In addition , we examined whether the PA14 virulence genes had a narrow phylogenetic distribution ( unique to PA14 , P . aeruginosa , Pseudomonas , or closely-related organisms ) or were broadly distributed across prokaryotic phylogeny . We used four sets of genes identified in our screen and a set of previously defined PA14 virulence genes downloaded from the Virulence Factor Database ( VFDB ) for all analyses . The sets of unique genes identified in the primary ( 294 ) , secondary ( 170 ) and tertiary ( 41 ) virulence-attenuated screens outlined above and the auxotrophic genes identified in the primary screen but subsequently discarded ( 76 ) were used and for simplicity are referred to below as primary , secondary , tertiary , and auxotroph sets . All statistical analyses of the virulence genes identified in the C . elegans screen were done in comparison to the genes represented in the non-redundant ( NR ) library , as opposed to the entire PA14 genome , because this was the starting set for the screen . We used all four sets of genes in our comparisons to gain statistical power because the final set of 41 verified virulence-related genes was so small that most analyses did not make statistical cutoffs . In addition , using sets of genes from subsequent rounds of screening allowed us to look for enrichment that correlated with the refinement of the screen . To compare the virulence-related genes identified in our screen to previously identified P . aeruginosa genes , we made use of a set of 241 P . aeruginosa strain PA14 virulence genes downloaded from the Virulence Factor Database ( VFDB ) [52] .
We set out to define the spectrum of genes required for P . aeruginosa PA14 infection in a single host organism with the ultimate goal of elucidating the mechanisms underlying pathogenesis in this multi-host opportunistic pathogen . A genome-wide unbiased screen for P . aeruginosa strain PA14 mutants defective in killing C . elegans identified a set of 180 putative virulence-related mutants ( corresponding to 170 genes ) after two rounds of screening . The screen was validated by the isolation of mutants previously shown to be required for P . aeruginosa virulence in both nematodes and mammals or known to regulate processes or pathways linked to pathogenesis including , but not limited to , genes involved in quorum sensing , two component regulators of virulence , transcriptional regulators , genes involved in type IV pilus production , and O-antigen biosynthesis . Twenty genes in the 170 gene set overlapped with a set of previously defined virulence factors in VFDB , a database of Pseudomonas-related virulence factors . Overall , the PA14 genes identified in the genome-wide screen have an overrepresentation of highly conserved genes present in many bacterial phyla and are part of the stable P . aeruginosa genome , rather than being located on pathogenicity islands . The set of 170 virulence-related genes is broadly distributed across 27 defined functional classes and DAVID GO term analysis and mapping onto KEGG pathways did not reveal any interpretable enrichment for particular functions or pathways . This breadth of functional classes parallels the virulence-attenuated mutants identified in an independent unbiased screen using signature tagged mutagenesis carried out by Potvin and coworkers in a rat chronic infection model [46] . Unlike the genes identified in our screen and by Potvin et al . , virulence-related factors in the VFDB are enriched for secretion- and adherence-related proteins . A major difference between the virulence-related genes in VFDB and the genes identified in our unbiased screen is that many of the genes in VFDB were included because they encode secreted toxins , secretion systems , or cell surface structures [52] , [67] . However , the sensitivity of our screen favored identification of mutants with strongly attenuated virulence . This was expected given the nature of the primary screen that required that both the parent nematodes live long enough to produce a significant brood and that the nematode brood mature on the mutant bacterial lawn . It is possible , therefore that mutants with a weak virulence-attenuated phenotype were not detected and this could potentially skew the collective analysis of virulence factors . One explanation for why so few mutants identified in our screen correspond to secretion pathways or to secreted effectors is that many of P . aeruginosa virulence effectors appear to function redundantly in the C . elegans killing assay . In support of this conclusion , disruption of the ExoU cytotoxic phospholipase had no statistically significant impact on virulence , but appeared to create a sensitized background that allowed the detection of other relatively weak virulence factors . Further , McEwan et al . have shown that whereas PA14 exotoxin A ( toxA ) mutants have no or little defect in virulence , overexpression of ToxA in E . coli activates the worm immune system and ultimately kills an immune-compromised animal , suggesting that ToxA may play an active , but to date undetected , role in PA14 pathogenesis in the nematode host [68] . Similarly , Dunbar et al . [69] have shown that ToxA inhibits protein synthesis in C . elegans intestinal cells during an infection . ExoU and ToxA are secreted by distinct systems and the weak virulence attenuation of secretion system mutants in C . elegans implies that multiple secretion systems and effectors may contribute to virulence in C . elegans with no single system being paramount . An alternative explanation for the identification of a limited number of secretion-related mutants in our screen may be linked to the fitness costs of maintaining a large set of effectors targeting a wide range of potential hosts . Only four Type III effectors have been identified in P . aeruginosa , whereas 46 families of effector proteins have been identified in various strains of the related plant pathogen P . syringae [70] and a typical P . syringae strain has 20 to 30 effectors [71] . In P . syringae , which has a much more limited host range than P . aeruginosa , type III effectors mostly target host defense signaling pathways and both enhance virulence in particular host plants , while eliciting a strong immune response in others [72] . This fact , combined with the observations that the genes encoding secreted effectors are often under diversifying selection [38] , [73] and are typically located in plastic regions of the genome [70] , [74] , suggests that P . syringae strains actively co-evolve with a limited number of hosts . In contrast , from first principles , it seems highly unlikely that a broad host-range pathogen like P . aeruginosa PA14 can be simultaneously co-evolving with all of its multiple hosts . Therefore P . aeruginosa might employ a broader set of strategies to ensure survival in diverse hosts instead of maintaining large sets of host-specific virulence-related effectors . In this context , a number of the strongly virulence-attenuated PA14 mutants identified in our screen may correspond to factors that enable survival of P . aeruginosa in the hostile environment of the C . elegans intestinal tract , which is acidic and filled with enzymes such as proteases , lipases , and DNAse that potentially disrupt bacteria [75] . Moreover , in response to pathogens , the nematode specifically upregulates transcription of many putative antimicrobial genes [18] . In order for P . aeruginosa to proliferate in the intestine and cause disease , it first has to survive . Both the two component potassium sensor KpdD and the nitrogen assimilation regulatory protein GlnK , identified in our screen , have recently been shown to play a role in the persistence of S . typhimurium in the C . elegans intestine and defects in outer membrane integrity may reduce the survival of kdpD and glnK mutants in the host [76] . In addition , the identification of two PA14 genes required for biosynthesis of glutathione , gshA and gshB , may be related to the role of glutathione as a protectant against stresses encountered in the worm intestine including reactive oxygen species and low pH [77] . Cold shock domain proteins , like PA0465 identified in our screen are another class of molecule that are induced by environmental stress and are generally thought to play a protective role in the cell [78] . Identification of a number of PA14 mutants corresponding to metabolic genes illustrates the importance of nutrient acquisition in virulence . Without specific biosynthetic or metabolic capabilities a pathogen may be unable to colonize or grow within a host . For example , P . aeruginosa mutants defective in purine biosynthesis are unable to replicate in neutropenic mice , presumably because the in vivo environment is deficient in purine [48] . In the cases of intercellular pathogens such as Listeria monocytogenes and Mycobacterium tuberculosis , specific amino acid and nucleotide auxotrophs are reduced in growth in vivo [79] , [80] . We identified a number of metabolic genes including prpB and prpC and several aru genes that may be important for bacterial metabolism and growth under the nutrient conditions within the nematode intestine . Further , some of the putative virulence-attenuated mutants identified in the primary screen that were set aside for further study because they were determined to be auxotrophs ( a typical step in many screens ) might specifically reflect nutrient availability in the nematode intestine . In this regard it is notable that mutations in nine purine , five pyrimidine , and six tryptophan biosynthetic genes were identified in the primary screen for virulence-attenuated mutants , and although some of these mutants exhibited reduced growth on the killing assay medium , many did not have any observable difference from wild-type PA14 , suggesting that the reduction in virulence might be due to aberrant growth of these auxotrophs in vivo . The predominant contributors to PA14 virulence in our C . elegans infection based assay appear not to be individual effectors , but genes that regulate numerous effectors ( like the quorum sensing regulators lasR and rhlR ) , genes that are vital for protecting the bacteria from the host defense onslaught , and genes that help P . aeruginosa obtain the necessary nutrients to survive in the host . Therefore the strategy of a broad host range opportunistic pathogen might fundamentally differ from a pathogen that targets specific hosts , relying more on multiple partially redundant secretion systems and their cognate effectors and strategies for survival under a wide-variety of metabolic and environmental conditions . The long-recognized association between virulence genes and regions of genomic plasticity , particularly genomic islands acquired by lateral transfer of genetic material [81] , has been attributed to the competitive advantage conferred by horizontally-acquired virulence factors in an ongoing co-evolutionary struggle between a host and pathogen . Two pathogenicity islands carrying plant and animal virulence-related genes have been identified and characterized in P . aeruginosa PA14 ( PAPI-1 and PAPI-2 ) . Of the 11 genes located on PAPI-1 previously shown to be required for normal levels of virulence in plants and mice [40] , only one ( rcsC ) was identified in our secondary screen as a weak mutant and it did not re-test in the tertiary screen . Overall our results showed no enrichment of virulence-associated genes on predicted genomic islands , or on the known genomic islands PAPI-1 , PAPI-2 , or PAGI-1 . More generally , whether or not virulence genes in aggregate are preponderantly associated with genomic islands in P . aeruginosa has not been experimentally demonstrated . The statistical power of analysis with respect to the genomic locations of genes is limited in part by the fact that not all genes are expected to segregate independently . In this regard , it is worth noting that the apparent enrichment of P . aeruginosa VFDB genes on predicted genomic islands is primarily due to the cluster of functionally interdependent Type III secretion apparatus genes located between gene loci PA14_42440 and PA14_42660 . In summary , our data and analysis of existing data suggest that neither P . aeruginosa VFDB genes nor the virulence-attenuated genes identified in our screen are preferentially found on genomic islands . Roughly paralleling our observations with genomic islands , our analysis of the frequency of PA14 virulence genes in the core and auxiliary genomes ( both from our screen and the VFDB ) showed no statistically significant over or underrepresentation . The main difference between the core versus auxiliary genome distinction , and that of genomic islands , is that auxiliary genes include both genes that are lost from the genomes of some isolates , as well as those genes that are newly acquired . Genomic islands by contrast , specifically include only newly acquired genes . Taken together , these results suggest that there is no specific enrichment of P . aeruginosa virulence-related genes on islands or in strain-specific regions of the P . aeruginosa genome , at least with respect to those that are involved in the C . elegans slow killing assay . To further test the hypothesis that the arsenal of P . aeruginosa virulence factors includes newly evolved and or newly acquired genes we investigated the phylogenetic breadth of distribution of the PA14 virulence genes , as well as the degree to which they belong to a set of putative old conserved genes , the so-called high-frequency-broad-phylogeny ( HFBP ) set , and probable newer genes , the Pseudomonas-genus-specific ( PGS ) set . We found that consistent with the result that virulence genes are not located on islands , that PA14 virulence genes involved in C . elegans killing are enriched in HFBP genes , which as a group are likely to be the most ancient and conserved prokaryotic genes . The vast majority of PA14 virulence genes appear not to be specific to the Pseudomonas genus , and in fact , such recently acquired or novel genes are underrepresented among our putative virulence genes . These observations underscore the point that many virulence factors with the most significant contribution to virulence can be old , highly conserved genes . Although the C . elegans model , in which nematodes are “force-fed” a monoculture of P . aeruginosa [13] , [43] , [44] , is somewhat artificial , so are other laboratory models of P . aeruginosa infection that require pricking of the body of a fly [45] , lung inoculation with agarose beads in rats [46] and mice [47] , and non-lethal cutaneous burns in mice [82] . The artificiality of these models and the need for a compromised host is in part dictated by the opportunistic nature of the pathogen . The fact that conserved immune defense pathways are activated in the C . elegans host by P . aeruginosa strongly supports the view that the nematode is responding to P . aeruginosa as a pathogen [83] . Although we don't yet know whether P . aeruginosa is a natural pathogen of C . elegans , since both organisms live in the soil , it is likely that they encounter one another , and it is not inconceivable that the pathogenicity interactions that we observe may approximate the interactions of P . aeruginosa with weakened individual C . elegans animals in the wild . Although some putative P . aeruginosa virulence factors are common to both mammalian and C . elegans host models , the degree to which P . aeruginosa virulence factors are shared , and implicitly , the degree to which P . aeruginosa virulence strategies are common to C . elegans and mammalian hosts is not yet clear . That only a single putative virulence-related gene is shared between the 170 virulence-attenuated genes from our secondary set , the rat chronic infection set defined by Potvin et al . , and the VFDB set , argues against the idea of a core set of virulence factors common to all infection models . The spectrum of virulence factors that play a role in a given host model is likely to depend on a wide variety of factors including the characteristics of the site of infection , such as pH , ionic strength , nutrient availability , and temperature , the type of immune compromise , the phase of infection , and the particulars of the immune response , including the presence of host factors , and even host behavior . Indeed , two additional PA14 pathogenicity assays in C . elegans have been developed in our laboratory , a toxin-mediated killing model [12] , [14] and a liquid killing assay ( N . Kirienko and F . Ausubel , unpublished ) . PA14 appears to utilize distinct mostly non-overlapping sets of virulence-related genes to kill nematodes in the three different models . Nevertheless , all three of these assays have identified virulence factors that play important roles in various aspects of mammalian pathogenesis . The implication for human pathology is that the predominant virulence factors that play a role in different types of P . aeruginosa infection in humans may be somewhat distinct . In summary , the data from our unbiased genome-wide screen for P . aeruginosa virulence factors involved in C . elegans killing in a specific infection model , suggest that in comparison to host-specific pathogens , P . aeruginosa may employ a smaller arsenal of host-specific effectors , and rely more on conserved , generic virulence factors and on its ability to endure host defense responses . While it is not yet clear that this same strategy is employed by multi-host pathogens beyond P . aeruginosa that are capable of infecting organisms from multiple phylogenetic kingdoms , this may explain why the major genes contributing to PA14 virulence in C . elegans are not overrepresented on genomic islands , are not PA14 or P . aeruginosa specific genes , and may in fact be biased for ancient genes common to many other prokaryotic species . These observations are consistent with the view of P . aeruginosa PA14 as a generalist pathogen for which the relationship with C . elegans is opportunistic rather than co-evolved . Indeed it is likely that no significant co-evolution occurs between P . aeruginosa and any of the hosts for which it is an opportunistic pathogen , both because of the rarity of pathogenic interaction , and because of the likelihood that co-evolution with multiple hosts would necessitate balancing opposing evolutionary pressures from those hosts . From a clinical perspective , the multiplicity and apparent combinatorial nature of P . aeruginosa's virulence factors may pose a challenge for the development of new therapeutics to fight Pseudomonas infection . This work does not settle the question of whether the profile of virulence factors of a multi-host pathogen is likely to be different from that of host-specific pathogens in terms of the reliance on conserved effectors that target highly conserved features of eukaryotic biology , but it is a question that deserves further inquiry .
P . aeruginosa strain PA14 mutants are gentamycin resistant MAR2xT7 transposon insertion mutants unless otherwise stated [42] . The lasR mutant used as a control was a deletion that removes the lasR ATG and carries a gentamycin cassette [84] . Similarly , the rhlR mutant used as a control in the pigment and motility assays was a deletion mutant that carries a gentamycin cassette [84] . These mutants have an identical avirulent phenotype to subsequently generated clean in-frame PA14 ΔlasR and ΔrhlR mutants ( data not shown ) . The pilA mutant is a tetracycline resistant Tn5-B30 transposon insertion mutant provided by G . O'Toole [85] . ΔPA0745 is a complete in-frame deletion of the ORF with a concomitant insertion of a PacI restriction site generated by the method previously described in Chand et . al . [86] . Unless otherwise indicated , MAR2xT7 mutant bacteria were streaked from frozen stocks onto LB agar containing 15 µg/ml gentamycin to isolate single colonies used for inoculation . WT PA14 ( as well as lasR and pilA mutants ) was streaked on LB agar containing 100 µg/ml rifampicin . N2 Bristol L4 animals were used for the primary and secondary screens . Tertiary tests were done with CF512 fer-15 ( b26 ) II;fem-1 ( hc17 ) IV temperature sensitive sterile nematodes . CF512 worms were propagated at 15°C , egg-prepped and hatched overnight at 20°C in M9 liquid media in the absence of food . Starved L1 animals were dropped onto NGM plates seeded with standard E . coli OP50 and grown for approximately 20 hours at 15°C and then 20 more hours at 25°C to generate staged L4 sterile animals for experimentation . We used the non-redundant PA14 MAR2xT7 transposon library of ordered mutants to screen for avirulent mutants [42] . The entire library was screened twice with the exception of a single 96-well plate of mutants ( plate 14 . 3 ) that was not screened because it contained mostly known slow-growing mutants . Bacteria were inoculated from the frozen 96-well stock plates into 150 µl of LB in 0 . 5 ml 96-well Masterblocks ( Greiner ) and grown overnight with shaking for 16 hours at 37°C . 10 µl of each culture was spotted onto slow killing ( SK ) agar ( standard nematode growth media , NGM , containing 0 . 35% instead of 0 . 25% peptone ) in each of two wells of a 6-well culture plate ( Falcon ) [12] . Two previously identified virulence-attenuated mutants , quorum sensing regulator lasR ( highly attenuated ) and typeIV pilin pilA ( moderately attenuated ) were included as positive controls in the screen . Plates were scored qualitatively after 4 days at 25°C and designated as either “strong” , indicating large numbers of gravid animals ( generally on lasR mutant bacteria hundreds of gravid nematodes were present and the bacterial lawn was mostly or completely consumed ) or “weak” , indicating an observable increase in the number of worms or gravid adults present compared to PA14 ( pilA usually resulted in an increase in the total brood size with a bias towards older stage larvae and gravid adults; the pilA control wells exhibited variability in the number of progeny alive and were near the limit of sensitivity in the screen ) . Each mutant was screened blind on two separate days for a total of four wells screened per mutant . Mutants that were scored as strong in either one or both repetitions or weak in both were retained for a total of 399 primary mutants . Approximately half way through the primary screen it became clear that lawn growth phenotype was relevant to our study , and from then on we scored for lawn growth . We therefore have a crude assessment of growth for about 50% of the mutants identified in the primary screen . All 399 of the mutants from the primary screen were streaked from the frozen library stock onto LB agar plates containing 15 µg/ml gentamycin and grown overnight at 37°C . To quickly test for potential auxotrophs , a single colony from each mutant was picked and streaked on 1 ) Neidhardt supplemented MOPS defined media without amino acids or nucleotides ( TEKNOVA EZ Rich Media ) , 2 ) Neidhardt supplemented MOPS defined media plus amino acids and nucleotides , and 3 ) LB containing 15 µg/ml gentamycin agar plates . Plates were incubated overnight at 37°C . 86 mutants that either exhibited no growth on minimal media minus supplements or significantly reduced growth were removed from the pool of putative avirulent mutants and not examined in further rounds of screening ( Table S2 ) . 15/56 of the auxotroph mutants for which a plate growth phenotype had been annotated ( see above ) had obvious growth defects on SK agar plates . The 313 mutants from the primary set that exhibited normal growth on minimal media without addition of amino acids and nucleotides ( Table S3 ) were screened for attenuation of virulence in standard SK assays and scored for both killing of parental ( P0 ) animals and the number and maturity of worm progeny produced [13] . A single bacterial colony was inoculated into 5 ml of LB media and grown for 14 . 5 hours at 37°C with aeration on a rotating wheel . 10 µl of this overnight culture was spread onto each of two SK agar plates ( 3 . 5 cm culture plates ( Falcon ) , plates were incubated for 24 hours at 37°C , and then 24 hours at 25°C . 30–40 N2 L4 animals ( raised on standard NGM plates with E . coli OP50 as food ) were picked to each SK plate and the plates were incubated at 25°C for a total of 60–80 animals per assay . Live and dead animals were counted 2–3 times over 1–3 days after transfer to the pathogen . Animals were considered dead if they did not respond to a gentle touch and were removed from the plate . After 4 days at 25°C , the number and age of the nematode progeny on each plate was qualitatively assessed as compared to those on PA14 WT . Secondary positives were ranked as “strong” ( strongly virulence-attenuated similar to the lasR mutant ) , “moderate” ( less attenuated than lasR but greater than or equal to pilA ) and “weak” ( less attenuated than pilA for parental killing but clearly exhibiting an increase in progeny number and age over WT PA14 ) . 180 MAR2xT7 mutants exhibited either attenuation of P0 parental killing or allowed increased production and development of nematode progeny ( Table S4 ) . The identity of the MAR2xT7 mutants was confirmed by sequencing of arbitrary PCR products as previously described [42] , see Table S4 . 58 MAR2xT7 mutants were re-screened in standard SK assays using sterile fer-15 ( b26 ) I:fem-1 ( hc17 ) IV animals ( Table S5 ) . Bacteria were grown as indicated above for the secondary screen . Three 3 . 5 cm SK agar plates were seeded with each bacterial strain and 35–50 L4 worms were transferred to each plate for a total of 100–150 animals per assay . Live and dead animals were counted every day over approximately 7 days . PA14 WT and lasR were included in each assay as controls . The resulting 41 virulence-attenuated mutants shown in Figure 2 were all tested in at least two separate experiments unless otherwise indicated . Statistical analysis of the curves summarized in Figure 2 was done using Prism 5 Log-rank ( Mantel-Cox ) test and the difference between the mutant and wild-type curves was highly significant in all cases with a p-value<0 . 0001 . Time to 50% survival was calculated ( Prism 5 . 0 linear regression Hill equation LogEC50 ) . All killing assays presented in the manuscript ( Figures 3 , 4 , 5 , 6 , S5 , S6 , S7 , S8 , S9 , S10 , and S11 ) are a representative example of two or more experiments that resulted in the summary shown in Figure 2 . Growth of bacterial mutants was evaluated by four methods: 1 ) Overnight cultures inoculated from a single colony into 5 ml of LB and grown for 14 . 5 hours at 37°C with shaking were visually compared to wild-type PA14 grown under the same conditions . Mutants that were observably less turbid were considered slow growers ( in the case of PA14_45650 mutant #54246 the cells appeared to be lysed ) . 2 ) Slow killing agar plates spread with each mutant were examined prior to transfer of worms . Bacterial lawns that were thin or exhibited other aberrant phenotypes ( nusA mutant #55834 had large colonies that emerged on top of the lawn ) were removed from the mutant pool . 3 ) Many of the secondary and tertiary positive mutants were grown overnight in LB and M63 minimal media in a 96-well plate without shaking and the OD600 was measured every 15 minutes for 15 hours at 37°C in a Molecular Devices Spectra Max M5 . The rate of growth during two hours of maximal growth was compared to WT ( Table S5 ) . 4 ) The growth of nine mutants ( pchH , pchI , PA14_27700 , PA2550 , PA0456 , PA0745 , clpA , PA1216 , vqsR ) in 5 ml of M63 minimal in a standard culture tube on a rotation wheel at 37°C was measured by counting colony forming units . All nine of these mutants grew as well as WT ( Figure S12 ) . A pyocyanin assay was modified from Essar et . al . 1990 [87] . A single colony of WT or mutant bacteria was inoculated into 5 ml of LB media and grown for 16 hours at 37°C on a rotating wheel . 1 ml of saturated overnight culture was transferred to a microfuge tube and cells were pelleted by centrifugation at 14 , 000 RPM for 2 minutes . 800 µl of the supernatant was transferred to a new tube , extracted with 600 µl of chloroform and the phases were separated by centrifugation for 5 minutes at 14 , 000 RPM . The chloroform phase was then re-extracted with 0 . 3 ml of 0 . 2 N hydrochloric acid . The pyocyanin content of 100 µl of the aqueous acidic phase was quantitated based on absorbance at 520 nm . The A520 was normalized to cell number ( A600 of the original overnight culture ) . A ΔphzA1-G1/ΔphzA2-G2 [88] mutant that does not produce any phenazines had no detectable A520 . The pyocyanin produced by each bacterial strain was measured from the growth of two individual colonies on two separate days . For each of the four cultures , the ratio of A520 ( normalized to cell number as measured by A600 ) of mutant to WT was calculated and the average of these four ratios is shown in Table 2 . The error presented is the SEM of the four mutant/WT ratios . A single colony of WT or mutant bacteria was inoculated into 5 ml of M9 media and grown overnight for 18 hours at 37°C on a rotating wheel . Cells were pelleted by centrifugation at 14 , 000 RPM for 2 minutes in a microfuge tube and the supernatant was diluted 10 fold in 10 mM Tris pH 7 . 4 . Pyoverdine content was determined by measurement of fluorescence ( 400 nm excitation , 460 nm emission ) [89] , [90] . No fluorescence above background was detected in pyoverdine biosynthetic mutants , pvdD ( #40342 ) and pvdA ( #30448 ) . The pyoverdine produced by each bacterial strain was measured from the outgrowth of two individual colonies on two separate days . For each of the four cultures , the ratio of fluorescence ( normalized to cell number as measured by A600 ) of mutant to WT was calculated and the average of these four ratios is shown in Table 2 . The error presented is the SEM of the four mutant/WT ratios . The swarming assay was modified from Overhage ( 2008 ) [91] . A single colony of WT or mutant bacteria was inoculated into 5 ml of LB media and grown for 15 hours at 37°C with aeration . 2 µl of each overnight culture was spotted onto the surface of LB 0 . 5% agar and SK 0 . 5% agar plates and then incubated at 37°C overnight . Each mutant was tested in triplicate on two separate days . As expected , the rhlR mutant ( deficient in rhamnolipid production ) and pilA mutant ( typeIV pilin ) were defective in swarming [62] . Mutants were visually compared to WT PA14 and were qualitatively evaluated for swarming radius and number of tendrils ( Table 2 ) . The swimming motility assay was based on Darzins ( 1993 ) [92] . A single bacterial colony was picked with a straight end loop and inoculated into LB swim agar ( 0 . 35% agar ) . Plates were incubated 8–12 hours at 37°C . The diameter of the flagellum-mediated motility generated turbid zone was measured . Each mutant was tested in triplicate and the average with SEM is presented in Table 2 . The twitching motility assay was based on O'Toole 1998 [85] . A portion of a single bacterial colony was picked with a straight end inoculation loop and stabbed to the bottom of a LB agar plate ( 1 . 5% agar ) . Plates were incubated overnight at 37°C and then 2 days at room temperature . The growth at the interface between the agar and the polystyrene plate ( radius from the inoculation point ) was measured . The pilA mutant exhibited no twitching motility . Each mutant was tested in triplicate and the average with SEM is presented in Table 2 . A database of orthologs of P . aeruginosa strain PA14 genes across 727 sequenced prokaryotic genomes ( including PA14 ) was created . Finished microbial genome sequences were obtained as downloaded packages from the NCBI ftp site ( ftp://ftp . ncbi . nlm . nih . gov/genomes/Bacteria/ ) on August 25 , 2008 . PA14 proteins were used as BLASTP queries against each bacterial genome . Putative orthologs were reciprocal best hits against the corresponding proteins in the subject genomes . Blast results against subject genomes were required to have an e-value equal to or less than 0 . 0001 . Reciprocal blasts against the PA14 genome were required to have e-values of 0 . 001 or less . Putative orthologs were required to align for at least 80 percent of their length and have less than 20% difference in protein sequence lengths , thereby conserving overall domain structure . The e-value constraint was permissive to allow detection of distant orthologs , but the requirement for alignment length was fairly stringent . Breadth of phylogenetic distribution , also called phylostratum , was a measure of the maximal phylogenetic distance at which an ortholog occurs . Breadth was defined as: 0 for proteins specific to PA14 , 1 for proteins that occur in multiple strains of Pseudomonas aeruginosa , but not in other species , 2 for proteins that occur in multiple Pseudomonas species but not in other genera , 3 if across gamma and beta proteobacteria , 4 if across proteobacteria , 5 if across eubacteria , and 6 if across eubacteria and archaea ( Figure S13 ) . P-values for overrepresentation and underrepresentation were calculated as Fisher Exact Test right and left probabilities , respectively , using version 1 . 21 of the Text::NSP::Measures::2D::Fisher Perl module , available from CPAN . org [93] . Multiple comparison correction using False Discovery Rate ( FDR ) was performed where indicated [94] . The maximal value of q was 0 . 05 . Genes of P . aeruginosa strain PA14 were mapped to KEGG pathways using the KEGG Mapper program ( http://www . genome . jp/kegg/tool/map_pathway1 . html ) . The total number of mutants in each pathway was summed for those genes in the PA14 NR set and for each mutant set analyzed . P-values for overrepresentation of each pathway were calculated using Fisher exact test right-probabilities , and their significance was assessed using the Bonferroni correction [94] , [95] . GO term analysis was performed using DAVID ( http://david . abcc . ncifcrf . gov/home . jsp ) using the genes in the NR set as a background , and the genes in the sets of analyzed positives as gene-lists . The DAVID software package calculates p-values for each GO term automatically , and also gives an EASE score , which was used to assess the significance of the overrepresentation of any given GO term .
|
Pseudomonas aeruginosa is an opportunistic human pathogen that can also infect a wide range of model organisms , including the nematode Caenorhabditis elegans . To identify P . aeruginosa genes that play key roles in the pathogenic process , we performed a screen for mutants that exhibited reduced ability to kill C . elegans using a previously constructed non-redundant library representing approximately 80% of the non-essential P . aeruginosa PA14 genes . We defined a functionally diverse set of 180 P . aeruginosa mutants ( representing 170 unique genes ) necessary for normal levels of virulence that included both known and novel virulence factors . The major contributors to P . aeruginosa virulence in the C . elegans infection model were not secretion systems or their corresponding effectors , but rather regulators ( particularly ones that are involved in quorum sensing ) and genes likely to play key roles in survival of P . aeruginosa within the host intestine . Moreover , these putative P . aeruginosa virulence genes are neither overrepresented in strain-specific regions nor in horizontally acquired genomic islands and furthermore tend to have orthologs that are widely distributed across sequenced prokaryotic species . These data underscore the diversity of pathways involved in virulence , and especially the importance of highly conserved genes for P . aeruginosa virulence in the C . elegans host model .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"microbiology",
"host-pathogen",
"interaction",
"animal",
"models",
"caenorhabditis",
"elegans",
"prokaryotic",
"models",
"model",
"organisms",
"microbial",
"evolution",
"bacterial",
"pathogens",
"microbial",
"pathogens",
"biology",
"pathogenesis",
"gram",
"negative",
"genetic",
"screens",
"genetics",
"evolutionary",
"biology",
"bacterial",
"evolution",
"genetics",
"and",
"genomics"
] |
2012
|
Genome-Wide Identification of Pseudomonas aeruginosa Virulence-Related Genes Using a Caenorhabditis elegans Infection Model
|
Depending on the environmental conditions , the pathogenic yeast Candida albicans can undergo different developmental programs , which are controlled by dedicated transcription factors and upstream signaling pathways . C . albicans strains that are homozygous at the mating type locus can switch from the normal yeast form ( white ) to an elongated cell type ( opaque ) , which is the mating-competent form of this fungus . Both white and opaque cells use the Ste11-Hst7-Cek1/Cek2 MAP kinase signaling pathway to react to the presence of mating pheromone . However , while opaque cells employ the transcription factor Cph1 to induce the mating response , white cells recruit a different downstream transcription factor , Tec1 , to promote the formation of a biofilm that facilitates mating of opaque cells in the population . The switch from the white to the opaque cell form is itself induced by environmental signals that result in the upregulation of the transcription factor Wor1 , the master regulator of white-opaque switching . To get insight into the upstream signaling pathways controlling the switch , we expressed all C . albicans protein kinases from a tetracycline-inducible promoter in a switching-competent strain . Screening of this library of strains showed that a hyperactive form of Ste11 lacking its N-terminal domain ( Ste11ΔN467 ) efficiently stimulated white cells to switch to the opaque phase , a behavior that did not occur in response to pheromone . Ste11ΔN467-induced switching specifically required the downstream MAP kinase Cek1 and its target transcription factor Cph1 , but not Cek2 and Tec1 , and forced expression of Cph1 also promoted white-opaque switching in a Wor1-dependent manner . Therefore , depending on the activation mechanism , components of the pheromone-responsive MAP kinase pathway can be reconnected to stimulate an alternative developmental program , switching of white cells to the mating-competent opaque phase .
The yeast Candida albicans is a member of the microbiota in the gastrointestinal and genitourinary tracts of most healthy persons , but it can also cause superficial as well as life-threatening systemic infections when host defenses are compromised . Morphological transitions play a major role in the biology of C . albicans and in the interactions of the fungus with its host . For example , in response to various environmental stimuli , C . albicans alters its morphology from a unicellular budding yeast to a multicellular hyphal form . The switch from yeast to filamentous growth facilitates tissue invasion and is associated with the transition of C . albicans from a harmless colonizer to a pathogen that causes symptomatic infections [1] . C . albicans can also switch from the normal , round-to-oval yeast morphology ( white ) to an elongated yeast cell type ( opaque ) , which is the mating-competent form of this diploid fungus [2] . Opaque cells can mate with opaque cells of opposite mating type to generate tetraploid fusion products , which may then undergo random chromosome loss to generate recombinant progeny in a parasexual cycle [3] . Switching of white cells to the opaque phase requires the transcription factor Wor1 , the master regulator of white-opaque switching [4] , [5] , [6] . WOR1 is expressed at very low levels in white cells , but an increase in the amount of Wor1 above a threshold induces switching to the opaque phase . Wor1 activates its own expression , resulting in a positive feedback loop that provides the high Wor1 levels required for maintenance of the opaque phenotype . Additional transcription factors , including the positive regulators Wor2 and Czf1 and the negative regulator Efg1 , which are themselves controlled by Wor1 , ensure bistable expression of WOR1 ( low in white and high in opaque cells ) and epigenetic inheritance of the two phases [7] . Switching to the mating-competent opaque cell type is restricted to strains that are homozygous at the mating type locus ( MTLa/a or MTLα/α ) . In heterozygous MTLa/α strains , switching of white cells to the opaque phase is inhibited by a heterodimeric repressor consisting of the homeodomain proteins a1 ( encoded by MTLa ) and α2 ( encoded by MTLα ) , which prevents WOR1 expression [2] , [4] , [5] , [6] . Most C . albicans strains in nature are MTL heterozygotes , but they can become homozygous by genomic rearrangements , which relieves them from repression by the a1-α2 repressor and renders the cells switching-competent [8] . White and opaque cells differ not only in their mating capacity , but also in many additional phenotypes and in the expression of genes that are unrelated to mating , suggesting that they are adapted to different environments within a mammalian host [9] , [10] , [11] . Opaque cells colonize skin more readily than do white cells , but they are much less virulent than white cells during a systemic infection [12] , [13] . Although opaque cells can form hyphae , they do not undergo the yeast-hypha transition under many conditions that stimulate hyphae formation in white cells [14] , [15] . This may result in a decreased capacity of opaque cells to escape from the bloodstream and invade into tissues . On the other hand , opaque cells can avoid recognition and phagocytosis by macrophages and neutrophils under conditions in which white cells are efficiently phagocytosed [16] , [17] , [18] . Therefore , switching to the opaque phase not only results in the acquisition of mating competence but may also allow escape from the native immune system in certain host niches in which the more aggressive white cells will be attacked , especially after hyphae formation [19] , [20] . Interestingly , both white and opaque cells respond to the presence of mating pheromone . α-pheromone , which is produced by opaque α-cells , binds to the receptor Ste2 on a-cells , and a-pheromone , produced by opaque a-cells , binds to the receptor Ste3 on α-cells . But while opaque cells induce a mating response that results in shmoo formation and cell fusion , pheromone binding to white cells induces the production of a biofilm , which stabilizes the pheromone gradient and facilitates mating of opaque cells [21] . This differential response of white and opaque cells is achieved by the pheromone-induced activation of a common MAP kinase cascade , consisting of the MAPKKK Ste11 , the MAPKK Hst7 , and the partially redundant MAPKs Cek1 and Cek2 , and the cell type-specific recruitment of different downstream transcription factors , Cph1 in opaque cells and Tec1 in white cells [22] , [23] . It was believed for a long time that white-to-opaque switching is a stochastic process that occurs spontaneously in few cells of a population . However , it has recently become evident that white cells can be induced under certain environmental conditions to switch en masse to the opaque phase [24] , [25] , [26] , [27] , [28] . Although several transcription factors have been identified that regulate white-opaque switching [4] , [5] , [6] , [7] , [28] , [29] , [30] , little is known about the upstream signal transduction pathways that allow white cells to respond to these inducing signals . Protein kinases are common components of signaling pathways that mediate cellular reactions to external and internal signals , and the protein kinase Tpk2 has recently been implicated in the induction of switching by environmental signals [26] . In the present study , we generated a comprehensive , tetracycline-inducible protein kinase expression library to investigate whether additional kinases are involved in the control of white-opaque switching . Intriguingly , we found that the pheromone-responsive MAP kinase pathway , which promotes the mating response in opaque cells and biofilm formation in white cells , can be rewired such that white cells recruit the opaque-specific transcription factor Cph1 instead of Tec1 , which then induces switching to the opaque phase .
Artificial expression of the master regulator WOR1 , and also of the positive regulator CZF1 , from a constitutive or inducible promoter induces switching of white cells to the opaque phase [4] , [5] , [6] , [7] , [28] , [30] . We reasoned that forced expression of protein kinases that act upstream of the transcriptional regulators to promote switching in response to environmental signals may similarly induce white cells to switch to the opaque phase . Therefore , we cloned all C . albicans genes encoding known or putative protein kinases and their regulators in a tetracycline-inducible gene expression cassette ( see Materials and Methods ) . We also included 21 putative hyperactive or dominant negative alleles of these kinases in the collection . The resulting library of 160 Tet-inducible protein kinases and regulators ( supplemental Table S1 ) was then integrated into the genome of the C . albicans MTLα/α strain WO-1 , in which white-opaque switching was originally discovered and which has been widely used as a model strain to study this developmental program . In each case , two independent transformants were kept to confirm the reproducibility of phenotypes that were induced by the expression of a kinase . To discover kinases whose forced expression from the Tet promoter induces switching to the opaque phase , white cells of the parental strain WO-1 and the strains containing the inducible protein kinase library were grown overnight in liquid medium in the presence or absence of doxycycline and then spread at an appropriate dilution on agar plates with or without doxycycline to allow the formation of colonies from individual cells . The results of this screening are summarized in supplemental Table S2 . As expected , all strains behaved like the parental strain WO-1 when they were grown in the absence of doxycycline and showed only basal levels of spontaneous switching . In contrast , forced expression of the protein kinases MPS1 , RAD53 , TPK1 , TPK2 , and the hyperactive STE11ΔN467 allele resulted in a strongly increased frequency of switching from the white to the opaque phase when the inducer doxycycline was present in the preculture and/or during subsequent colony growth on the agar plates . Ste11 is the MAPKKK of the pheromone-responsive MAP kinase signaling pathway ( but also functions in other pathways ) , and in the present study we focus on the role of this pathway in the regulation of white-opaque switching in C . albicans . The other identified kinases will be subject of future investigations . The pheromone-responsive signaling pathway contains two partially redundant MAP kinases , Cek1 and Cek2 , which activate the downstream transcription factors Cph1 and Tec1 [22] , [23] , [31] . Of note , the functions of the two transcription factors in the pheromone response differ . While Cph1 induces the expression of mating-specific genes in opaque cells and is required for mating , Tec1 has no role in mating of opaque cells , but promotes biofilm formation of white cells in response to pheromone produced by opaque cells . This response of white cells is thought to stabilize the pheromone gradient and facilitate mating of opaque cells in a population that contains a majority of white cells [21] . We therefore investigated if these downstream MAP kinases and transcription factors are also involved in the induction of white-opaque switching by the activated Ste11ΔN467 . To this aim , we generated deletion mutants of strain WO-1 lacking CEK1 , CEK2 , CPH1 , or TEC1 and expressed the hyperactive STE11ΔN467 allele from the Tet promoter in the various mutants , all of which were constructed twice independently ( see supplemental Table S3 ) . As can be seen in Fig . 1 , inactivation of CEK1 abolished STE11ΔN467-induced white-opaque switching , while deletion of CEK2 had no effect . The switching defect of the cek1Δ mutants was complemented by reintroduction of a functional CEK1 copy . These results indicate that the two MAP kinases , which have redundant functions in the mating pheromone response of opaque cells and in pheromone-induced biofilm formation of white cells [23] , [31] , also have divergent roles: CEK1 is required for the Ste11-induced switching of white cells to the mating-competent opaque form , whereas CEK2 is dispensable for this developmental program . Deletion of TEC1 did not affect the ability of the hyperactive Ste11 to induce white-opaque switching . In contrast , Ste11-induced switching was abolished in the cph1Δ mutants , and this defect was complemented by reintroduction of a functional CPH1 copy ( Fig . 1 ) . Therefore , Cph1 not only functions in the mating response of opaque cells but is also required for Ste11-induced switching of white cells to the opaque phase . In contrast , Tec1 induces biofilm formation in white cells in response to pheromone [22] , but is not required for the induction of white-opaque switching . We hypothesized that , if Cph1 is the downstream transcription factor that mediates Ste11-induced white-opaque switching , forced expression of CPH1 might also promote switching of white cells to the opaque phase . Indeed , expression of CPH1 from the Tet promoter in strain WO-1 strongly induced white-opaque switching ( Fig . 2 ) . Doxycycline-induced expression of TEC1 also caused an increase in the frequency of white-opaque switching , but this was not comparable to the stimulation by CPH1 , and TEC1 was also not required for Cph1-induced white-opaque switching ( Fig . 2 ) . Together , these results demonstrate that activation of the pheromone-responsive MAP kinase cascade in white cells by a hyperactive form of the MAPKKK Ste11 induces one arm of this signaling pathway , including the MAPK Cek1 and the transcription factor Cph1 , to promote switching to the opaque phase . White-opaque switching is controlled by a network of feedback loops comprising the positive regulators Wor1 , Wor2 , and Czf1 and the negative regulator Efg1 [7] . To investigate if the core positive regulators were required for Ste11/Cph1-induced white-opaque switching , we expressed CPH1 and the hyperactive STE11ΔN467 allele in mutants of strain WO-1 from which WOR1 , WOR2 , or CZF1 were deleted . Initial experiments showed that the colony phenotypes of the mutants were altered in some cases , making it difficult to decide whether cells had truly switched to the opaque phase ( see Fig . 3B , top panels ) . We therefore generated derivatives of strain WO-1 and of the mutants that expressed GFP or RFP from the opaque-specific OP4 promoter to determine whether opaque-like colonies generated after Tet-induced expression of CPH1 or STE11ΔN467 contained bona fide opaque cells ( see Fig . 3B , bottom panels ) . CPH1 and STE11ΔN467 induced white-opaque switching in the labeled wild-type reporter strains with the same efficiency as in the original parental strain WO-1 ( Fig . 3A , compare with Fig . 1 and 2 ) . No switching to the opaque phase was observed in wor1Δ and wor2Δ mutants , demonstrating that the master regulator Wor1 as well as Wor2 were also required for Cph1-induced switching . However , doxycycline-induced CPH1 expression promoted switching to the opaque phase also in the absence of CZF1 , while forced expression of STE11ΔN467 failed to increase the switching frequency above background levels ( Fig . 3A ) . Apparently , the overexpression of CPH1 was strong enough to promote the switch in the absence of Czf1 , which is part of a positive feedback loop that facilitates switching . In contrast , the comparatively weaker induction by the hyperactive STE11ΔN467 allele was not sufficient to overcome the threshold for the switch to occur in the absence of Czf1 . These results demonstrate that , when sufficiently active , Cph1 does not depend on Czf1 to promote white-opaque switching , but it still requires the master regulator Wor1 . Tet-induced expression of TEC1 or STE11 , but not CPH1 , in white cells of a C . albicans MTLa/a strain has been shown to promote biofilm formation under normally noninducing conditions [22] . In line with these results , we found that Tet-induced expression of the hyperactive STE11ΔN467 allele in the MTLα/α strain WO-1 also induced biofilm formation in addition to white-opaque switching ( Fig . 4 ) . To investigate which downstream MAP kinases and transcription factors were required for STE11ΔN467-induced biofilm growth , we assayed biofilm formation in cek1Δ , cek2Δ , cph1Δ , and tec1Δ mutants expressing the hyperactive STE11 allele . As expected , STE11ΔN467-induced biofilm formation was abolished in tec1Δ mutants . Surprisingly , however , the hyperactive Ste11 was also unable to promote biofilm formation in mutants lacking CPH1 . Similarly , no induction of biofilm formation was seen in the absence of CEK1 , while deletion of CEK2 had no effect in these assays , demonstrating that the two MAP kinases are not redundant for Ste11ΔN467-induced biofilm development in strain WO-1 . Reintroduction of an intact copy of CPH1 and CEK1 into the respective mutants restored STE11ΔN467-induced biofilm growth , confirming that the mutant phenotype was caused by the deletion of these genes ( Fig . 4 ) . As both CPH1 and TEC1 were required for the induction of biofilm formation by the hyperactive STE11ΔN467 allele , we tested whether forced expression of either of these transcription factors would promote biofilm formation in the absence of the other in strain WO-1 . Tet-induced CPH1 expression indeed caused biofilm formation in this strain background , but this induction depended on the presence of TEC1 ( Fig . 5 ) . In contrast , Tet-induced expression of TEC1 promoted biofilm formation both in the presence and absence of CPH1 . The results presented above demonstrate that Tet-induced expression of the hyperactive STE11ΔN467 allele or the downstream transcription factor CPH1 promoted both biofilm formation and white-opaque switching in strain WO-1 . Biofilm formation apparently was not a prerequisite for the induction of white-opaque switching , because switching was efficiently induced also in tec1Δ mutants , which did not form biofilms . Since biofilm formation has often been linked to hyphal morphogenesis , we examined the phenotype of the cells after Tet-induced STE11ΔN467 or CPH1 expression before plating for subsequent colony formation . For comparison , we also tested cells that expressed the known positive regulator CZF1 from the Tet promoter . These latter cells did not yet exhibit the opaque morphology after induction of CZF1 expression in liquid culture ( Fig . 6A , bottom left panel ) , but were programmed to switch to the opaque phase , as almost all of them formed opaque colonies after plating ( see supplemental Table S2 ) . Wild-type cells expressing STE11ΔN467 or CPH1 mainly grew as filaments , similar to cells expressing TEC1 , explaining the formation of biofilms on the plastic surface ( Fig . 6A , middle panels ) . In the absence of TEC1 , filamentation of cells expressing STE11ΔN467 was strongly reduced and most cells exhibited the normal white yeast morphology ( Fig . 6A , top right panel ) . In contrast , Tet-induced expression of CPH1 in a tec1Δ background resulted in switching to the opaque phase already during growth in liquid medium , indicating that when filamentation is blocked by the absence of Tec1 , upregulation of CPH1 expression in a switching-competent strain directly promotes opaque cell formation ( Fig . 6A , bottom right panel ) . Strikingly , many of the opaque cells formed shmoos , in accord with the previously reported finding that CPH1 overexpression in opaque cells induces the mating response [32] . In order to understand why the filamentous cells observed after Tet-induced STE11ΔN467 or CPH1 expression were programmed to switch to the opaque phase after plating , we determined the expression levels of the master regulator WOR1 in these filamentous cells by RT-qPCR . Despite some variation between biological replicates , WOR1 transcript levels were consistently increased upon STE11ΔN467 expression in all experiments with the two independently constructed strains , on average ca . 15-fold above those in the parental control strain ( Fig . 6B , top left panel ) . CPH1 transcript levels were also elevated , whereas TEC1 mRNA levels remained unchanged ( Fig . 6B , top middle and right panels ) . The latter result is in agreement with the fact that Cph1 , but not Tec1 , is the downstream transcription factor that promotes STE11ΔN467-induced white-opaque switching . An even stronger upregulation of WOR1 ( >800-fold ) was observed upon Tet-induced expression of CPH1 itself ( Fig . 6B , bottom left panel ) , which is explained by the significantly higher CPH1 transcript levels in these cells compared to those seen after expression of STE11ΔN467 from the Tet promoter ( Fig . 6B , middle panels ) . These data demonstrate that the activation of the MAP kinase cascade in white cells by the hyperactive STE11ΔN467 has a different outcome compared to the induction by pheromone . Instead of causing TEC1 upregulation , Ste11 lacking its N-terminal inhibitory domain increases CPH1 expression , which in turn induces WOR1 expression , thereby programming the cells to switch to the opaque phase . White-opaque switching can be induced by different environmental signals , but there are strain-specific differences in the response of switching-competent strains to the various stimuli [24] , [25] , [26] , [28] . In strain WO-1 , switching is strongly induced by a transient incubation in an anaerobic environment ( 0% O2 , 18% CO2 ) , while other tested strains did not switch to the opaque phase under these conditions [28] . We therefore investigated if Cph1-induced white-opaque switching might be a peculiar characteristic of strain WO-1 or if Cph1 can promote switching also in other C . albicans strains and in both mating types . For this purpose , we deleted either the MTLa or the MTLα locus in the commonly used reference strain SC5314 to generate switching-competent α and a derivatives , respectively , into which the Ptet-CPH1 fusion was subsequently introduced . Doxycycline did not induce white-opaque switching in these strains; however , using a control Ptet-GFP reporter fusion we found that the Tet promoter was much less efficiently induced in this strain background than in strain WO-1 ( unpublished results ) . Hence , we expressed CPH1 from another regulatable promoter , POPT3 , which is efficiently induced in strain SC5314 when the cells grow on BSA as a nitrogen source [33] . Expression of CPH1 from the OPT3 promoter during growth in YCB-BSA-YE medium strongly stimulated white-opaque switching in independently generated a and α derivatives of strain SC5314 ( Fig . 7 ) . Incubation of the untransformed parental strains in the same growth medium did not promote white-opaque switching , confirming that switching was induced by CPH1 expression . We consistently observed a higher switching frequency when the POPT3-CPH1 fusion was integrated into the OPT3-1 allele as compared to the OPT3-2 allele . Allele-specific differences in the activity of the OPT3 promoter did not seem to be the reason , because a POPT3-GFP reporter fusion was expressed at comparable levels after integration at either of the two loci ( data not shown ) . Therefore , minor differences in the resulting OPT3-CPH1 hybrid transcripts ( e . g . , stability or translational efficiency ) may specifically affect Cph1 levels and , consequently , the switching frequency . Regardless , these results demonstrate that Cph1-induced white-opaque switching is not specific to strain WO-1 and is independent of mating type . The finding that activation of the Cph1-dependent MAP kinase pathway in white cells promoted switching to the opaque phase suggested that the presence of mating pheromone might also stimulate white-opaque switching . However , in contrast to opaque cells , white cells do not [22] , [23] or not strongly ( Fig . 8C ) upregulate CPH1 expression in response to pheromone and no pheromone-induced white-opaque switching has been reported so far . In line with this , the addition of synthetic α-pheromone to white a-cells derived from strain SC5314 did not induce switching to the opaque phase under various growth conditions tested , and a mixture of opaque a- and α-derivatives of strain SC5314 ( used as a source of a-pheromone ) also did not stimulate white-opaque switching in the MTLα/α strain WO-1 when the cells were coincubated ( data not shown , see Materials and Methods for details ) . As white cells use Tec1 to induce biofilm formation in response to pheromone , we also tested the effect of a-pheromone on tec1Δ mutants of strain WO-1; however , no pheromone-induced switching was observed in the mutants . Consequently , other signals may activate the MAP kinase pathway in a different way from that stimulated by pheromone to result in CPH1 instead of TEC1 upregulation in white cells , similar to the effect of the hyperactive Ste11ΔN467 . We therefore investigated whether Cph1 is required for white-opaque switching in response to signals that efficiently stimulate switching in strain WO-1 . For this purpose , we determined the switching frequency of the wild-type strain WO-1 and the cph1Δ mutants under various inducing conditions; mutants lacking the master regulator Wor1 or the positive regulator Czf1 were included for comparison . In contrast to czf1Δ mutants , in which the frequency of switching was drastically reduced under all tested conditions ( no switching was observed in wor1Δ mutants , as expected ) , cells lacking Cph1 switched to the opaque phase with the same efficiency as the parental wild-type strain WO-1 when switching was induced by incubation in an anaerobic jar , in the presence of ketoconazole , or by growth on GlcNAc as carbon source ( Fig . 10 ) . Therefore , conditions that induce white cells to switch to the opaque phase by activating CPH1 remain to be discovered ( see discussion ) .
The recent observation by several groups that switching of white cells to the opaque phase does not only occur stochastically in few cells of a population , but can be strongly stimulated by certain environmental cues [24] , [25] , [26] , [28] , suggested that expression of the master regulator WOR1 is induced by upstream signal transduction pathways in response to these signals . The cAMP/PKA signaling pathway has been implicated in the environmental induction of white-opaque switching , because deletion of components of this pathway reduced , albeit did not abolish , the stimulation of switching by GlcNAc and elevated CO2 concentrations [25] , [26] . In this work , we used an overexpression approach to identify protein kinases that stimulate white-opaque switching . By generating and screening a comprehensive library of C . albicans strains that express all protein kinases of this organism from a tetracycline-inducible promoter , we discovered several kinases that could induce white cells to switch to the opaque phase . In addition to the recently identified Tpk2 [26] , we found that the homologous kinase Tpk1 as well as the kinases Mps1 and Rad53 also stimulated white-opaque switching when overexpressed from the Tet promoter . The involvement of different protein kinases in the regulation of white-opaque switching probably reflects the fact that a variety of signals stimulate white cells to switch to the opaque phase . It should be noted that additional kinases not identified in our screening could nevertheless be involved in the control of white-opaque switching , because overexpression of a wild-type kinase may not necessarily have a phenotypic effect ( as was the case for wild-type STE11 , which did not stimulate switching ) . A particularly intriguing result of our present study was that a hyperactive form of the MAPKKK Ste11 , which acts in the pheromone response pathway , also promoted switching . This finding came as a surprise , because the pheromone-responsive MAP kinase signaling pathway is known to induce the mating response in opaque cells and biofilm formation in white cells , but so far it has not been implicated in the regulation of the switching event itself . No induction of white-opaque switching by pheromone has been reported in previous studies investigating the pheromone response of white cells [21] , [23] , and we also did not observe stimulation of switching by pheromone , suggesting that the MAP kinase pathway can be activated in an alternative way in white cells to promote switching to the opaque phase . The other kinases identified in our study do not seem to function via the MAP kinase pathway , because Tet-induced expression of MPS1 , RAD53 , TPK1 , and TPK2 stimulated switching also in cph1Δ and ste11Δ mutants ( see supplemental Fig . S1 ) , arguing that these kinases function via alternative signaling pathways to activate WOR1 . Unlike the hyperactive STE11 , these other kinases also did not promote biofilm formation in our assays ( data not shown ) , supporting the assumption that they do not activate the MAP kinase pathway under these conditions . White and opaque cells use the same upstream components of the pheromone-induced MAP kinase signaling pathway , but different transcription factors to effect their specific responses . White cells induce expression of TEC1 , but not CPH1 , whereas opaque cells induce expression of CPH1 , but not TEC1 , in response to pheromone [22] , [23] , [34] . It is not currently known how the upregulation of the alternative downstream transcription factor , CPH1 in white cells and TEC1 in opaque cells , in the presence of pheromone is blocked in the two cell types . In Saccharomyces cerevisiae , components of the pheromone-responsive MAP kinase cascade are also used for the induction of a different developmental program , invasive growth , under starvation conditions [36] . Here , the scaffold protein Ste5 , which binds all three components of the MAP kinase cascade , acts as a conformational switch that gates the flow of information to ensure a proper physiological response to different inducing signals [37] . In C . albicans , the homologous scaffold protein Cst5 is required for both the Cph1-mediated mating response of opaque cells and Tec1-dependent biofilm formation of white cells [32] . The hyperactive Ste11ΔN467 , which lacks the binding site for Cst5 [35] , caused upregulation of CPH1 instead of TEC1 in white cells , i . e . , these cells had overcome the block in CPH1 induction upon activation of the MAP kinase pathway ( Fig . 11 ) . However , the absence of Cst5 was not sufficient to enable wild-type Ste11 to induce switching of white cells to the opaque phase , indicating that additional signals are required for Ste11 to recruit Cph1 and promote switching . Overexpression of CPH1 has been reported already a decade ago to increase mating efficiency [31] . At the time of that study , the linkage of mating to white-opaque switching had not yet been uncovered and the mating experiments were performed with white cells . While overexpression of CPH1 may have enhanced the mating response of opaque cells that were already in the population , our results suggest that the increased mating efficiency might also have been caused by Cph1-induced switching of white cells to the opaque phase . We observed that some of the kinases identified in our study efficiently stimulated white-opaque switching only when their expression was induced in the preculture , but not during growth on agar plates ( STE11ΔN467 , TPK1 ) , while others ( RAD53 , TPK2 ) promoted switching under both conditions ( see supplemental Table S2 ) . Indeed , growth conditions affected the inducibility of white-opaque switching by overexpressed STE11ΔN467 and CPH1: Growth of the cells under static conditions in microtiter plates , a procedure that we used to streamline screening of the library , resulted in more efficient switching than the usual culturing in Erlenmeyer flasks or glass tubes with shaking . Under the latter conditions , the switching rates obtained after Tet-induced CPH1 expression were only slightly reduced , but no significant induction of switching by Ste11ΔN467 was observed , indicating that white cells could more easily be stimulated to switch when growing on the bottom of a microtiter plate ( data not shown ) . A reduced growth rate , which facilitates Wor1 accumulation in a cell [24] , or local buildup of higher CO2 levels , which promote white-opaque switching [25] , may contribute to this effect . In this regard , it is interesting to note that increased CO2 concentrations were reported to induce white-opaque switching only when the cells were grown on agar plates , but not in liquid culture [25] , and GlcNAc induces switching only in aged , but not in fresh cultures [26] . In addition , Tet-induced expression of Flo8 , another transcription factor that was recently found to be involved in the regulation of white-opaque switching , induced switching only in the presence of elevated CO2 concentrations [29] . Evidently , the ability of a regulatory factor to induce switching depends on the environmental conditions , because these will affect the activity of additional regulators . It has been proposed that the pheromone-induced biofilm formation response observed in white cells has evolved via the recruitment of components from the ancestral pheromone response pathway ( all upstream components , from the pheromone receptor to the MAP kinases Cek1 and Cek2 ) and the filamentation pathway ( the transcription factor Tec1 ) as well as target genes for biofilm formation [22] . Our results demonstrate that this signaling pathway can be used in a highly flexible way , depending not only on the cell type but also on the manner in which it is activated . The downstream transcription factor Cph1 mediates the mating response of opaque cells , but can also stimulate white cells to switch to the opaque phase to become mating-competent . In the latter case , one of the two MAP kinases , Cek1 , is specifically recruited to transmit the signal from activated Ste11 to Cph1 , whereas Cek1 and Cek2 have at least partially redundant functions in the pheromone-induced biofilm formation of white cells and in the mating response of opaque cells [23] , [31] . CPH1 was only required for the induction of white-opaque switching by the artificially activated Ste11 , but not for the induction of switching by the tested environmental conditions and for spontaneous switching , in agreement with findings by other researchers [23] . Of note , Cek1 is also known to be activated upon cell wall stress , for example after treatment with the cell wall disturbing agent tunicamycin , but Cph1 has not been implicated as a Cek1 downstream target under these conditions [38] , [39] , and we were unable to stimulate white-opaque switching with tunicamycin ( unpublished results ) . In MTL heterozygous cells , the Cph1-dependent MAP kinase pathway induces filamentous growth under starvation conditions [40] , [41] . Our results provide further evidence that certain transcription factors ( Cph1 , Czf1 , Efg1 , Flo8 ) are involved in the regulation of both filamentation and white-opaque switching . It is conceivable that the N-terminally truncated , hyperactive form of Ste11 , which induced CPH1 upregulation in white cells and switching to the opaque phase , reflects a normal function of the MAP kinase pathway when activated by an unknown physiological signal . In this respect , it is remarkable that white-opaque switching , which was thought to be restricted to MTL homozygous strains , has recently been observed in MTLa/α strains under specific growth conditions [42] . So far , we have not found environmental conditions that promote white-opaque switching in a Cph1-dependent fashion and such conditions might be encountered only in suitable host niches in vivo . It is even possible that the presence of pheromone is a prerequisite , and that additional signals are required to overcome the block in CPH1 upregulation that is seen upon pheromone treatment of white cells in vitro . This , in turn , raises the intriguing hypothesis that C . albicans white cells may have the ability to react to the presence of a potential mating partner ( pheromone-secreting opaque cells ) by switching to the opaque phase and thus become themselves mating-competent in an appropriate environment .
The C . albicans strains used in this study are listed in supplemental Table S3 . All strains were stored as frozen stocks with 15% glycerol at −80°C . The strains were subcultured separately in the white and opaque phases at room temperature on agar plates containing Lee's medium , pH 6 . 8 [43] , and 5 µg/ml phloxine B , which selectively stains opaque colonies pink [44] . Strains were routinely grown in YPD ( 10 g yeast extract , 20 g peptone , 20 g glucose ) or SD ( 1 . 7 g yeast nitrogen base without amino acids [YNB; BIO 101 , Vista , Calif . ] , 20 g glucose per liter ) liquid medium at 30°C in a shaking incubator . To prepare solid media , 1 . 5% agar was added to the media before autoclaving . For selection of nourseothricin-resistant transformants , 200 µg/ml nourseothricin ( Werner Bioagents , Jena , Germany ) was added to YPD agar plates . To obtain nourseothricin-sensitive derivatives in which the SAT1 flipper cassette was excised by FLP-mediated recombination , transformants were grown overnight in YCB-BSA-YE medium ( 23 . 4 g yeast carbon base , 4 g bovine serum albumin , 2 g yeast extract per liter , pH 4 . 0 ) without selective pressure to induce the SAP2 promoter controlling caFLP expression . Alternatively , strains containing a SAT1 flipper cassette in which the caFLP gene is expressed from the MAL2 promoter ( as in plasmids pOP4G4 , pOP4R2 , and pMTLΔ1 ) were grown overnight in YPM medium ( 10 g yeast extract , 20 g peptone , 20 g maltose per liter ) instead of YCB-BSA-YE to induce the MAL2 promoter . Appropriate dilutions were plated on YPD agar plates containing 10 µg/ml nourseothricin and grown for 2 days at 30°C . Nourseothricin-sensitive clones were identified by their small colony size and confirmed by restreaking on YPD plates containing 200 µg/ml nourseothricin as described previously [45] . YCB-BSA-YE medium was also used to induce gene expression from the OPT3 promoter . To induce gene expression from the Tet promoter , white cells of the strains containing the protein kinase library were grown for 24 h at 30°C in Lee's medium in 96-well microtiter plates in the absence or presence of 50 µg/ml doxycycline , diluted 10−5 , and spread on Lee's agar plates with or without 50 µg/ml doxycycline . The plates were incubated for 7 days at room temperature and the number of white , opaque , and mixed white/opaque colonies was determined . Induction of white-opaque switching by environmental signals was performed as described in the legend to Fig . 9 . Incubation under anaerobic conditions was performed in an anaerobic jar ( Anaerocult , Merck KGaA , Darmstadt , Germany ) that generates an oxygen-free milieu in a CO2 atmosphere ( 18% CO2 ) within one hour . To generate a comprehensive library containing all known or putative protein kinases and kinase regulators of C . albicans , we searched the Candida genome database ( CGD , http://www . candidagenome . org ) for genes that were annotated with this function . These genes were amplified from genomic DNA of strain SC5314 by PCR with primers that introduced a SalI site in front of the start codon and a BglII site behind the stop codon ( primer sequences are available upon request ) . For genes with internal SalI or BglII sites , primers containing compatible XhoI and/or BamHI sites were used . The PCR products were appropriately digested and cloned in place of the GFP reporter gene in the SalI/BglII-digested vector pNIM6 [28] , which is identical to the originally developed Tet-On vector pNIM1 [46] , except that it contains the TEF3 transcription termination sequence instead of the ACT1 terminator behind GFP . All cloned genes were completely sequenced to confirm their identity and to exclude PCR errors . For the deletion of CEK1 , CEK2 , CPH1 , CST5 , STE11 , TEC1 , and WOR2 , the upstream and downstream regions of the genes were amplified as SacI-SacII and XhoI-ApaI fragments , respectively , and cloned on both sides of the SAT1 flipper cassette in plasmid pSFS5 , a derivative of plasmid pSFS2 in which the caFLP gene is placed under the control of the SAP2 promoter instead of the MAL2 promoter [47] to result in pCEK1M1 , pCEK2M1 , pCPH1M1 , pCST5M1 , pSTE11M1 , pTEC1M1 , and pWOR2M3 , respectively . For complementation of the cek1Δ and cph1Δ mutants , the complete CEK1 and CPH1 coding regions and flanking sequences were cloned as SacI-SacII fragments and inserted in place of the upstream flanking region of pCEK1M1 and pCPH1M1 , generating pCEK1K1 and pCPH1K1 , respectively . The CPH1 and TEC1 coding regions were also amplified and inserted in plasmid pNIM6 to generate doxycycline-inducible expression cassettes , as described above for the protein kinase expression library . For deletion of the MTLa or MTLα locus of strain SC5314 , ca . 0 . 8 kb of the common flanking regions were amplified as SacI-SacII and XhoI-ApaI fragments and cloned on both sides of the SAT1 flipper cassette in plasmid pSFS2 to generate pMTLΔ1 . To express CPH1 from the OPT3 promoter , the CPH1 coding sequence was substituted for GFP in plasmid pOPT3G22 [33] , resulting in pOPT3-CPH1 . HA-tagged versions of CEK1 , CEK2 , and CPH1 were generated by amplifying the C-terminal parts of the genes with primers that introduced a BamHI site ( encoding a Gly-Ser linker ) instead of the stop codons . The PCR products were digested with BamHI and at internal or introduced SacI sites , fused to a PCR-amplified BamHI-SacII fragment containing the 3×HA-TACT1 sequences from pZCF36H2 [48] , and substituted for the upstream flanking sequences in the corresponding deletion cassettes , resulting in pCEK1H1 , pCEK2H1 , and pCPH1H1 . C . albicans strains were transformed by electroporation [45] with the gel-purified inserts from the plasmids described above . The cassettes from the Tet-inducible protein kinase expression library were separated from the plasmid backbone by digestion with SacII/ApaI or SacII/KpnI ( in some cases a partial digest was required ) . Gene deletion and reinsertion cassettes were excised from the corresponding plasmids by SacI/ApaI digestion . For insertion of POP4-GFP and POP4-RFP reporter fusions into various strain backgrounds , the ApaI-SacI fragments from pOP4G2 ( with the caSAT1 marker alone ) or from pOP4G4 and pOP4R2 ( with the recyclable SAT1 flipper cassette ) [18] were used . The correct integration of each construct was confirmed by Southern hybridization using the flanking sequences as probes . In each case , two independent series of strains were generated and used for further analysis . Genomic DNA from C . albicans strains was isolated as described previously [45] . The DNA was digested with appropriate restriction enzymes , separated on a 1% agarose gel , transferred by vacuum blotting onto a nylon membrane , and fixed by UV crosslinking . Southern hybridization with enhanced chemiluminescence-labeled probes was performed with the Amersham ECL Direct Nucleic Acid Labelling and Detection System ( GE Healthcare UK Limited , Little Chalfont Buckinghamshire , UK ) according to the instructions of the manufacturer . Overnight cultures of C . albicans strains in SD medium were diluted to 107 cells/ml in Lee's medium with or without 50 µg/ml doxycycline and 500 µl of these suspensions was added to each well of a 24-well polystyrene microtiter plate . After 24 h of incubation at 30°C , the wells were gently washed with PBS and imaged . Quantification of biofilm formation was performed as previously described [49] , with some modifications . Three hundred microliters of the cell suspensions was added to each well of a 96-well polystyrene microtiter plate . A well containing Lee's medium without cells was included as reference . After 24 h of incubation at 30°C , the medium was aspirated and the wells were washed three times with 200 µl sterile PBS . Subsequently , 100 µl crystal violet solution ( 1% ) was added to each well for 5 min . The wells were washed three times with sterile water and bound crystal violet was released by adding 200 µl of 33% acetic acid . The obtained solution was transferred to a new microtiter plate and the absorbance read at 595 nm . Overnight cultures of C . albicans strains were diluted 10−2 in Lee's medium with 50 µg/ml doxycycline and grown for 18 h at 30°C in individual wells of a flat-bottomed 96-well polystyrene microtiter plate . Cells were harvested and total RNA was extracted by the hot acidic phenol method [50] combined with a purification step with the RNeasy mini kit ( Qiagen , Hilden , Germany ) , and treated with Turbo DNA-free DNase ( Ambion , Austin , TX ) . cDNA was synthesized using 500 ng of total RNA from each sample with the Superscript III Super Mix ( Invitrogen , Karlsruhe , Germany ) . PCR was performed on a MyiQ Real-time PCR system using the iQ SYBR Green Supermix kit ( Bio-Rad Laboratories , Hercules , CA ) . Relative mRNA levels , adjusted to ACT1 mRNA levels , were calculated using expression levels in the wild-type strain WO-1 ( set to 1 ) as a reference . The primers used are listed in supplemental Table S4 . Overnight cultures of the strains were tenfold diluted in 10 ml fresh Lee's medium with and without doxycycline or α-pheromone ( see below ) and grown without shaking for 18 h at 30°C . The cells were pelleted by centrifugation , washed in 2 . 5 ml breaking buffer ( 100 mM Tris-HCl [pH 7 . 5] , 200 mM NaCl , 20% glycerol , 5 mM EDTA ) , and resuspended in 500 µl breaking buffer supplemented with protease and phosphatase inhibitors ( 100 mM Tris-HCl [pH 7 . 5] , 200 mM NaCl , 20% glycerol , 5 mM EDTA , 0 . 1% β-mercaptoethanol , cOmplete EDTA-free Protease Inhibitor Cocktail and PhosStop Phosphatase Inhibitor Cocktail [Roche Diagnostics GmbH , Mannheim , Germany] ) . An equal volume of 0 . 5-mm acid-washed beads was added to each tube . Cells were mechanically disrupted on a FastPrep-24 cell-homogenizer ( MP Biomedicals , Solon , USA ) for two 40-seconds intervals , with 1 min on ice between each cycle . Samples were centrifuged at 13 , 000 rpm for 10 min at 4°C , the supernatant was collected , and the protein concentration was quantified using the Bradford protein assay . Extracts were heated for 5 min at 95°C , and equal amounts of protein of each sample were separated on an SDS-10% polyacrylamide gel . Separated proteins were transferred onto a nitrocellulose membrane with a Mini-Protean System ( Bio-Rad , Munich , Germany ) and stained with Ponceau S to control for equal loading . Membranes were blocked in 5% milk in PBS at room temperature for 1 hour and subsequently incubated overnight at 4°C with rat monoclonal anti-HA-Peroxidase antibody ( Roche Diagnostics GmbH , Mannheim , Germany ) . Membranes were washed in PBS containing 0 . 1% Tween-20 and signals detected with ECL chemiluminescence detection system ( GE Healthcare Bio-Sciences GmbH , Munich , Germany ) . For phosphatase treatment , cell extracts were washed in Amicon-ultra 10 K columns ( Millipore Corporation , Billerica , USA ) with three volumes of breaking buffer without EDTA and phosphatase inhibitors , and subsequently washed with two volumes of phosphatase buffer supplemented with 1 mM MnCl2 . Washed cell extracts were incubated with λ Protein Phosphatase ( New England Biolabs , Ipswich , USA ) at 30°C for 30 min . To investigate whether white-opaque switching could be induced by pheromone , white cells of MTLa/Δ derivatives of strain SC5314 were grown overnight in SD medium at room temperature , washed with water , and resuspended in Lee's medium or Spider medium to a concentration of 107 cells/ml . α-pheromone ( Seqlab , Göttingen , Germany ) was added at a concentration of 10 µg/ml , and the cultures were incubated at room temperature for 24 h . An appropriate dilution of the cultures was spread on Lee's agar plates with phloxin B and incubated for 7 days at room temperature to determine the frequency of opaque colonies . The experiment was repeated several times using additional additives and conditions along the pheromone treatment , including addition of 1 µg/ml pepstatin A ( to prevent pheromone degradation by secreted aspartic proteases ) and/or 10% FCS , incubation in the presence of 1% CO2 , incubation at 30°C or 37°C , and elongation of the treatment time to 48 h . In addition , the cells were also grown in the absence or presence of α-pheromone on Lee's agar plates without or with 1 . 2 M sorbitol in normal air or in a 1% CO2 atmosphere . To test the effect of a-pheromone on the MTLα/α strain WO-1 , opaque cells of MTLa/Δ and MTLα/Δ derivatives of strain SC5314 were mixed ( to serve as a source of both a- and α-pheromone ) and incubated with white cells of nourseothricin-resistant derivatives of strain WO-1 ( the wild-type strains WOP4G2A and -B and the tec1Δmutants WTEC1M3A and -B ) to a final concentration of 107 cells/ml in Lee's medium . The proportions of cells in the mixture were 50% white test cells , 25% opaque a-cells , and 25% opaque α-cells . The cell suspensions were incubated at room temperature for 24 h , diluted , and spread on Lee's agar plates with phloxin B and 100 µg/ml nourseothricin , on which only the nourseothricin-resistant test strains could grow . The plates were incubated for 7 days at room temperature and the presence of opaque colonies was recorded . In control experiments , we incubated RFP-expressing MTLa/Δ and MTLα/Δ opaque cells ( strains SCMTLαM2BOP4R22 and SCMTLaM2BOP4R22 ) together with GFP-expressing opaque cells of strain WO-1 ( strains WOP4G2A and -B ) . Microscopic inspection of the cells after 6 h of incubation showed that both GFP- and RFP-expressing cells formed shmoos , demonstrating that sufficient a-pheromone was produced by the mixture to induce the mating response in the MTLα/α strain WO-1 .
|
The pathogenic yeast Candida albicans can switch from the white yeast form to the mating-competent opaque form . Opaque cells are less virulent than white cells , but they can avoid recognition by phagocytes , indicating that white-opaque switching has evolved as an adaptation mechanism of C . albicans to specific host niches . Both white and opaque cells respond to mating pheromone by activating the Ste11-Hst7-Cek1/Cek2 MAP kinase pathway , but with different outcomes . Opaque cells utilize the transcription factor Cph1 to induce the mating response , whereas white cells recruit a different downstream transcription factor , Tec1 , to promote biofilm formation . We used a comprehensive protein kinase expression library to gain insight into the signaling pathways that regulate switching from the white to the opaque phase and found that a hyperactive form of the upstream kinase Ste11 induced white opaque-switching , a behavior that did not occur in response to pheromone . Hyperactive Ste11 functions via the opaque-specific transcription factor Cph1 instead of the white-specific transcription factor Tec1 to promote this alternative developmental program . Therefore , depending on the activation mechanism , components of the pheromone-responsive MAP kinase pathway can be rewired to stimulate a transition from the more virulent white form to the less aggressive , but mating-competent opaque form .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Activation of the Cph1-Dependent MAP Kinase Signaling Pathway Induces White-Opaque Switching in Candida albicans
|
We present a novel regularization scheme called The Generalized Elastic Net ( GELnet ) that incorporates gene pathway information into feature selection . The proposed formulation is applicable to a wide variety of problems in which the interpretation of predictive features using known molecular interactions is desired . The method naturally steers solutions toward sets of mechanistically interlinked genes . Using experiments on synthetic data , we demonstrate that pathway-guided results maintain , and often improve , the accuracy of predictors even in cases where the full gene network is unknown . We apply the method to predict the drug response of breast cancer cell lines . GELnet is able to reveal genetic determinants of sensitivity and resistance for several compounds . In particular , for an EGFR/HER2 inhibitor , it finds a possible trans-differentiation resistance mechanism missed by the corresponding pathway agnostic approach .
Previous approaches that combine gene interaction data with genomic features can be roughly divided into three categories . The first category focuses on feature modification , with the most popular approach being dimensionality reduction by grouping genes together according to functional categories . A single summary measure is then derived for each category , resulting in a feature space of much lower dimensionality than the original . All of the traditional analysis methods can then be applied in the resulting lower-dimensional setting , yielding higher robustness of the trained models [9–12] . Outside the realm of dimensionality reduction , methods that perform feature modification on the basis of gene interaction data include computing local entropy measures [13] , spin states in an Ising model [14] , and diffusion kernel values [15] . At the other end , the second category of methods operates by first applying predictors to obtain a set of discriminative scores—one for each feature / gene—and then using gene sets to elucidate pathways saturated with highly discriminative scores . Gene Set Enrichment Analysis [16] , Significance Analysis of Function and Expression [17] , and a method by Lee , et al . [56] do exactly this to arrive at a single differential score for each gene set of interest . Rather than using curated sets of genes , one may also place differential scores directly onto the gene interaction network and look for saturated subnetworks . The problem is NP-hard [18 , 19] and thus requires approximation techniques such as Simulated Annealing [18] , node color coding methods [19] and diffusion heat models on graphs with single [20] and multiple [21] data types . The first two categories utilize gene interaction information in a way that is entirely decoupled from the underlying predictor method: any algorithm that maps genomic input features to phenotypic outcomes and produces discriminative scores for each feature can be used . While this allows for a higher level of generality , intuitively , one would expect to achieve better accuracy if the predictor was able to utilize gene interactions directly , as part of its training . Methods that follow this intuition make up the third category , and we highlight several methods that make direct use of gene interaction networks as part of training . Dutkowski and Ideker proposed Network-Guided Forests [22] , a method that uses discriminative genomic features as decision tree nodes while forcing the edges between the nodes to coincide with a gene interaction network . The intuition behind Network-Guided Forests was later extended to a clustering setting , where gene network information was used with somatic mutations to derive Network-Based Stratification , a method for identifying clinically-relevant cancer subtypes [23] . In a linear model setting , several methods have used gene network information to guide feature selection during training . Johannes , et al performed recursive feature elimination for Support Vector Machines ( SVMs ) , using GeneRank to assign network-based importance weights to features at each elimination step [24] . Jang , et al proposed Stepwise Group Sparse Regression ( SGSR ) that utilized the grouping of genes according to functional pathways , rather than network connectivity directly [25] . SGSR is an iterative procedure that is initialized to a sparse LASSO-regularized linear model; at each iteration , the method adds a functional group of genes that results in large improvement in classification accuracy until the model is saturated and no further improvement is possible . Another LASSO-based model is Sparse Group LASSO ( SGL ) , which was utilized by Silver , et al . to find genes and pathways associated with high-density lipoprotein cholesterol in a genome-wide association study [52] . SGL accepts a collection of pathways as input and induces sparsity at both the pathway and the gene level [53] . Perhaps the closest two methods to the approach presented here are Network-Induced Classification Kernels ( NICK ) [26] and Network-Constrained Regularization by Li & Li [55] , which use the Laplacian of a gene interaction graph to force neighboring genes to have similar weights . As described below , both methods can be seen as a special case of our proposed framework . Methods that take advantage of gene interaction data in training are often closely tied to the underlying choice of a predictor . This places a severe limitation on the scope of prediction tasks ( e . g . , classification ) that can be addressed , and generalization to other tasks ( e . g . , regression ) may be difficult . We propose using model regularization to integrate feature relationship information , extendable to a large spectrum of supervised and unsupervised prediction tasks . In doing so , we bridge the gap between the freedom in choosing the underlying predictor and the ability of that predictor to utilize domain knowledge .
To make the learning problem in Eq ( 1 ) well-defined , the regularizer must be bound from below . This translates to the requirement that dj ≥ 0 , ∀j , and P must be a positive semi-definite matrix . The latter is satisfied by any kernel matrix , allowing one to directly apply the wealth of kernels defined in the literature , with one caveat . Kernels are often treated as measures of similarity [29] . However , P in Eq ( 2 ) drives the L2 penalty term and should , therefore , align with a measure of dissimilarity . Intuitively , we would like to penalize high values of wi and wj if features i and j are , in some sense , dissimilar . Such a penalty is more conducive towards finding correlated features , mimicking the behavior of the traditional elastic nets [8] . For this reason , we advocate using the pseudo-inverse of a kernel matrix as a choice for P , not the kernel matrix itself . The pseudo-inverse maintains the positive semi-definite property , while being more in line with the aforementioned intuition . To further motivate the use of the pseudo-inverse , consider the following example . When dealing with gene interaction networks , we may be interested in assigning similar weights to genes that are close together on the network [26 , 55] . Given an adjacency matrix A for a graph , we formulate the following regularizer: R ( w ) = λ 2 2 1 2 ∑ i ∑ j ( w i - w j ) 2 A i j ( 3 ) This is closely related to graph embedding [30] , where a graph structure is imposed over a set of samples and one seeks to reduce data dimensionality in a way that preserves node proximity in the lower-dimensional space . The regularizer in Eq ( 3 ) can be simplified to R ( w ) = λ 2 2 w T L w , ( 4 ) where L is the graph Laplacian [31] . This is a GELnet with P = L and d = 0 . The spectral decomposition of the Laplacian constitutes a Hilbert space , while its pseudo-inverse , L+ , is the reproducing kernel of that space [32] . By now , the connection should be clear: L+ is a kernel matrix that captures similarity between features using their proximity on a graph , while its pseudo-inverse , ( L+ ) + = L , appears in the L2 regularizer to penalize high weights of distant features in the graph . Lavi , et al . use a similar intuition to develop a method called Network-Induced Classification Kernels ( NICK ) for SVMs [26] . Rather than using L directly , the authors formulate an L2 regularizer around a linear combination of the Laplacian with the identity matrix: ( I+βL ) for some β ≥ 0 . In their method , the parameter β provides a trade-off between graph-driven regularization and the traditional ridge regression penalty of the SVMs . The NICK method can be seen as a special case of the framework proposed in this paper , where a GELnet with P = ( I + βL ) and d = 0 regularizes the hinge loss of the SVM . Likewise , Li & Li use the graph Laplacian to solve a set of regression tasks [55] . Their method can be seen as a special case of our framework , where a GELnet with P = L and d = 1 regularizes the squared-error loss . One drawback of the graph Laplacian is that it characterizes a node’s immediate neighborhood only , which may be inadequate for some applications . A natural extension beyond immediate adjacency is the diffusion kernel [33] . The kernel arises from a simulated physical process , where “heat” is applied to one node in the graph and the “temperature” is measured in another node , after the heat is allowed to diffuse along the graph edges . If the two nodes are localized to the same subgraph , this heat-based similarity measure will be high . When dealing with gene interaction networks , such subgraphs may correspond to genetic pathways , motivating the use of the diffusion kernel in place of L+ to discover the underlying molecular mechanisms . This intuition lies behind other diffusion-based methods , such as HotNet [20] . If D is a diffusion kernel , computed as a matrix exponential of the graph Laplacian , and I is the identity matrix , then setting the penalty matrix P = I − D will correctly assign lower penalty to “hot” pairs of nodes . Note that because all eigenvalues of D lie in [0 , 1] , P is positive semi-definite . While most of our attention has been given to gene interaction networks , we reiterate that GELnets are more general and can accommodate any positive semi-definite measure of dissimilarity between pairs of features . For example , we may be interested in grouping features together according to some predetermined factor and defining P in such a way as to penalize selection of feature pairs that do not belong to the same group . This is closely related to group LASSO [34] , where the regularization penalty behaves as LASSO for predefined groups of variables , and as ridge regression for individual variables within those groups . Group LASSO is limited by its inability to identify and remove noisy variables within a particular group , without excluding the entire group altogether . This limitation is overcome by GELnets , where the LASSO penalty is assigned to individual variables . We discuss how to solve Eq ( 1 ) for a specific form of the loss function and then show how several of the common learning problems can be expressed using this loss . The presented loss function arises directly from standard regression and is defined by the weighted sum of squared residuals . Consider the problem min L = min w 1 2 n ∑ i = 1 n a i y i - ( w T ϕ ( x i ) + b ) 2 + λ 1 ∑ j = 1 p d j | w j | + λ 2 2 w T P w , ( 5 ) where ( xi , yi ) i=1n is the training data , ai are the sample weights , and dj , P encode domain-specific information regarding feature importance and association . Note that we will generally use i to iterate over the samples and j to iterate over the features . We solve the problem in Eq ( 5 ) through cyclic coordinate descent by changing one wk at a time , while keeping the values of wj fixed for all j ≠ k[35] . The coordinate descent methods have been recently growing in popularity , giving rise to libraries like glmnet[36] and LIBLINEAR[37] . Their primary advantage is the fact that objective functions with a single variable can be solved in closed-form , leading to simple update rules and efficient implementations . Friedman and Hastie demonstrated that this can sometimes lead to ten-fold decreases in run time over the more traditional optimization methods for linear models [35] . For notational convenience , we define y i ( k ) = ∑ j ≠ k w j ϕ j ( x i ) + b , which is the prediction for sample i , made by the model when feature k is excluded . Our goal is to find the value of wk that minimizes the remaining residual . We solve the subproblem min L k = min w k 1 2 n ∑ i a i y i - y i ( k ) - w k ϕ k ( x i ) 2 + λ 1 d k | w k | + ∑ j ≠ k d j | w j | + λ 2 2 P k k w k 2 + 2 ∑ j ≠ k P k j w j w k + ∑ j ≠ k ∑ j ′ ≠ k P j j ′ w j w j ′ by taking a partial derivative with respect to wk and setting it equal to zero: ∂ L k ∂ w k = - 1 n ∑ i a i ϕ k ( x i ) y i - y i ( k ) - w k ϕ k ( x i ) + λ 1 d k ∂ | w k | ∂ w k + λ 2 P k k w k + ∑ j ≠ k P k j w j = 0 . This results in the following update rule for wk: w k ← S 1 n ∑ i = 1 n a i ϕ k ( x i ) y i - y i ( k ) - λ 2 ∑ j ≠ k P k j w j , λ 1 d k 1 n ∑ i = 1 n a i ϕ k ( x i ) 2 + λ 2 P k k , ( 6 ) where S ( v , γ ) = sgn ( v ) ( |v| − γ ) + is the soft-threshold operator that “snaps” values within γ of zero to be exactly zero [35] . The soft-threshold operator contributes greatly to faster run times when the LASSO penalty coefficient λ1 is not zero . The reason for this speedup is the fact that a wk that was previously “snapped” to be exactly zero will remain at zero , unless its absolute value exceeds λ1 dk . When wk is zero , it makes no contribution to partial fits y i ( j ) for all other j ≠ k . Thus , if the value of wk remains at zero , no updates to y i ( j ) are required , allowing those quantities to be cached . Higher values of λ1 will therefore lead to both sparser solutions and faster convergence times . Similar to the weight updates above , we can differentiate the objective with respect to the bias term . The derivative is given by ∂ L ∂ b = - 1 n ∑ i a i y i - ( w T ϕ ( x i ) + b ) , ( 7 ) which leads to the following update rule for b: b ← ∑ i a i y i - w T ϕ ( x i ) ∑ i a i . ( 8 ) Note that this is a simple weighted average of the residuals . Using Eqs ( 6 ) and ( 8 ) , we can derive an upper bound on the “meaningful” values of the λ1 meta-parameter . Specifically , by initializing all wk to zero and b to ∑ i a i y i ∑ i a i , setting λ1 to any value higher than λ 1 m a x = max j 1 n ∑ i a i ϕ j ( x i ) ( y i - b ) d j ( 9 ) guarantees that all wk will remain at zero and no updates will be made . The training procedure cycles through all the coordinates and the bias term until the desired stopping criterion is reached . In our experiments , we used both the number of iterations and the difference in the objective value between updates as the convergence criteria . For the latter , we terminated training whenever that difference fell below a certain threshold ϵ . We make the code available as an R package gelnet . Many other loss functions can be reduced to regression . In S1 Text , we review how this can be done for several popular methods via Taylor-series expansion [36 , 47 , 48] . Our review also shows how to handle loss functions that are non-covex ratios of quadratic norms ( such as Principal Component Analysis [50] and Linear Discriminant Analysis [49] ) using a method developed by Witten and Tibshirani [51] . We begin with experiments on synthetic data to investigate a key question: under what circumstances does the prior information about the gene regulatory network help prediction performance ? To answer this question , we generate synthetic data from predefined gene-gene relationships and then compare the performance of classical elastic nets to GELnets , where the gene interactions are provided to the latter via the penalty matrix P . As we show below , GELnets are able to correctly utilize such prior information to achieve better accuracy . For synthetic data experiments , we consider a randomly-generated scale-free graph . The associated adjacency matrix A has an entry of 1 if two nodes share an edge in the graph and an entry of 0 otherwise . We begin the experiment by using A to generate the “true” weight vector w . The goal of w is to simulate a signaling pathway , whose activity contributes to the phenotypic observations . The prediction task then aims to uncover this pathway from the observable data . To simulate a signaling pathway , we select a connected subcomponent of our scale-free graph via a random walk . The walk is terminated when 10% of the nodes are selected . The feature weights wj are set to 1 for j in the selected set and to 0 for all other nodes . We simulate gene expression data from a multivariate normal distribution: X ∼ N ( 0 , S ) , where we consider two scenarios for specifying the covariance matrix S . In the first scenario , a random covariance matrix is used . This creates a decoupling between the simulated expression data X and the simulated signaling pathway w , modeling the negative control case in which the observable data has no relationship to the gene regulatory network . The second scenario assumes that the feature covariance structure is dictated by the graph adjacency matrix A . A Gaussian Graphical Model ( GGM ) is used , with S selected such that S−1 closely approximates A[38] . The GGM imposes a coupling between X and w that models a biological scenario , where the observable phenotype is driven by a small number of genomic correlates belonging to the same signaling pathway , while the rest of the simulated transcriptome is expressed according to the regulatory relationships encoded by A . To simulate a typical high-dimensional low-sample scenario found in biological applications , we made use of a 5000-by-5000 adjacency matrix and generated 50 samples for each of the two scenarios above . The observable phenotypic response in all cases was computed as y = wT X . Note that because the data dimensionality vastly exceeds the number of samples , the problem of reconstructing the signaling pathway w from gene expression X and phenotypic observations y is under-determined . The simulated data ( X , y ) defines a regression problem , to which we apply the classical Elastic Nets and the GELnets , comparing the two regularization schemes . We provide additional information about feature-feature relationships to GELnets through the penalty matrix P , as in Eq ( 2 ) . All individual penalty weights dj are left at 1 . 0 . To evaluate how performance is affected by prior knowledge in the form of feature-feature relationships , we consider two distinct choices for P . The first choice encapsulates the true information via the normalized Laplacian of the graph adjacency matrix A . For the second choice of P , we investigate the effect of providing the “wrong” information to the GELnets , by using the normalized Laplacian of another graph adjacency matrix A′ . This matrix A′ is constructed by randomly permuting the columns ( and , to subsequently maintain symmetry , rows ) of A; such a permutation operator maintains the overall structure of the graph , while scrambling the individual feature-feature relationships . We introduced two ways to generate the data matrix X and two ways to specify the feature-feature penalty matrix P . Together , this setup gives rise to four scenarios; we refer to these as Rand+ , Rand- , GGM+ , and GGM- , where the prefix specifies whether the data is generated with a random covariance matrix ( Rand ) or via a GGM ( GGM ) , and the suffix denotes whether the normalized Laplacian is computed over the true adjacency matrix ( + ) or the permuted one ( − ) . Fig 1 summarizes how the four scenarios differ from each other . As we show below , the relative performance of Elastic Nets and GELnets can vary drastically from one scenario to the next . The performance was evaluated in a leave-pair-out cross-validation ( LPOCV ) setting , due to its tendency to yield less bias in performance estimation [5] . We focus on three specific performance metrics: The first performance measure captures how accurately we are able to recover the original signaling pathway that gave rise to the observable phenotypic response . Note that in real applications , the true weight vector w is unknown , and the reconstruction error is therefore not directly observable . RMSE is a standard performance metric for regression problems , capturing the deviation between predicted and observed response . Dispersion measures the degree to which features found to be predictive are near one another in network space . The metric we use arises directly from the L2-norm regularization term and acts as a positive control . Specifically , when the set of nodes Z is completely disconnected , the corresponding normalized Laplacian is the identity matrix and dispersion is equal to 1 . Conversely , for every edge that appears between the nodes in Z , the corresponding off-diagonal entry in the normalized Laplacian will be negative , resulting in a lower dispersion value . This is a positive control , because GELnet regularization directly minimizes dispersion in its L2-norm term . Consequently , we expect dispersion to always be lower in the GELnet models , compared to their Elastic Net counterparts . To address the question of meta-parameter selection , we average performance measures across a grid of meta-parameter values to obtain a marginalized estimate . Specifically , we iterate λ2 over { 10 , 000 , 1 , 000 , 100 , 10 , 1 } for both regularization schemes . For the Elastic Nets , we also iterate λ1 over { λ 1 m a x 27 , λ 1 m a x 9 , λ 1 m a x 3 } , where λ 1 m a x is defined in Eq ( 9 ) . Basing the choice of λ1 off λ 1 m a x allows us to consider models of varying sparsity . The values of λ1 for the GELnet models were specified such that the number of non-zero feature weights equaled the corresponding Elastic Net models to allow for a fair comparison . Based on our preliminary experiments with parameter tuning , we found that such a grid covers a wide range of models . As discussed in the literature , marginalized performance estimates are useful for method comparison [4 , 39] . Note that while we marginalize over the meta-parameters , the performance estimates are still conditional on the training data , which effectively allows us to ask “which of the two methods yields better performance , given a particular training set ? ” . This is important as we are not claiming that GELnet regularization is universally better than classical Elastic Nets , nor should we expect it to be . Besides the “No Free Lunch” considerations [40] , we expect a given biological network to be relevant in a subset of prediction tasks . Thus , a key question is not whether GELnet pathway-based regularization is better , but under what conditions does it boost performance . Answering this question will help us properly utilize prior biological information to gain novel insight in bioinformatics applications .
Fig 2 presents the performance of both regularization schemes in the two scenarios where the data was generated with a GGM; the same network was used to simulate both the gene expression X and the signaling network w . The GELnets are able to more accurately recover the simulated signaling pathway w when the information about true feature-feature relationships is provided ( GGM+ case ) , and the inverse is true when the GELnets are given the scrambled relationship information ( GGM- case ) . The latter is explained by GELnets selecting features in close proximity on the scrambled network , which is unlikely to contain the connected subcomponent encoded by w . As a positive control , we note that all GELnet solutions have lower dispersion than the corresponding Elastic Net models , when evaluated on the network provided to the GELnets . This indicates that the feature-feature penalties are working as intended . Fig 3 demonstrates that the improvement in dispersion is more pronounced in the GGM+ case . As expected , a higher improvement in RMSE is observed when the GELnets are provided with the correct network . Surprisingly , GELnet regularization consistently led to better RMSE , regardless of whether the method was given the true or the scrambled network . We speculated that this was due to the effect of the false network sharing some neighbors and paths as the true network . To test this idea , GELnet-based models were retrained with increasingly scrambled information about true feature-feature relationships . To compose a partially-scrambled network , we randomly permuted a fraction of rows ( and symmetrically columns ) in the graph adjacency matrix , before using the graph’s Laplacian to train a GELnet model . We refer to this fraction as the “Scramble Factor” and present the results from 100 runs of the experiment in S6 Fig . Note that the left-hand side of the plots , where the GELnets are provided with the true network , corresponds to the GGM+ case . Likewise , the right-hand side , where the entire network is scrambled , is the GGM- case . From S6 Fig , we observe that GELnets maintain their performance edge over Elastic Nets in the presence of up to 20% noise in the feature-feature relationship network . We also consider the performance of the two regularization methods on data generated with a random covariance matrix , with results presented in S1 and S2 Figs . As in the GGM case , providing the true network to GELnets leads to reconstruction improvement over Elastic Nets . Likewise , GELnets always yield better dispersion values than Elastic Nets , indicating once again that the feature-feature penalties are working as intended . Unlike in the GGM scenarios , both regularization schemes produce comparable RMSE values and , as depicted in S2 Fig , there is little to no distinction between Rand+ and Rand- cases . This evidence suggests that GELnets gain no benefit from the true gene-gene network when the network captures the signaling pathway that gave rise to the observed phenotypic response but not the expression data . When viewed together , these synthetic data results allow us to reason about the relevance of prior information to the application of a given dataset . Specifically , by training a model regularized with the GELnet , we are not only able to extract pathway-aligned features , but to also estimate how well those pathways represent the underlying biological mechanism that gave rise to the observed phenotype . We do this by comparing the performance to a model regularized by the classical Elastic Nets . We compare the ability of GELnets to make use of prior biological information relative to other regularization schemes by repeating our synthetic data experiments with Sparse Group LASSO ( SGL ) [53] . SGL takes as input a grouping of features and induces sparsity at both the group level and the individual feature level . By providing the method with a set of pathways , SGL can be readily used in bioinformatics applications , as has been done by Silver , et al . , who identified pathways and genes associated with high-density lipoprotein cholesterol in a genome-wide association study [52] . We present the comparison of SGL and GELnets in S7 and S8 Figs . Similar to our application of GELnets , we marginalize the performance of SGL across a range of its parameter values using the R package SGL . Because SGL requires a collection of pathways rather a single graph , we apply community-based clustering ( R package igraph ) to split up the synthetic network before providing it to SGL . In S7 Fig , it can be observed that SGL produces a better fit to the GGM-generated data , as measured by RMSE . However , GELnets produce more tightly-clustered solutions ( lower dispersion ) that better capture the simulated signaling pathway w ( lower reconstruction error ) , suggesting that SGL overfits the data in this situation . A similar trend with dispersion and reconstruction error can be observed in the Rand+/- scenarios ( S8 Fig ) , except that both regularization schemes produce comparable fits to the data , as measured by RMSE . We also evaluated the use of a diffusion kernel ( specifically I − D , as described in the Methods section ) as an alternative penalty matrix for the GGM+/- scenarios . S10 Fig demonstrates that models trained using the diffusion kernel penalty yield vastly lower dispersion than the Laplacian-based models . Additionally , models regularized by the diffusion kernel have lower RMSE ( particularly in the GGM- scenario ) , but the reconstruction error is slightly worse , suggestion minor overfitting to the training data . Because the Laplacian penalty matrix produces models with lower reconstruction error , we chose to use it when building models of drug sensitivity in Gray Cell lines , which we discuss in the following section . The GELNet can always be applied as a standalone regularization method; it provides as good a fit to the data in terms of RMSE as Elastic Net . However , the underlying network identified may or may not be related to the true underlying mechanism . By comparing the performance to Elastic Nets , we are able to identify the situations in which the network improves modeling accuracy . From these experiments and comparing the results of Fig 3 with S2 Fig we find that using an appreciable improvement in either RMSE ( at least 2 . 5% ) or dispersion ( at least 5% ) give an indication that a network model is relevant . However , to maintain a conservative interpretation , we suggest using both RMSE and dispersion as criteria for identifying when an underlying network is consistent with the data observations . We trained linear regression models to predict drug sensitivity in breast cancer cell lines [41] , comparing the performance of classical Elastic Nets to GELnets . In light of our results in the previous section , we reason that when GELnet regularization outperforms Elastic Nets by a large margin , it is evidence that the mechanism of resistance and the gene expression data are both captured by the same gene regulatory network ( GRN ) . Note that because the reconstruction error is not directly observable , we have to rely on RMSE and dispersion to determine whether this network corresponds with the one provided to GELnets . Fig 3 shows that providing to GELnets the network used to generate the data ( scenario GGM+ ) yields higher improvement in RMSE and lower dispersion over Elastic Nets compared to when the “wrong” network is provided ( scenario GGM- ) . It is important to note that even though GELnets always attains lower dispersion than Elastic Nets , our simulations reveal that there is information in the relative difference in RMSE and dispersion between the two methods . As revealed in the simulation experiments above , if the level of difference in performance between GELnets and Elastic Nets exceeds critical levels , it strongly suggests the pathway model is applicable to the learning task . Specifically , we use the difference in performance as an indicator of the network prior relevance . Thus , we aim to identify drugs for which we observe the highest improvement in RMSE and dispersion over Elastic Nets . The dataset by Heiser , et al . ( available for download from the supplement of [41] ) is comprised of RNAseq expression assays of 54 breast cancer cell lines and their sensitivity profiles to 74 compounds . The RNAseq data contains expression values for 18 , 632 genes . The sensitivity is measured as −log10 ( GI50 ) , where GI50 is the amount of compound needed to inhibit cell growth by 50% . For every drug , we trained two linear regression models , one regularized by an Elastic Net and another by a GELnet . We used the same grid of values for the λ1 and λ2 meta-parameters as in the synthetic data experiments , and provided all GELnet models with an interaction network from Pathway Commons ( http://www . pathwaycommons . org/ ) [42] , reducing the feature space of the dataset to 9 , 984 genes that occur in the network . Fig 4 presents the results for 27 of the 74 drugs where we observed lower RMSE values in GELnet models . The figure presents improvement in RMSE values over Elastic Net models , with the raw RMSE values shown in S5 Fig . GELnet models for 47 of the 74 drugs failed to provide an improvement in RMSE over Elastic Nets , suggesting that PathwayCommons is unable to accurately capture the underlying mechanism of drug resistance . Additionally , we note that GELnet models for all 74 drugs had lower dispersion values compared to their Elastic Net counterparts , as expected . As in the case of synthetic data experiments , the values presented in Fig 4 are averages over the grid of meta-parameter values . In the cases of BIBW2992 and 5-FdUR , GELnets outperformed Elastic Nets for all values of the λ1 and λ2 meta-parameters . Many of the drugs used in breast cancer were developed to target specific subtypes that have clear expression signatures ( e . g . luminals versus basals versus HER2-amplified ) . We indicate drugs whose sensitivity profiles correlate significantly with breast cancer subtypes ( subtype calls specified by Heiser , et al . [41] ) with diamond shapes in Fig 4 . For these drugs , we have to be mindful of the fact that the models of sensitivity are likely to be confounded by the subtype . Note that the two drugs , BIBW2992 and 5-FdUR , where GELnets outperformed Elastic Nets under all parameter settings do not fall into this category ( single-factor ANOVA; p-values greater than 0 . 05 ) . We note that while we expected BIBW2992 to be specific to the HER2 subtype , the sensitivity spectrum across the cell lines does indeed seem to sensitize additional lines w/o the amplification . While we observe the most consistent improvement in RMSE for both BIBW2992 and 5-FdUR , the GELnet models for BIBW2992 also yield the largest reduction in dispersion over Elastic Nets . A large improvement in both performance metrics suggests that BIBW2992 falls into what we called the GGM+ scenario in our synthetic data experiments: the expression data and the mechanism of resistance are both captured by the network that is provided to GELnets . We further tested this intuition by training 30 GELnet models for BIBW2992 sensitivity using randomly scrambled versions of the PathwayCommons network . S9 Fig presents the distribution of performance values for these models . We observe a substantial decrease in RMSE and dispersion relative to when the unscrambled version of the network is used , providing further support that PathwayCommons captures the underlying mechanism of resistance . We now take a closer look at the solutions obtained by GELnets to investigate potential novel mechanisms of resistance to BIBW2992 . BIBW2992 ( also known as Afatinib ) is an inhibitor of kinases from the epidermal growth factor receptor family , specifically EGFR and ERBB2 ( Her2 ) . It acts by covalently binding to and irreversibly blocking the receptors , thereby shutting down the signaling networks whose deregulation is commonly known to be implicated in epithelial cancer growth and proliferation [43] . Consequently , one expects that higher expression of these receptor genes will lead to higher sensitivity to the inhibitor . This is indeed one of the trends we observe . Fig 5 presents the GELnet models trained to predict BIBW2992 sensitivity across a grid of λ1 and λ2 meta-parameter values . The models are sorted by their improvement in RMSE over the Elastic Net equivalents . For each model , we show the feature weights for 30 genes that had the highest median rank across all meta-parameter values , where the ranking was according to the absolute values of the weights . As expected , higher expression of the HER2 gene is correlated with higher sensitivity to BIBW2992 . This is demonstrated by the relatively high negative model weights of ERBB2 and GRB7 , a positional neighbor of ERBB2 on the chromosome and often co-amplified and co-expressed . Furthermore , a positive model weight for FGFR2 and GABRA3 suggests that cells resistant to BIBW2992 may be responding to alternative stimuli ( the fibroblast growth factor and GABAA signaling ) from these overexpressed receptors . In support of this observation , the overexpression of FGFR2 has been previously observed in cells resistant to Lapatinib , another Her2 inhibitor [44] . Azuma , et al . speculated that FGFR2-targeted therapy may provide a promising salvage strategy after Lapatinib failure [44] , and our findings here suggest that the same may hold true for BIBW2992 as well . Note that the above trend of model weights for cell surface receptors is observed on the right-hand side of Fig 5 only , where Elastic Nets and GELnets perform comparably . The part of the figure is also associated with the lower values of λ2 , implying that there is little distinction between the GELnet and Elastic Net models . Indeed , the correlation between BIBW2992 sensitivity and the expression of the cell surface receptors above is also found by the Elastic Net regularization . As the value of the λ2 meta-parameter increases , Elastic Net and GELnet models begin to diverge and a new trend emerges on the left-hand side of Fig 5 , which is associated with a higher improvement in prediction accuracy of GELnet-regularized models over those of Elastic net . Importantly , the GELnet models emphasize an entirely different set of genes for predicting BIBW2992 sensitivity . These models identify the expression of KLF7 and its transcriptional co-activator FBXO38 as predictors of resistance . KLF7 is a transcription factor that was recently shown to play a regulatory role in differentiation of several cell lineages , including neuronal and osteocytic [45] . Its role in breast cancer is largely unknown , but the gene’s regulation of Map2 , NGF and TrkA suggests an involvement in cell proliferation and renewal . Note that while GELnets demonstrate the largest improvement in performance over Elastic Nets for high values of λ2 , the model with the lowest RMSE was obtained when the parameters were set to λ 1 = λ 1 m a x / 9 and λ2 = 1 . The latter appears on the right-hand side of Fig 5 , where the model of resistance is dominated by the cell surface receptors . These results demonstrate that the most accurate model does not necessarily recapitulate the entire biological story , and further exploration of the parameter space can produce additional insight . We present the two mechanisms of resistance in S3 and S4 Figs , as well as their interaction on a gene regulatory network in Fig 6 . Taken together , our findings in this section suggest that cells resistant to BIBW2992 might have undergone partial transdifferentiation , as indicated by the active KLF7 transcription factor and overexpressed fibroblast growth factor and GABAA receptors . This hypothesis is further supported by a very strong signal of SCGB2A2 and SCGB1D2 being downregulated in resistant cells , as indicated by their large negative weights in the GELnet models . The two genes are considered to be highly specific markers for the breast tissue , where their proteins form a covalent complex [46] . Further experimental validation is required to confirm the transdifferentiation hypothesis . Because KLF7 appears to play a central role in these transdifferentiated cells , the observation may suggest shRNA-mediated silencing of this transcription factor to get around resistance to BIBW2992 . All of the prediction problems we considered so far are supervised methods . To illustrate the generality of the GELnet regularization framework , we sought to apply it to an unsupervised task as well . We constructed a regularized Principal Component decomposition of the TCGA “PanCan12” dataset representing RNA-Seq data from twelve different types of cancer . The technical details of this problem can be found in S1 Text where we discuss non-convex ratios of quadratic norms . We downloaded the data from the Synapse TCGA_Pancancer repository ( https://www . synapse . org/# ! Synapse:syn300013 ) . For each principal component , we constructed two GELnet models . The first model used the Laplacian of PathwayCommons as its penalty matrix , as in the previous experiments . For the second model , we set the penalty matrix as P = I − D , where D is the diffusion kernel of PathwayCommons and I is the identity matrix . Our intuition is that , by capturing indirect gene connectivity , the diffusion kernel will produce models that more tightly cluster on the corresponding interaction network . The empirical results presented in Fig 7 confirm this intuition . We projected the PanCan12 dataset onto the first two unregularized principal components and estimated the quality of GELnet models constructed for those two components . Specifically , we measured performance of GELnet models according to how well they approximate the original , unregularized principal components ( measured via RMSE ) and by how tightly-clustered the solutions are on the PathwayCommons network ( measured via the dispersion metric ) . We considered the same grid of values for the λ1 and λ2 meta-parameters . Note that for both principal components , the use of the diffusion kernel produced models with lower dispersion while maintaining the same level of reconstruction accuracy compared to the Laplacian . We looked at the genes picked up by the diffusion-driven models and found enrichment for many pathways associated with organism development and tissue differentiation , confirming the findings of the TCGA consortium that found that cell-of-origin signatures drive the dominant information in the data [54] . The top 10 genes selected by each model are shown in S1 Table . The Gene Ontology ( GO ) enrichment analysis revealed a nearly identical set of GO terms enriched in both models , with the following terms appearing at the top: ECTODERM_DEVELOPMENT ( GO:0007398 ) , TISSUE_DEVELOPMENT ( GO:0009888 ) and ORGAN_DEVELOPMENT ( GO:0048513 ) . The close similarity of the enriched terms between the two models is expected , because the enrichment analysis acts as a “smoothing function” on the Laplacian-based solution effectively elucidating the same set of pathways as those found by the diffusion-based solution .
In molecular biology , genetic interactions provide a rich source of information encoding what is known about cellular circuitry . The proposed GELnet regularization method capitalizes on this information to improve the accuracy and interpretability of linear regression-based solutions to genome-based prediction tasks . The novel regularization scheme allows the use of domain knowledge to guide the selection of related features to steer toward intuitive solutions . Because our knowledge about genetic pathways is incomplete , we expect this new framework to be applicable only in situations where current knowledge aligns adequately with the underlying biological mechanism . Obviously , this information is usually not available; the puzzle is to determine if and when such genetic pathway representations are indeed relevant for a particular study . We have shown here , through a series of simulation experiments , how to identify such situations . We demonstrated that GELnets outperform their non-pathway-based counterpart , Elastic Nets , when both the dataset and the phenotype are simulated from the same genetic network , and where GELnet regularization is provided with that network . Importantly , we found critical levels in the relative difference between the methods in accuracy in prediction and the mutual closeness of features on the networks to indicate when the network used for simulation matches the network used for modeling . We describe how one can use this observation to detect when drug resistance mechanisms might be inferred from regression models . In a panel of breast cancer cell lines , we show that both expected and novel mechanisms are revealed for over one third of the drugs tested in the cell line panel . One such case is the model for response to the dual EGFR/ERBB2 inhibitor , BIBW2992 . Consistent with the known drug action , we find over-expression of ERBB2 and GRB7 are sensitivity markers . In addition , concrete receptors regulating parallel growth response pathways , such as FGFR2 , are revealed as resistance mechanisms that may provide synergistic targets . Our approach is general enough to extend to other machine learning problems , where a comparable regularization scheme can be introduced to reward and/or penalize the selection of features based on their mutual proximity within a genetic pathway diagram . Applications include linear and non-linear approaches for supervised , unsupervised , and semi-supervised strategies ( see S1 Text ) . Although not considered here , the use of feature-specific weights dj in Eq ( 2 ) can be used to further guide feature selection by placing more or less penalty on individual features . Other graph-based penalty matrices in place of the Laplacian can also be used .
|
The low costs of sequencing and other high-throughput technologies have made available large amounts of data to address molecular biology problems . However , often this means thousands of measurements , for example on gene expression , are assayed for a much smaller number of samples . The imbalance complicates the identification of genes that generalize to new samples and in finding a collection of genes that suggest a theme for interpreting the data . Pathway and network-based approaches have proven their worth in these situations . They force solutions onto known biology and they produce more robust predictors . In this manuscript , we describe a new formulation of statistical learning approaches that naturally incorporates gene-gene relationships , like those found in gene network databases . The theory we present helps unify and codify an explicit formulation for gene pathway-informed machine-learning that should have wide reach .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"infographics",
"medicine",
"and",
"health",
"sciences",
"genetic",
"networks",
"breast",
"tumors",
"cancers",
"and",
"neoplasms",
"random",
"variables",
"covariance",
"oncology",
"mathematics",
"pharmaceutics",
"network",
"analysis",
"genome",
"analysis",
"pharmacology",
"computer",
"and",
"information",
"sciences",
"gene",
"expression",
"breast",
"cancer",
"probability",
"theory",
"drug",
"synthesis",
"data",
"visualization",
"gene",
"regulatory",
"networks",
"graphs",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"drug",
"interactions",
"pharmaceutical",
"processing",
"technology",
"computational",
"biology"
] |
2016
|
Pathway-Based Genomics Prediction using Generalized Elastic Net
|
Host encounters with viruses lead to an innate immune response that must be rapid and broadly targeted but also tightly regulated to avoid the detrimental effects of unregulated interferon expression . Viral stimulation of host negative regulatory mechanisms is an alternate method of suppressing the host innate immune response . We examined three key mediators of the innate immune response: NF-KB , STAT1 and STAT2 during HCV infection in order to investigate the paradoxical induction of an innate immune response by HCV despite a multitude of mechanisms combating the host response . During infection , we find that all three are repressed only in HCV infected cells but not in uninfected bystander cells , both in vivo in chimeric mouse livers and in cultured Huh7 . 5 cells after IFNα treatment . We show here that HCV and Flaviviruses suppress the innate immune response by upregulation of PDLIM2 , independent of the host interferon response . We show PDLIM2 is an E3 ubiquitin ligase that also acts to stimulate nuclear degradation of STAT2 . Interferon dependent relocalization of STAT1/2 to the nucleus leads to PDLIM2 ubiquitination of STAT2 but not STAT1 and the proteasome-dependent degradation of STAT2 , predominantly within the nucleus . CRISPR/Cas9 knockout of PDLIM2 results in increased levels of STAT2 following IFNα treatment , retention of STAT2 within the nucleus of HCV infected cells after IFNα stimulation , increased interferon response , and increased resistance to infection by several flaviviruses , indicating that PDLIM2 is a global regulator of the interferon response .
Host cells sense the presence of an infecting virus using pathogen pattern recognition receptors which initiate signaling pathways that converge on the activation of NF-κB and the IRF family of transcription factors leading to the production and secretion of cytokines , type I ( IFNα and IFNβ ) and type III ( IFN-λ ) interferons [1–5] . Binding of type I and type III IFN to receptors on the infected cell and neighbouring cells leads to nuclear translocation of STAT1/STAT2/IRF9 ( ISGF3 ) [6] and the transcription of numerous IFN stimulated genes ( ISGs ) which in turn inhibit viral replication [7–9] . Cooperation between the NF-κB and ISGF3 pathways is required to induce an innate immune response that controls a variety of pathogens [9–12] . However , because of the wide variety of biologic functions regulated during the innate immune response , continuous induction of the anti-viral state is detrimental to normal cell function [13 , 14] , and fine control of the antiviral response is required . Host cells have developed a multitude of mechanisms to shut down the interferon response and viruses have evolved mechanisms to prematurely activate these shut down pathways to enhance their own replication [15] . Canonical interferon signaling initiated by IFN-α or IFN-λ induces dimerization of their receptors , IFNAR1 and IFNAR2 or IFNLR1 and IL-10R2 respectively , activating the associated protein kinases janus kinase1 ( JAK1 ) and tyrosine kinase2 ( TYK2 ) . The major substrates of these kinases are signal transducer and activator of transcription 1 and 2 ( STAT1 and STAT2 ) . Phosphorylated STAT1 and STAT2 bind IRF9 to form the transcription factor ISGF3 , which migrates to the nucleus to induce ISG transcription [13] . However , it has also been shown that STAT2 can homodimerize and stimulate ISG expression independently [16] , and while STAT1 is essential for IFN-γ and IFN-λ signaling , STAT2 is essential for both IFN-λ and IFN-α signaling in Huh7 . 5 cells [17] . In addition , unlike STAT1 knockdown , STAT2 knockdown resulted in elevation of HCV levels in stem cell derived hepatocytes [18] . Control of STAT signaling involves numerous post translational modifications: acetylation [19] , methylation [20 , 21] , and ISGylation [22 , 23] promote signaling , while dephosphorylation [24] or sumoylation [25] inhibit it . Nuclear ubiquitination of STAT followed by degradation by the proteasome is another mechanism to shut down STAT mediated transcription [26–28] . Ubiquitination of key signaling proteins and their subsequent degradation in the proteasome has emerged as a method for shutting down a variety of processes including the innate immune response [29–31] . Ubiquitin E3 ligases are the elements of the ubiquitination pathway responsible for substrate specificity . PDLIM2 is one such ubiquitin E3 ligase . Its known targets include NF-κB [32 , 33] , as well as STAT1 , STAT3 , and STAT4 in mouse cells or when overexpressed from plasmids [28 , 34 , 35] . In the case of NF-κB it was shown that the active form of the transcription factor is targeted for degradation in insoluble proteasome complexes within the nucleus . It has been shown that HCV infected patients have an ongoing innate interferon response in their livers and that a very similar response develops in the chimeric livers of HCV infected SCID/Alb-uPA mice in the absence of an adaptive immune system [36–40] . The host interferon response persists despite a variety of mechanisms that HCV uses to inhibit the interferon response [41] . Studies in cell culture revealed RIGI and downstream IRF3 activation occurs early in infection and subsequently controlled by HCV NS3/4A later in infection [42] . Consistently , studies of patient biopsies revealed that the ISG response originated in HCV infected cells and was also high in neighboring bystander cells [43] . In mice where HCV was expressed in liver cells , the STAT response was suppressed . When overexpressed both HCV core and NS5a reduced STAT1 phosphorylation and interferon signaling [44 , 45] . HCV core also stimulates protein phosphatase 2A to inhibit IFN-α signaling but this can be overcome by high concentrations of IFN-α [46 , 47] . Surprisingly , STAT1 was not essential for inhibition of HCV by IFN-α [17] . We previously examined NF-κB protein levels [48] in the chimeric livers of HCV infected SCID/Alb-uPA mice and found that NF-kB levels were decreased in HCV infected cells but elevated in surrounding uninfected human hepatocytes . This implied that HCV can blunt innate immunity in infected cells and was the first indication that uninfected bystander cells , not infected cells , were responsible for the global interferon response seen in vivo [48] . Here we extend our studies to the mechanisms by which HCV and flaviviruses limit innate immunity . In these studies we focused on the effects of viral infection on the levels and cellular location of two key signal transduction molecules mentioned above , STAT1 and STAT2 . During HCV infection of chimeric mouse livers , HCV infected cells have reduced levels of STAT1 and STAT2 compared to bystander cells , and during interferon treatment HCV infected Huh7 . 5 cells have reduced levels of STAT1 and STAT2 compared to bystander cells . The reduction in levels of nuclear of STAT2 , but not STAT1 , is blocked by the proteasome inhibitor MG132 implicating the ubiquitination pathway in this process . Infection by HCV and flaviviruses strongly induces transcription of the ubiquitin E3 ligase PDLIM2 . Knockout of the PDLIM2 gene results in increased levels of STAT2 after IFNα treatment , the retention of STAT2 in the nucleus of HCV infected cells , a decrease in viral replication , and a more robust interferon response .
We have previously shown that the interferon response to HCV in chimeric livers of SCID/Alb-uPA mice was similar to that found in liver biopsies of HCV infected patients [40] . We , and others , have shown that despite having several known mechanisms to combat the host interferon response , HCV infection still induces numerous ISGs [49–53] . To address this apparent paradox , we examined how the levels and locations of key mediators of the innate immune response , STAT1 and STAT2 and NF-κB , changed in response to HCV infection in the livers of chimeric mice using confocal microscopy . Comparison of immunostained liver sections from uninfected mice to those from infected mice showed that the overall levels of STAT1 , STAT2 , and NF-κB trended higher in HCV infected mice ( Figs 1A , S1 and S2 ) . Comparison of STAT1 and STAT2 protein levels in uninfected livers and infected livers using confocal microscopy revealed that STAT1 levels increased 1 . 7±1 . 3 fold in infected livers while STAT2 increased 1 . 5±0 . 45 fold in infected livers . While this was not significant it was consistent with the known induction of an interferon response by HCV in the tissue [40 , 48] as well as Huh7 . 5 cells ( S1 Table ) . However , within infected livers , when we compared the levels of STAT1 or STAT 2 in HCV infected cells ( red arrows ) with the levels in uninfected cells ( white arrows ) we saw reduced levels of STAT1 and STAT2 in the HCV infected cells . Quantification revealed that STAT1 and STAT2 levels in infected cells were approximately half that in uninfected cells within an infected liver ( Figs 1A and S1B ) . The same phenomenon was seen with NF-κB expression in infected chimeric liver tissue ( S2 Fig ) , consistent with our previous results [48] . These results indicate that upregulation of STAT signaling and a concomitant interferon response and induction of ISGs ( S1 Table and [40] ) occurs in bystander cells but HCV infected cells were able to thwart STAT signaling . This is consistent with an inverse correlation between the ISG , IFITM3 and HCV RNA in individual cells of HCV patient livers isolated using laser dissection microscopy [54] , and also consistent with the notion that the stimulus for ISG expression originates in HCV infected cells [43] . To further investigate the mechanism leading to decreased levels of STAT1 and STAT2 in HCV infected cells in vivo , we extended our observations to a HCV cell culture system . We confirmed previous findings [55] showing ISG induction during HCV JFH-1 infection of Huh7 . 5 cells ( S1 Table ) . The mRNA of 37/39 ISGs examined increased their expression at least 2 fold during HCV infection . To examine the effect of HCV infection on STAT protein translocation we treated Huh7 . 5 cells with exogenous interferon and examined STAT1 and STAT2 levels by confocal microscopy . Cells were treated with interferon for either 15 min or 12h then stained using antibodies specific for STAT1 or STAT2 ( green ) and HCV core ( red-Fig 1B and 1C ) . Prior to IFNα treatment , both STATs were present at low levels predominantly in the cytoplasm of both uninfected and HCV infected cells; after 15 minutes of IFNα treatment , the majority of STAT1 and STAT2 migrated to the cell nuclei and there was little difference in either protein levels or localization between uninfected and HCV infected cells , indicating that HCV is not blocking nuclear translocation . After 12h of IFNα treatment there was a notable increase in the levels of both STAT proteins . In cells not exposed to HCV both STATs were primarily cytoplasmic after 12h of IFNα treatment . In cells exposed to HCV there was a clear difference between infected cells and uninfected bystander cells . Infected cells are stained red for HCV core protein , marked by red arrows and outlined in white . Quantitation of the STAT levels revealed that infected cells had low overall levels of both STATs , similar to those found in cells not treated with IFNα . However , during interferon treatment uninfected cells had 3 fold higher levels of STAT proteins than HCV infected cells . These features of STAT1 and STAT2 localization resemble that seen in the livers of chimeric mice with chronic HCV infection: low levels of cytoplasmic STATs in infected cells and higher levels of STATs in bystander human hepatocytes , indicating that the global interferon response noted in transcriptomics arises from uninfected bystander cells . Low levels of STAT1 and STAT2 seen in HCV infected cells may be the result of either a lack of induction or increased degradation . If degradation is involved then STAT protein levels should increase upon treatment with a proteasome inhibitor such as MG132 . We included MG132 during 12h IFNα treatment of uninfected and HCV infected cells and then examined STAT1 and STAT2 levels and location ( Fig 2A and 2B ) . Treatment with MG132 alone does not change either localization or amount of STAT1 or STAT2 . A small induction of STAT1 was caused by IFNα in the presence of MG132 . In contrast to STAT1 , inclusion of MG132 during IFNα treatment resulted in significantly increased levels of STAT2 in infected cells indicating that STAT2 is degraded following IFNα treatment . Even more remarkable is the localization of the STAT proteins . Compared to what is seen when cells were treated with IFNα alone ( Fig 1B and 1C ) , treatment with IFNα and MG132 for 12h caused retention of both STAT1 ( Fig 2A ) and STAT2 ( Fig 2B ) in the nucleus of infected cells . A similar analysis of the levels of STAT and phospho-STAT was consistent with our confocal analysis ( S3 Fig ) . First , HCV infection led to a decrease in STAT2 but not STAT1 protein levels , indicating HCV infection affects STAT2 levels ( S3A Fig ) . Treatment of cells with IFNα induced phosphorylation of both STAT1 and STAT2 , and additional treatment with the proteasome inhibitor MG132 increased the levels of phospho-STAT1 and phospho-STAT2 , consistent with nuclear localization . The levels of total STAT1 remained constant after IFNα treatment; however , total STAT2 levels declined during interferon treatment , and this was prevented by MG132 . The levels of total STAT2 were also lower in infected and IFNα treated cells ( S3B Fig ) . The levels of HCV core protein did not change significantly during interferon treatment ( S3C Fig ) . Similar experiments monitoring the localization of NF-κB after stimulation with LPS/IL-1β also showed that MG132 treatment led to nuclear retention of NF-κB ( S4 Fig ) [33 , 56] . Since STAT1 and STAT2 protein levels are themselves induced by IFNα , to more clearly examine whether they were degraded during IFNα treatment we performed a time course of IFNα treatment in the presence of the protein synthesis inhibitor cycloheximide ( Fig 2C and 2D ) . Total cell lysates from uninfected or HCV infected Huh7 . 5 cells were prepared after IFNα treatment , and STAT1 and STAT2 levels were measured by western blot analysis . The levels of STAT1 remained constant over the course of the experiment , and did not change when the proteasome inhibitor MG132 was included during the experiment , indicating STAT1 was not degraded in the absence of new protein synthesis . The levels of STAT2 declined in the presence of IFNα in both uninfected and HCV infected cells . The decrease in STAT2 levels was dependent on the proteasome since it was prevented by MG132 treatment . To examine the sub-cellular compartment where degradation occurred , we fractionated cells into cytoplasmic and nuclear fractions after 12h of IFNα treatment again in the presence of cycloheximide to prevent new protein synthesis ( Fig 2D ) . STAT1 was significantly more abundant in the nucleus than the cytoplasm . Consistent with time course experiments , STAT1 levels did not change upon MG132 treatment . An apparent molecular weight increase was observed in the presence of MG132 which may indicate hyper-phosphorylation STAT1 [57] , [20] . Unlike STAT1 , when proteasomal degradation was inhibited by MG132 , STAT2 levels were elevated . This effect was most dramatic in the nuclear fraction , showing that STAT2 but not STAT1 is degraded predominantly in the nucleus following interferon treatment . Taken together these results may indicate inhibition of STAT2 degradation by MG132 also results in nuclear retention of STAT1 , its partner in ISGF3 . The E3 ubiquitin ligase PDLIM2 has been shown to be one factor capable of shutting down NF-κB , STAT1 , STAT3 , and STAT4 signaling . It can ubiquitinate these substrates and target them for degradation in the proteasome [28 , 32–35 , 58] . Since we had originally observed low levels of STAT1 , STAT2 and NF-κB in the livers of chimeric mice , we performed similar experiments examining the localization of NF-κB . Stimulation of Huh7 . 5 cells with IL-1β and LPS led to the nuclear translocation of NF-κB ( S4 Fig ) . HCV infected cells contained less NF-κB than surrounding bystander cells after 7h of LPS/IL-1β stimulation . Taken together these results suggested that a common factor was involved in diminishing STAT and NF-κB activity in HCV infected hepatocytes . We therefore examined the levels of PDLIM2 mRNA in HCV JFH-1 infected Huh7 . 5 cells by quantitative RT-PCR ( qRT-PCR ) . In time course experiments we found that PDLIM2 mRNA levels were 13 fold higher in Huh7 . 5 cells 4 days after infection , when viral titers peaked , than in uninfected cells ( Fig 3A ) . In dose response experiments , we found PDLIM2 transcript levels correlated with increasing amounts of HCV JFH-1 virus used to infect Huh7 . 5 cells ( Fig 3E ) . We also examined PDLIM2 protein levels after infection with HCV ( Fig 3B ) and found that PDLIM2 increased in the nucleus during infection , consistent with its location and action during NF-KB and HTLV Tax degradation [33 , 59] . Interestingly PDLIM2 sequestration in the cytosol by association with the cytoskeleton is associated with increased activity of NF-kB [32 , 60] , and HCV and flaviviruses are known to disrupt the cytoskeleton [61–65] . Since degradation of STAT2 has been observed in cells infected with other flaviviruses [66–68] , we also examined PDLIM2 expression in Huh7 . 5 cells infected with Zika virus or Dengue virus ( Fig 3C and 3D ) . Expression of PDLIM2 mRNA increased 14 fold after Zika virus infection , and was dependent on the amount of initial inoculum ( Fig 3C and 3E ) . Dengue virus infection had the most dramatic effect on PDLIM2 expression with a 20 fold increase in PDLIM2 expression soon after infection followed by a rapid decline , again in a MOI dependent manner ( Fig 3D and 3E ) . Representative images showing infection at different MOI with each of HCV , ZIKV , and DENV are shown in S5 Fig . The increase in PDLIM2 levels in response to a variety of infections implied that PDLIM2 may be an interferon response gene . To examine whether PDLIM2 is an ISG , we treated Huh7 . 5 cells with 1000 IU/mL of IFNα2 for 2h and 7h ( Fig 3F ) . Transcripts for the known ISGs IRF9 , IFIT2 , ISG15 , OASL , and MDA5 increased strongly , as expected . In contrast , PDLIM2 did not change , nor did the non-ISG’s PCNA and HSPA5 indicating that PDLIM2 is not an ISG . In separate experiments PDLIM2 levels also did not change after induction with either IFNα or IFNλ for 24h ( Fig 3F ) . Together , these results indicate viral infection leads to induction of PDLIM2 , a non-ISG which targets key signal transduction intermediates such as STATs and NF-κB for degradation , and may be a common mechanism whereby flaviviruses inhibit the host interferon response . Since STAT2 was degraded predominantly in the nucleus and PDLIM2 is a ubiquitin E3 ligase , known to degrade NF-κB p65 in the nucleus , we wanted to assess whether STAT2 associates with ubiquitin , where this association occurs and if PDLIM2 is involved in this reaction . We used proximity ligation assays ( PLA ) to localize the interaction between STAT proteins and ubiquitin . The PLA fluorescent signal is generated if the 2 proteins that the primary antibodies recognize ( in this case either STAT1 or STAT2 , and ubiquitin ) are within 40 nm [69] . We treated Huh7 . 5 cells with IFN-α for 15 min or 7h , and then examined whether either STAT1 or STAT2 were associated with ubiquitin using PLA ( Fig 4A ) . To assess background , we omitted each primary antibody independently , and the highest background was found when the ubiquitin antibody was omitted . Therefore , the overall fluorescence in the Cy3 channel in a control omitting anti-ubiquitin antibodies ( FK2 ) from the assay was used as background . The Cy3 fluorescent signal produced by PLA in untreated cells was approximately 3 fold higher than that in cells where the primary antibody was omitted . Ubiquitin and STAT1 were associated throughout the cell , and this association did not change location upon IFNα or MG132 treatment . Quantification indicated that there was a small but significant increase in the association between ubiquitin and STAT1 in the nucleus after 7h of IFNα treatment , likely due to an overall increase in the amount of STAT1 caused by IFNα . However , the STAT1/ubiquitin interaction ( PLA positive signal ) did not reflect the nuclear re-localization of STAT1 either after 15 min of treatment with IFNα , or when MG132 was included during IFNα treatment . In contrast , the interaction between STAT2 and ubiquitin was much more dynamic and reflective of STAT2 localization . Immediately after IFNα treatment , STAT2-ubiquitin association was strikingly nuclear , and by 7h of treatment STAT2-ubiquitin was noticeably absent from the nuclei and increased in the cytoplasm . Consistent with proteasomal degradation of poly-ubiquitinated proteins , inhibition of the proteasome by MG132 resulted in continued STAT2/Ub association in the nucleus . The pattern of STAT2/ubiquitin interaction reflected the pattern of nuclear re-localization of STAT2 seen in response to IFNα treatment . We also performed STAT2 immunoprecipitation to determine if ubiquitin associates with STAT2 when Huh7 . 5 cells are treated with IFNα in the presence and absence of MG132 for 2h ( S6A Fig ) . The levels of STAT2 increased upon IFNα treatment and STAT2 was not precipitated if anti-STAT2 antibodies were omitted . Immunoblotting of STAT2 immunoprecipitate with anti-ubiquitin antibodies revealed that the levels of polyubiquitined STAT2 increased after treatment with both IFNα and MG132 . Polyubiquitin was not precipitated if anti-STAT2 antibodies were omitted . To examine whether HCV infection increases association between STAT2 and ubiquitin within the nuclear matrix , similar to p65 , uninfected and HCV infected cells were left untreated , or treated with IFNα for 15 min , then treated and fixed as in Tanaka et al . [33] prior to performing a PLA using anti-STAT2 and anti-ubiquitin antibodies ( S6B Fig ) . There was greater association between STAT2 and ubiquitin in the nucleus of HCV infected cells both when untreated or treated with IFNα . Taken together these results are consistent with ubiquitin-dependent STAT2 degradation primarily in the nucleus induced by IFNα treatment . We next examined whether STAT1 or STAT2 interacted with PDLIM2 after IFNα treatment . We expressed FLAG-tagged PDLIM2 in Huh7 . 5 cells , treated the cells with IFNα , and examined the interaction of either STAT1 or STAT2 with the tagged PDLIM2 by PLA ( Fig 4B ) . If the anti-FLAG antibody was omitted from the PLA assay there was little background fluorescent signal . Transfected cells ( green ) expressing tagged PDLIM were identified using anti-FLAG antibodies after the PLA reaction ( yellow ) . There was minimal interaction between STAT1 and tagged PDLIM2 detected by the proximity ligation assay . Again , minimal interaction could be detected between tagged PDLIM2 and STAT2 in untreated cells or in cells treated with IFNα for 7h . However , inhibition of the proteasome with MG132 increased the interaction between STAT2 and PDLIM2 significantly , indicating that inhibition of degradation of STAT2 increased its association with PDLIM2 ( Fig 4B ) . Quantification of PLA indicated that interactions between STAT2 and PDLIM2 increased greater than 2 fold upon interferon treatment in the presence of MG132 whereas the interactions between STAT1 and PDLIM2 , while significant , were much lower . These results indicate that interaction between PDLIM2 and STAT2 are increased by IFNα treatment . To further assess the interaction between STAT2 and PDLIM2 , STAT2 and PDLIM2-FLAG were expressed in Huh7 . 5 cells followed by STAT2 immuno-precipitation , and western blotting with anti-FLAG antibodies ( Fig 4C ) . Interactions were detected between STAT2 and PDLIM2-FLAG even in untreated cells , however , treatment with IFNα and MG132 to prevent degradation resulted in increased immuno-precipitation of PDLIM2-FLAG by STAT2 antibodies . Omission of STAT2 antibodies resulted in no PDLIM2-FLAG signal . Taken together these results indicate that IFNα increases the interaction between ubiquitin and STAT2 and also between PDLIM2 and STAT2; however , it does not change STAT1 interactions to the same extent . We postulate that the increased interactions between STAT2 and PDLIM2 lead to ubiquitination of STAT2 and its degradation . To determine whether PDLIM2 is a regulator of STAT2 we made a knockout of PDLIM2 in Huh7 . 5 cells using the CRISPR/CAS9 system . The knockout is an 800 bp deletion , starting in the first exon of PDLIM2 , which changes the reading frame of all PDLIM2 transcripts . Any putative translation products would be truncated and lack the LIM domain needed for ubiquitin ligase activity ( S7 Fig ) . The absence of PDLIM2 expression was confirmed by western blot ( S7 Fig ) . The effects of knocking out PDLIM2 expression are expected to be pleomorphic , as a number of targets for its ubiquitin ligase activity have been proposed [28 , 32–35 , 58] . Here we have focused on its involvement in the degradation of STAT2 protein in the nucleus of HCV infected cells ( Fig 5 ) . We treated either Huh7 . 5 cells or PDLIM K/O cells with IFNα and examined the levels of STAT2 . In untreated cells , STAT2 levels were lower in Huh7 . 5 cells than in PDLIM K/O cells . However , treatment with IFNα led to greater increases of STAT2 in PDLIM K/O cells than in Huh7 . 5 cells , ultimately resulting in more STAT2 overall after 24h ( Fig 5A ) . To determine whether PDLIM2 was responsible for STAT2/ubiquitin association , we examined IFNα dependent STAT2/ubiquitin association in the nuclear matrix of Huh7 . 5 and PDLIM K/O cells using PLA ( Fig 5B ) . Huh7 . 5 or PDLIM K/O cells were treated with IFNα for either 15 min or 7h to stimulate STAT2 nuclear translocation and degradation . MG132 was included during the 7h incubation to prevent degradation . Ubiquitin was associated with STAT2 in the nuclear matrix of Huh7 . 5 cells after both 15 min and 7h , whereas the STAT2/ubiquitin association was significantly less in PDLIM K/O cells . We also examined whether PDLIM2 knockout resulted in nuclear retention of STAT2 , similar to what we saw when cells were treated with IFNα while inhibiting the proteasome with MG132 in Huh7 . 5 cells ( Fig 2B ) . We infected both parental Huh7 . 5 and PDLIM2 K/O cells with HCV and then treated with IFNα for 7h to stimulate STAT2 nuclear translocation ( Fig 5C ) . In wild type Huh7 . 5 cells little STAT2 was seen in the nuclei of infected or uninfected cells as we saw in Fig 1B . However , in cells lacking PDLIM2 , STAT2 staining was predominantly nuclear in both infected and uninfected cells . This is consistent with what we saw when we inhibited STAT2 degradation during IFNα treatment using MG132 ( Fig 2B ) ; however , no MG132 was used . Quantification revealed increased STAT2 in the nucleus of PDLIM K/O cells . Taken together these results indicate that PDLIM2 knockout results in decreased ubiquitination of STAT2 , an increase in STAT2 levels following IFNα treatment and nuclear retention of STAT2 during IFNα treatment . This indicates that in addition to its previously elucidated functions PDLIM2 directs the ubiquitination and degradation of STAT2 . Knockout of PDLIM2 is predicted to have a number of effects because it has many targets that are involved in stimulating the interferon response . We therefore compared the interferon response in the knockout cells to that in Huh7 . 5 cells using quantitative RT-PCR for a set of ISGs that are induced by HCV infection . For each ISG we examined the response to IFNα at 6 times: untreated , the initial response at 2h and 7h post treatment , then the “so called” refractory interferon response [70 , 71] . To induce the refractory response , cells were treated with IFNα for 12h followed by 12h without treatment , and then re-inducing with IFNα for 2h and 7h . These responses are shown as the clusters of six bars for each ISG in Fig 6 . We displayed ISGs in groups: those involved in initial intracellular signaling by the innate immune response ( Fig 6A–6B ) , inhibitors ( Fig 6C ) of the innate immune response , extracellular signals ( Fig 6D–6F ) , and effectors ( Fig 6G–6H ) of the innate immune response . In general , there was a greater induction of ISG mRNA in the PDLIM2 knockout cells ( red bars ) than in Huh7 . 5 cells ( white bars ) , with the exception of IFIT3 . There was relatively little induction of , or difference between , Huh7 . 5 and PDLIM2 K/O cells seen in expression of genes involved in initiation of the innate response or inhibitors ( Fig 6A–6C ) , and some changes were observed in ISGs involved in extracellular signaling ( Fig 6D–6F ) . Most notably , the down regulation of IFNA2 and IFNB mRNA that occurs when exogenous IFNα is added was markedly less in PDLIM2 K/O cells ( Fig 6D ) . The greatest differences between Huh7 . 5 cells and PDLIM2 K/O cells were in interferon effector genes ( Fig 6G and 6H ) . In some specific genes such as XAF1 , OAS3 , IFI6 , MX1 , and HLA-A the refractoriness of IFNα signaling was abolished in the PDLIM2 K/O cells; however , in most cases the refractory pattern of expression where the expression of a given ISG was lower after the second IFNα stimulation remained approximately the same . When PDLIM2 K/O cells were first transfected with a PDLIM2 overexpression plasmid then treated with IFNα for 7h , the ISG expression was decreased when compared with PDLIM2 K/O cells transfected with the equivalent vector ( S7E Fig ) . These results indicate that overall the response to IFNα is greater in PDLIM2 K/O cells than in parental Huh7 . 5 cells . To evaluate whether increased ISG expression in PDLIM2 K/O cells led to a greater antiviral effect , we examined the susceptibility of these cells to vesicular stomatitis virus ( VSV ) infection because this virus is extremely sensitive to inhibition by the innate immune response [72] . Productive VSV infection leads to cell lysis . Both Huh7 . 5 cells and PDLIM2 K/O cells were infected with VSV and the extent of cell lysis 24h after infection was used as a measure of productive infection . PDLIM2 K/O cells were significantly more resistant to VSV infection than the parental Huh7 . 5 cells ( Fig 6I ) . This suggests that the generally increased interferon response detected in our survey of ISGs in PDLIM2 K/O cells is significant enough to inhibit VSV infection . HCV induces a significant interferon response during infection ( S1 Table ) that it counters by multiple mechanisms . The stronger innate immune response in PDLIM2 K/O cells that we demonstrated above would be predicted to make these cells less susceptible to HCV infection . We therefore investigated the ability of HCV to infect PDLIM2 K/O and Huh7 . 5 cells ( Fig 7A ) . When infected with identical amounts of virus PDLIM2 K/O cells expressed less HCV core protein and in fewer cells than did Huh7 . 5 cells when examined by immunofluorescence microscopy ( Fig 7A , upper panel ) . They also had less intracellular HCV RNA and secreted approximately 30 fold less HCV when infected with 1 genome/cell in time course experiments ( Fig 7A , middle panel ) . In a limiting dilution assay ( Fig 7A , lower panel ) , the TCID 50 was lower when determined using Huh7 . 5 cells than in PDLIM2 K/O cells . Stated differently , it took more virus to infect a similar number of PDLIM2 K/O cells than Huh7 . 5 cells . Although HCV and HAV are both hepatotropic , HAV infection induces a more limited type I interferon response in host cells than does HCV infection [73 , 74] . We therefore examined whether PDLIM2 K/O cells are more resistant to HAV infection ( Fig 7B ) than parental Huh7 . 5 cells . Infection with equal amounts of HAV resulted in less HAV core expression , less HAV RNA in cells , and less HAV in the supernatant , indicating that HAV can be controlled by an increased interferon response as expected ( Fig 7B ) . However , it should be noted that the difference in HAV levels between PDLIM2 K/O cells and Huh7 . 5 cells was smaller than the difference in HCV levels produced in the same cells . Since we have shown that infection of Huh7 . 5 cells by ZIKV or DENV increased expression of PDLIM2 ( Fig 3 ) and both viruses are known to degrade STAT2 [67 , 68 , 75] , we examined whether PDLIM2 K/O cells were more resistant to ZIKV or DENV infections . We infected PDLIM2 K/O and parental Huh7 . 5 cells with several different MOIs of ZIKV . We found that PDLIM2 K/O cells had lower levels of intracellular Zika virus RNA and secreted less Zika virus into culture supernatants than did Huh7 . 5 cells particularly at low MOI ( Fig 7C ) . In a similar way , we examined whether the PDLIM2 K/O cells were more resistant to DENV infection by infecting both Huh7 . 5 and PDLIM2 K/O cells with two different MOI . After 36h of infection we stained the cells using DENV capsid specific antibodies and quantified the number of infected cells ( Fig 7D ) . We found that fewer PDLIM2 K/O cells were infected by an equivalent MOI , indicating that cells lacking PDLIM2 are more resistant to both ZIKV and DENV infections . Taken together these results indicate that PDLIM2 is a negative regulator of the interferon response and may be a common target used by a number of flaviviruses to combat the host antiviral response .
The host antiviral response is a complex and highly regulated process , which must be rapid to create a profoundly antiviral state within the infected cell and in surrounding cells under threat of infection . Induction of multiple cytokines during infection is key to resistance to infection . Cross talk between the type I interferon and NF-κB pathways are required for resistance to lethal ectormelia virus infections , and improves response to HCV and hepatitis E virus infection [10 , 76] . Consistently , a recent report shows that STAT and NF-κB pathways cooperate by concerted recruitment of the mediator complex to interferon responsive promoters [12] . This coordinated host antiviral response must also be capable of rapid shut down to avoid the detrimental effects of the antiviral state on normal cell function . Prolonged activation of pro-inflammatory pathways is associated with inflammatory and autoimmune diseases [13] as well as the development and proliferation of multiple forms of cancer [77–79] . Inhibition of the NF-κB and STAT pathways can suppress the growth of a number of tumors [80–83] . Given their importance in the innate antiviral response and the connection between HCV and hepatocellular carcinoma , we examined the expression of several key mediators of the host antiviral response , STAT1 , STAT2 , and NF-κB . We found that all 3 were lower in HCV infected hepatocytes compared with surrounding bystander cells , both in vivo in chimeric mouse livers and in Huh7 . 5 cells , indicating that the interferon response seen during HCV infection is due to uninfected or newly infected hepatocytes [42] . This is consistent with an inverse correlation between HCV RNA and an ISG , IFITM3 , found in individual liver cells [54] . We investigated whether lower levels of STAT1 , STAT2 , and NF-κB in HCV infected cells were due to degradation . Previous work using mouse cells had shown that PDLIM2/Mystique/SLIM could direct the degradation of NF-κB , STAT1 , and STAT4 , and that overexpression of PDLIM2 in human cell lines targeted NF-κB p65 , STAT1 , and STAT4 for ubiquitination and degradation [33 , 34 , 84] . We found that addition of the proteasome inhibitor MG132 during either IFNα or LPS/IL-1β treatment led to activation and nuclear retention of STAT1 , STAT2 , or NF-κB . Since PDLIM2 was shown to act in the nucleus following NF-κB activation [33] , we investigated the effects of viral infection on PDLIM2 expression in the human hepatocyte cell line Huh7 . 5 . We found that HCV , Zika virus , and Dengue virus infections upregulated PDLIM2 in a dose and time dependent manner , but not in an interferon dependent manner . We next focused our investigations on the mechanism of STAT1 and STAT2 down-regulation in HCV infected cells in the context of interferon treatment . We found that following interferon treatment , STAT2 was degraded in the nucleus but there was little degradation of STAT1 . Consistent with the targeting of STAT2 for degradation by PDLIM2 , we found that interferon treatment led to association of PDLIM2 with STAT2 , but not with STAT1 , and interferon treatment led to the ubiquitination of STAT2 , but not STAT1 in the nucleus . These results indicate that STAT2 is a target for PDLIM2 in Huh7 . 5 cells while STAT1 is not , and that targeting STAT2 for degradation may be a common theme among Flaviviruses as has been shown for Zika virus and Dengue virus [66–68 , 75] and now HCV . Zika virus NS5 protein interacts with STAT2 to stimulate its degradation while DENV NS5 protein bridges the interaction between STAT2 and UBR4 to stimulate its degradation . We , and others , have shown that HCV NS5A interacts with STAT1 [44] . We speculate by virtue of the ISGF3 complex NS5A may also form a complex with STAT2 and therefore PDLIM2 . In our system , overexpression of PDLIM2 led to its association with STAT1 , which is consistent with previous reports of STAT1 association with PDLIM2 and its degradation [34 , 85] . However , the association between STAT1 and PDLIM2 did not respond to IFNα administration nor did we observe degradation of STAT1 in the presence of cycloheximide , which prevents new protein synthesis . We did observe a lack of induction of STAT1 protein during HCV infection possibly due to the interactions with HCV core [45] . Since STAT1 , STAT2 and likely PDLIM2 protein levels vary among cell types and during infection , we cannot rule out that PDLIM2 does not direct interferon independent degradation of STAT1 in other situations . Indeed , a number of IFNα inducible E3 ubiquitin ligases that are known to shut down type I interferon signaling have been reported [86–88] . However , a key component of many efficient and quickly inducible/repressible signaling systems is that of default repression [89] , and PDLIM2 is uniquely placed to fulfill this role in type I IFN signaling . Selective degradation of STAT2 has an additional implication for type I and type II interferon signaling . It has been recently shown that STAT2 inhibits type II interferon signaling [90] . We propose that initial shutdown of the type I interferon response is due to recognition of STAT2 by PDLIM2 , followed by ubiquitination , and degradation in a nuclear proteasome , while STAT1 is preserved . We observe that STAT1 is not degraded and is retained in the nucleus when STAT2 degradation is inhibited by MG132 . Since STAT1 homodimers bind to gamma activation site ( GAS ) promoter sequences in response to type II IFN ( IFN-γ ) stimulation , the preservation of STAT1 may prime the cells to respond to type II interferon as the type I response abates . This mechanism is consistent with previous reports of both the induction of a GAS binding factor and GAS promoter driven transcription by IFN-α and IFN-β [91 , 92] . Indeed in separate experiments , we observed that induction of a GAS reporter gene by interferon-γ is more pronounced in PDLIM2 K/O cells than in the parental Huh7 . 5 cells . Preservation of STAT1 and selective depletion of STAT2 by PDLIM2 during IFNα administration is also consistent with the shift in expression from uniquely IFN-α/β induced proteins at early times after IFN-α treatment , to proteins that are also induced by IFN-γ at later time points [93] . Given the complex activation of STAT proteins and modulation of the IFN response by additional cytokines [94 , 95] , more complex explanations cannot be ruled out . Viral promotion of PDLIM2 activity could be a mechanism for premature shut down of the innate immune response . The specific mechanisms used by these viruses to regulate PDLIM2 are not known; however , HCV , ZIKV , and DENV all disrupt the cytoskeleton and both HCV and ZIKV have been proposed to alter the epithelial to mesenchymal transition ( EMT ) [96–100] . It has been proposed that sequestration of PDLIM2 in the cytosol regulates its function in the nucleus , and that PDLIM2 regulates the EMT [32 , 101] . It is possible that disruption of the cytoskeleton during infection results in the nuclear localization that we showed in Fig 3B . Such an alteration would be expected to have pleiotropic effects because this E3 ubiquitin ligase’s known targets , NF-κB and STAT2 , and perhaps others occupy key positions in antiviral , proliferative , and apoptotic pathways . NF-κB activation predominantly leads to expression of extracellular cytokines while IFN signaling leads to the activation of intracellular factors [102] . We found that knockout of PDLIM2 and stimulation by IFNα led to activation of genes for both extracellular cytokines and receptors such as CXCL9 , 10 , HLA-A , and HLA-C , as well as intracellular effector ISGs such as MX-1 , OAS3 and ISG20 . Intracellular signaling molecules were not as affected . Consistent with an increase in the interferon response in PDLIM K/O cells , viral infection , by a variety of viruses including HCV , ZIKV , DENV and VSV , was attenuated . Taken together these results indicate that PDLIM2 is a key regulator of the innate immune response , and may be manipulated by a variety of viruses to promote their replication . The consequences of dysregulated PDLIM2 expression may range from inflammatory conditions , which during HCV infection , leads to fibrosis and cirrhosis , to the development of tumors in the affected cells [103–108] .
Experimental approval for mouse experiments came from the University of Alberta Animal Welfare Committee according to the Canadian Council on Animal Care guidelines . Study approval #00000348 . Prior to harvesting chimeric livers , animals were given isoflurane and put into the surgical plane prior to cervical dislocation . All mice were housed VAF and frozen human hepatocytes were purchased from Cellz Direct or ThermoFisher Scientific . Mice were transplanted and infected as described previously [40 , 48] . Mice were infected with HCV genotype 2a strain JFH-1 virus and had serum titers greater than 1x104 copies/ml . Huh7 . 5 cells ( Dr . Charles Rice ) were cultured in DMEM ( Sigma , D5796 ) with 10% FBS ( Sigma , F1051 ) . Tissue culture adapted JFH-1 was used for all HCV infections , HM175/p16 for HAV infections ( Stanley Lemon , University of North Carolina ) , and PLCal-ZIKV for Zika virus infections . ZIKV , VSV , and Dengue virus were provided by Tom Hobman [68] . Cells were infected for 4-6h at 40%-50% ( HCV , DENV ) , 85% ( VSV ) , or 90% ( HAV , ZIKV ) confluence and then washed 4x with media . For activation of NF-κB or STAT proteins , recombinant Human IL-1β ( PeproTech , 200-01B , 10 ng/ml ) , LPS ( Sigma , 10 ug/ml ) or IFNα2b ( Schering , 02238675 , 1000 IU/mL ) or IFN-λ3 ( R&D Systems , 5259IL-025 ) were diluted in media immediately before use . Cycloheximide ( Sigma , 7698 , 50 μM ) and MG132 ( Sigma , 2211 , 10 μM ) were added in conjunction as indicated . Commercially available antibodies were used to detect FK2 ( Enzo Life Sciences , BML-PW8810 ) , DDK tag ( Origene , TA50011 ) , HCV NS3 ( TORDJI-22 , Abcam ) , HCV core ( ThermoFisher Scientific , MA1-080 ) , HAV capsid ( Commonwealth Serum Laboratories , K24F2 ) , Dengue capsid ( Dr . Tom Hobman , University of Alberta ) , STAT1 ( Cell Signaling Technologies , CST9175 ) , STAT2 ( Santa Cruz , sc-476 ) , Y701-phosphoSTAT1 ( Cell Signaling Technologies , 9167S ) , Y690-phosphoSTAT2 ( Cell Signaling Technologies , 88410S ) , PDLIM2 ( Abcam , 246868 ) , p65 NF-κB ( Santa Cruz , sc-372 ) , Actin ( EMD Millipore , MAB1501 ) , PARP-1 ( BD Pharmingen , BD556362 ) , Lamin ( Zymed , 33–2000 ) , B-tubulin ( Abcam , AB6046 ) , Anti-NS5a ( gift from Charlie Rice , 9E10 ) . For western blotting , goat anti-mouse IR Dye 800 ( Licor , 926–32210 ) and goat anti-rabbit IR Dye 680 ( Licor , 926–32221 ) were used . For immunofluorescence , Alexa Fluor 546 goat anti-mouse IgG ( ThermoFisher Scientific , A11030 ) , Alexa Fluor 488 goat anti-rabbit IgG , ( ThermoFisher Scientific , A11008 ) , or Alexa Fluor 647 goat anti-mouse IgG ( ThermoFisher Scientific , A21236 ) were used . Custom primer/probe sets were used for HCV , HAV , and ZIKV ( S2 Table ) . TaqMan assays ( ThermoFisher Scientific ) were used for PDLIM2 ( Hs00917389_m1 ) , and HPRT ( Hs99999909_m1 ) . For the interferon response gene panel , a Gene Expression MicroFluidics Card with a custom Human ISG array from Applied Biosystems was used . See S2 Table for genes and assay numbers . Additionally , Primers and probes for CXCL9 , CXCL10 , IFI6 , IRF9 , MX1 , SOCS1 , and XAF ( S2 Table ) were used to assess ISG response using IDT Prime Time quantitative RT-PCR assays when PDLIM2 K/O cells were transfected with PDLIM2 . Cells were collected using Accutase ( Millipore , SCR005 ) and lysed in Cytoplasmic Extract buffer ( CE ) ( 10mM HEPES , 10mM KCl , 0 . 1mM EDTA , 0 . 1 mM EGTA pH8 , 0 . 1% Triton X-100 , Sigma Protease Inhibitor 11873580001 ) . After 40 minutes gentle rotation at 4°C , cytoplasmic fractions were separated from nuclear fractions by centrifugation at 1000x g for 5 minutes and then collected . Nuclear pellets were washed 4 times in CE buffer , then lysed with Nuclear Extract buffer ( 20mM HEPES , 25% glycerol , 0 . 5M NaCl , 1mM EDTA , 1mM EGTA , Sigma Protease Inhibitor ) . For phosphoSTAT western blots , cells were collected in Eppendorf tubes by scraping on ice in ice cold Radioimmunopercipitation assay buffer ( RIPA ) ( 50mM Tris pH 7 . 4 , 150mM Sodium Chloride , 0 . 5% Sodium Deoxycholate , 1% Nonidet P-40 ) supplemented with 1mM Sodium-Orthovanadate , 10mM Beta-Glycerophosphate , 50mM Sodium Fluoride , and 1x Protease Inhibitor Cocktail ( Sigma Aldrich ) . Samples were allowed to incubate at -20°C for 15 min , after which they were briefly vortexed and clarified at 14 , 000g for 10 min at 4°C . Supernatants were collected and 5x Protein Sample Buffer ( 250mM Tris-Cl pH6 . 8 , 5% SDS , 45% Glycerol , 9% Beta-Mercaptoethanol , 0 . 01% Bromophenol Blue ) was added to result in 1x samples for western blotting . For immunoprecipitation: Huh7 . 5 cells were lysed as in [33] and then incubated at 22°C with rabbit anti STAT2 antibodies for 5 . 5 h , then incubated with sheep anti-rabbit IgG Dynabeads ( Thermo , 11204D ) overnight at 4°C , followed by 8 washes with 50mM Tris-HCL pH 7 . 5 , 150mM NaCl , 0 . 05% Triton X-100 . Western blotting was performed as per standard methods [109] . Signal was detected with an Odyssey Infrared Imaging system ( Licor ) . Cells were transfected using Lipofectamine 2000 ( ThermoFisher Scientific , 11668019 ) according to manufacturer’s instructions . The GFP-tagged STAT2 construct and the DDK-tagged PDLIM2 construct was purchased from Origene ( RG208592 , accession number NM_005419 . 2 and RC210022 , accession number NM_021630 . 4 ) . The PDLIM2 construct corresponds to PDLIM2 transcript variant 2 . For comparison of ISG levels after 7h of IFNα treatment PDLIM2 K/O cells were transfected with the DDK tagged PDLIM2 construct or an equivalent construct expressing a GFP-GFP fusion protein 24h prior to IFNα treatment ( Viromer Red - Lipocalyx , VR-01LB-00 ) essentially according to the manufactures protocol except twice the recommended amount of plasmid was used . The control was a CMV expressed double GFP [110] . PDLIM2 K/O cells were generated from Huh7 . 5 cells . A homozygous deletion between positions 6914 and 7715 of the PDLIM2 gene ( RefSeq NG_030435 . 1 ) was generated by CRISPR-based gene editing following the protocol of Ran et al [111] . Two target sequences within the gene ( CCCTGGGGCTTCCGTATCAC and CATCAACGGGGAAAGCGCGG ) were each inserted into the vector pSpCas9 ( BB ) -2A-Puro v2 . 0 ( Addgene , PX459 ) . The two constructs were co-transfected into Huh7 . 5 cells using GenJet reagent ( II ) ( SL100489-HUH ) . After 3 days of selection in media containing 2 ug/mL puromycin , live cells were selected by flow cytometry . Serial dilutions of the cells were plated on mitomycin C-inactivated NIH-3TC feeder cells . An initial clone , C4 , was subcloned by rounds of serial dilution until a clone was obtained that , by PCR and sequence analysis , showed precise deletion of the sequences between the two predicted CRISPR-CAS9 cleavage sites in the PDLIM2 gene ( S4 Fig ) . The deletion alters the reading frame of all PDLIM2 transcripts such that putative translation products would be truncated near their N-termini; they would have a truncated PDZ domain implicated in protein:protein interactions and lack the LIM domain needed for ubiquitin E3 ligase activity . Huh7 . 5 cells grown on glass coverslips were fixed with 1:1 methanol: acetone at -20°C for a minimum of 30 minutes , then washed with PBS and blocked for 1hr in PBS with 1% bovine serum albumin ( Sigma , A3912 ) and 2 . 5 mM EDTA . For STAT1/STAT2 IF , cells were fixed with 4% paraformaldehyde for 40 min , permeabilized in 0 . 1% saponin in PBS for 40 min , blocked 1% BSA/2 . 5 mM EDTA/0 . 1% saponin for 1hr , and washed with PBS/0 . 1% saponin . Cells were incubated with primary and secondary antibodies ( see above ) , diluted in block for 1hr at room temperature . Following secondary antibody application , cells were mounted with DAPI Fluoromount-G ( Southern Biotech , 0100–20 ) , or incubated with 1/5000 Hoechst 33342 ( ThermoFisher Scientific ) in PBS for 5 minutes , then washed and mounted on slides with Vectashield Mounting Medium for Fluorescence ( H-1000 ) and sealed with nail polish . Confocal images were obtained using a Quorum Wave FX-2 spinning disk microscope using 20X/0 . 85 NA or 40x/1 . 3NA oil immersion lenses . Quantification was done using ImageJ ( NIH ) or Velocity 6 . 2 . 1 ( PerkinElmer ) . Brightfield photos were obtained using a Zeiss Axiovert 200M microscope with a Fluar 10x/0 . 5 NA lens . Samples were treated as described above for growing , fixation , blocking , and primary antibody incubation , except for coverslips prepared for nuclear matrices , which were first treated as described by Tanaka et al . [33] , then treated as above . Following primary antibody incubation , a Duolink proximity ligation assay kit ( Sigma , DUO92102 ) was used according to manufacturer’s instructions except that following the final wash coverslips were immediately mounted onto slides rather than being allowed to dry . Extracellular HAV and HCV RNA were purified using the Roche High Pure Viral Nucleic Acid kit ( 11858874001 ) as per manufacturer’s instructions . cDNA was synthesized with Superscript III ( ThermoFisher Scientific , 18080044 ) and random primers ( for HAV , ThermoFisher Scientific , 48190011 ) or a gene-specific primer for HCV ( HCV reverse primer , see above ) . Intracellular RNA was harvested using QIAzol ( Qiagen , 79306 ) according to manufacturer’s instructions , and cDNA was synthesized using M-MLV ( ThermoFisher Scientific , 28025013 ) . Prime Time ( IDT ) PCR assay mix was used with for qRT-PCR using primers listed in S2 Table . Extracellular ZIKV was measured by plaque assay on Vero cells ( Dr . Tom Hobman , University of Alberta ) . ZIKV virus was diluted in DMEM to the MOI indicated , and the media applied to Huh7 . 5 or PDLIM2 K/O cells . After 1h , media was removed , cells washed in DMEM , and fresh DMEM + FBS ( 10% ) applied . Supernatant was harvested after 1d , 2d , or 3d , and applied to Vero cells for 1h , after which media was removed , cells washed , and media containing 1 . 5% carboxymethylcellulose ( Sigma-Aldrich , 21902 ) added . After 4 days , paraformaldehyde ( PFA ) was added to the media to a final concentration of 5% PFA for 30 minutes . Following fixation , a 0 . 1% crystal violet solution was used to stain the plaques .
|
The response of cells to an invading pathogen must be swift and well controlled because of the detrimental effects of chronic inflammation . However , viruses often hijack host control mechanisms . HCV and flaviviruses are known to suppress the innate immune response in cells by a variety of mechanisms . This study clarifies and expands a specific cellular mechanism for global control of the antiviral response after the induction of interferon expression . It shows how several viruses hijack this control mechanism to suppress the innate interferon response .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"nuclear",
"staining",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"light",
"microscopy",
"viruses",
"rna",
"viruses",
"immunoprecipitation",
"microscopy",
"confocal",
"microscopy",
"stat",
"proteins",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"stat",
"signaling",
"immune",
"response",
"precipitation",
"techniques",
"biochemistry",
"signal",
"transduction",
"dapi",
"staining",
"cell",
"biology",
"flaviviruses",
"viral",
"pathogens",
"interferons",
"biology",
"and",
"life",
"sciences",
"cell",
"signaling",
"organisms",
"zika",
"virus"
] |
2019
|
HCV and flaviviruses hijack cellular mechanisms for nuclear STAT2 degradation: Up-regulation of PDLIM2 suppresses the innate immune response
|
Variation in gene expression is a fundamental aspect of human phenotypic variation . Several recent studies have analyzed gene expression levels in populations of different continental ancestry and reported population differences at a large number of genes . However , these differences could largely be due to non-genetic ( e . g . , environmental ) effects . Here , we analyze gene expression levels in African American cell lines , which differ from previously analyzed cell lines in that individuals from this population inherit variable proportions of two continental ancestries . We first relate gene expression levels in individual African Americans to their genome-wide proportion of European ancestry . The results provide strong evidence of a genetic contribution to expression differences between European and African populations , validating previous findings . Second , we infer local ancestry ( 0 , 1 , or 2 European chromosomes ) at each location in the genome and investigate the effects of ancestry proximal to the expressed gene ( cis ) versus ancestry elsewhere in the genome ( trans ) . Both effects are highly significant , and we estimate that 12±3% of all heritable variation in human gene expression is due to cis variants .
Admixed populations are uniquely useful for analyzing the genetic contribution to phenotypic differences among humans . Phenotypic differences that are observed among human populations may have systematic non-genetic causes , such as differences in environment [1] , [2] . However , in an admixed population such as African Americans , such differences are minimized and the only systematic differences among individuals are in the proportion of European ancestry , which can be accurately inferred using genetic data . Several recent epidemiological studies in African Americans have taken advantage of this , showing that many phenotypic traits vary with the proportion of European ancestry [3]–[5] . Here , we apply this idea to analyze population differences in gene expression . Gene expression is a fundamental determinant of cellular phenotypes , and understanding how gene expression variation is apportioned among human populations is an important aspect of biomedical research , as has been true for apportionment of human genetic variation at the DNA level [6] . Recently , four studies analyzed lymphoblastoid cell lines from HapMap samples and reported that a large number of expressed genes exhibit significant differences in gene expression among continental populations [7]–[10] . However , results of these studies may be affected by non-genetic factors such as differences in environment , differences in preparation of cell lines , or batch effects [2] , [9] , [11] , [12] . In particular , a recent review article has suggested that much of the expression variation across populations is caused by environmental factors [13] . On the other hand , analyses of expression differences that are correlated to ancestry within an admixed population are robust to all of these concerns . In this study , we analyzed lymphoblastoid cell lines from 89 African-American samples and investigated the relationship between expression levels of ∼4 , 200 genes and the proportion of European ancestry . We compared the results with those predicted from the differences in expression levels between 60 European samples ( CEU from the International HapMap Project ) and 60 African samples ( YRI from HapMap ) [6] . We confirmed the existence of heritable gene expression differences between CEU and YRI by showing a highly significant correspondence between observed CEU vs . YRI differences ( i . e . differences between sample means ) and the expression differences predicted by ancestry differences among African Americans . Notably , the correspondence holds regardless of whether differences between CEU and YRI are large or small . This suggests that the effects of heritable population differences on variation in gene expression are widespread across genes , mirroring population differences at the DNA level [6] . Heritable variation in gene expression may be due to cis or trans variants . Previous studies in humans have been successful in mapping both cis and trans effects , but the results they provide are far from complete , due to limited sample sizes [14] , [15] , [9] , [16]–[20] . In particular , the relative number of cis vs . trans associations that were reported varies widely across these studies , perhaps due to differences in power or choices of significance thresholds [13] . Thus , the overall extent of cis vs . trans regulatory variation in human gene expression has not yet been established . Here , by measuring how gene expression levels across all genes vary with local ancestry ( 0 , 1 or 2 European chromosomes ) either proximal to the expressed gene ( cis ) or elsewhere in the genome ( trans ) , we estimate that 12±3% of heritable variation in human gene expression is due to cis variants .
100 African-American ( AA ) samples from the Coriell HD100AA panel were genotyped on the Affymetrix SNP 6 . 0 GeneChip . Genotyping was conducted at the Coriell Genotyping and Microarray Center , and the genotype data was obtained from the NIGMS Human Genetic Cell Repository at Coriell ( see Web Resources ) . In addition , genotype data from 60 European ( CEU ) , 60 African ( YRI ) , 45 Chinese ( CHB ) and 44 Japanese ( JPT ) samples was obtained from Phase 2 HapMap [6] ( see Web Resources ) . We restricted all analyses to 595 , 964 autosomal markers with <5% missing data in AA samples and <5% missing data in Phase 2 HapMap samples , with A/T and C/G markers excluded so as to preclude any ambiguity in strand complementarity . Our analyses were not sensitive to the number of markers used . Two AA samples which we identified as cryptically related to other AA samples were excluded from the set of samples used for principal components analysis . Local ancestry ( 0 , 1 or 2 European chromosomes ) at each location in the genome was estimated for each AA sample using the HAPMIX program , a haplotype-based approach that has been shown to attain an r2 of 0 . 98 between inferred local ancestry and true local ancestry in simulated African-American data sets ( A . L . P . , N . P . , D . R . & S . M . , unpublished data; see Web Resources , specifically http://www . stats . ox . ac . uk/̃myers/software . html ) . The HAPMIX program inputs AA genotype data and phased CEU and YRI data from Phase II HapMap [6] , and outputs the estimated probability of 0 , 1 or 2 European chromosomes at each location in the genome . The weighted sum of these probabilities ( multiplied by 0 . 00 , 0 . 50 or 1 . 00 , respectively ) forms an estimate of local % European ancestry . Genome-wide ancestry was computed as the average of estimated local ancestry throughout the genome . Lymphoblastoid cell lines for 60 HapMap CEU , 60 HapMap YRI and the Coriell HD100AA samples were obtained from Coriell Cell Repositories ( see Web Resources ) . Gene expression was assayed using the Affymetrix Genome Focus Array , as described previously [7] . We restricted our analysis to the 4 , 197 genes on the array that are expressed in lymphoblastoid cell lines [7] . The gene expression data is publicly available ( GEO accession number GSE10824 ) ( see Web Resources ) . For HD100AA samples , we excluded two cryptically related samples ( see above ) , four samples identified as genetic outliers ( see Results ) , and five samples for which gene expression measurements were not obtained , so that 89 AA samples were included in gene expression analyses . For each gene g , we normalized gene expression measurements for CEU and YRI to have mean 0 and variance 1 across 120 CEU+YRI samples , and normalized gene expression measurements for AA by applying the same normalization for consistency . We implicitly assume an additive genetic model in which gene expression has genetic and non-genetic components , with part of the genetic component predicted by ancestry . Let egs denote normalized gene expression of gene g for sample ( i . e . individual ) s . Let θs denote the genome-wide European ancestry proportion of sample s , so that θs has value 1 for CEU samples and 0 for YRI samples as above , and fractional values for AA samples . We consider a model in which egs = agθs+νgs for CEU and YRI samples and egs = cagθs+νgs for AA samples , where c is a global parameter and νgs represents the residual contribution to gene expression that is not predicted by ancestry . Thus , the parameter c represents a validation coefficient measuring the aggregate extent to which the observed gene expression differences ag between CEU and YRI ( differences between sample means ) are heritable . We implemented two different approaches for fitting the parameters c and ag of this model: ( 1 ) Starting with the initial guess c = 1 , we alternated computing maximum likelihood estimates for ag ( for all g ) conditional on c , and computing a maximum likelihood estimate for c conditional on ag ( for all g ) , and iterated to convergence . In each case , the maximum likelihood estimates were obtained via linear regression ( with a separate linear regression for each g when estimating ag , and a single linear regression when estimating c ) . ( 2 ) For each g , we estimated values ãg , CEU+YRI by regressing egs against θs using CEU and YRI data only , and ãg , AA by regressing egs against θs using AA data only . We then regressed ãg , AA against ãg , CEU+YRI to obtain an estimate of c . In this computation , we scaled our estimates of ãg , CEU+YRI using the sampling error correction ξ ( described below in Computation of QST ) to remove the effect of sampling error on the denominator Σg ( ãg , CEU+YRI ) 2 of our estimate of c . ( On the other hand , we note that sampling noise in the AA data does not bias our computation of c , whose expected value does not change when noise is added to ãg , AA ) . We observed that approaches ( 1 ) and ( 2 ) produced identical estimates of c , indicating that both approaches are effective in finding the best fit to the model . We followed approach ( 2 ) to plot ãg , AA vs . ãg , CEU+YRI and to compute estimates of c specific to different values of |ãg , CEU+YRI| . We repeated the above computation using genotype data instead of gene expression data . We restricted the analysis to markers in which the average of CEU and YRI frequencies was between 0 . 05 and 0 . 95 . Although AA genotypes at each marker were used twice in this computation—both for estimating genome-wide ancestry using all markers and for measuring the effect of genome-wide ancestry on genotype at a specific marker—we note that with hundreds of thousands of markers , our estimate of genome-wide ancestry is negligibly impacted by data from a specific marker . We investigated the effects of cis ancestry and trans ancestry on gene expression in AA . Roughly , we define cis ancestry as the local ancestry at the gene whose expression is being analyzed , and trans ancestry as the average ancestry at non-cis regions . We extended our above model by letting egs = ccisagγgs+ctransagθs+νgs for AA samples , where γgs denotes the estimated local ancestry of sample s at the SNP closest to the center of gene g ( cis locus; average of transcription start and transcription end positions ) . We note that although trans ancestry is theoretically defined as the average ancestry at non-cis regions , this quantity is in practice virtually identical to θs because cis regions ( regardless of the precise definition of cis ) form an extremely small proportion of the genome . Because chromosomal segments of ancestry in AA typically span >10 Mb [21] , it is nearly always the case that a gene lies completely within a single ancestry block , so that our analysis is not sensitive to the choice of genomic location used to define cis ancestry γgs . The probabilistic estimates of local ancestry produced by HAPMIX are extremely accurate ( see above ) , so that γgs is typically close to 0 . 00 , 0 . 50 or 1 . 00 ( corresponding to 0 , 1 or 2 copies of European ancestry ) . To avoid complications in local ancestry analyses on the X chromosome , we restricted this analysis to 4 , 015 autosomal genes . ( Analyses involving global ancestry were not affected by inclusion or exclusion of genes on the X chromosome . ) We estimated the global parameters ccis and ctrans as above , accounting for the correlation between genome-wide and local ancestry by using residual values of γgs ( adjusted for θs ) to compute ãcis , g , AA ( and conversely for ãtrans , g , AA ) . Let F denote the proportion of total variance in gene expression that is attributable to population differences . For quantitative traits with an additive genetic basis , the quantity that is analogous to single-locus estimates of FST is not F , but rather QST = F/ ( 2−F ) ( reviewed in [22] ) . This is a consequence of the contributions of genetic variation on two distinct haploid chromosomes , magnifying the effect of population differences under an additive genetic model . We computed both F and QST . For each gene g , we normalized gene expression measurements for CEU and YRI to have mean 0 and variance 1 across 120 CEU+YRI samples . We defined the ancestry θs of sample s to be 1 if s is a CEU sample , and 0 if s is a YRI sample . As above , we modeled normalized expression of gene g for sample s as egs = agθs+νgs . Equivalently , under this definition , ag is equal to the difference in normalized gene expression between CEU and YRI samples . We defined F to be the quantity such that the true value of ag has mean 0 and variance 2F across genes [23] . For a specific gene , agθs has variance 0 . 25ag2 and νgs has variance 1–0 . 25ag2 across CEU+YRI samples ( these variances have expected value 0 . 5F and 1–0 . 5F , respectively ) . Due to sampling error , the observed difference ãg in normalized gene expression between CEU and YRI samples ( i . e . the coefficient obtained from a regression of egs on θs ) has variance 2F+ ( 1–0 . 5F ) /30 , where 1/30 is the sum of reciprocals of CEU and YRI sample sizes . We thus estimated mean F as ( Varg ( ãg ) – 1/30 ) / ( 2 – 0 . 5/30 ) . The ratio between mean F and Varg ( ãg ) /2 represents a sampling error correction that we call ξ . We estimated median F as the median value of ãg2/2 times ξ . The value of ξ was 0 . 93 , indicating that the sampling error correction had only a minor effect on these computations . To account for differences between CEU and YRI due to non-genetic factors , we adjusted F by multiplying it by c . ( We note that the scaled population differences cag have variance that is c2 times the variance of ag , but explain only the proportion c of the true component of variance that is attributable to ancestry . ) We then computed QST = F/ ( 2−F ) . We calculated the standard error of our estimate of F via jackknife , repeating the computation of F 120 times with one of the 120 CEU+YRI samples excluded in each computation , and estimating the standard error as the standard deviation of the 120 estimates times the square root of 120 .
We analyzed Affymetrix 6 . 0 genotype data from the African-American panel of 100 samples from Coriell Cell Repositories , together with HapMap samples ( see Materials and Methods ) . We first ran principal components analysis , using the EIGENSOFT software [24] . The top two principal components are displayed in Figure 1 , in which most AA samples roughly lie on a straight line running from CEU to YRI ( we excluded three genetic outliers with partial East Asian ancestry and one genetic outlier whose ancestry is very close to CEU from subsequent analyses ) . This suggests that the ancestry of the AA samples might be reasonably approximated as a mixture of varying amounts of CEU and YRI ancestry , as reported previously [21] . However , given the wide range of genetic diversity across Europe and particularly across Africa [23] , we sought to test this hypothesis further . We removed related samples , genetic outliers , and samples without valid gene expression measurements to obtain a reduced set of 89 AA samples for subsequent analysis ( see Materials and Methods ) . We computed FST values between the set of 89 AA samples and possible linear combinations αCEU+ ( 1−α ) YRI , adjusting for sample size . The lowest value of FST = 0 . 0009 was obtained at α = 0 . 21 . Thus , the 89 AA samples are extremely well-modeled as a mix of CEU and YRI , with average ancestry proportions of 21% CEU and 79% YRI . Though this justifies our modeling approach using CEU and YRI , we caution against drawing historical inferences from this finding: because FST scales with the square of admixture proportion , it is possible that African Americans inherit a small percentage of their ancestry from a more diverse set of populations . We estimated the genome-wide proportion of European ancestry for each the 89 AA samples ( see Materials and Methods ) . Genome-wide ancestry proportions varied from 1% to 62% with a mean±SD of 21±14%; this ancestry distribution is similar to that in other AA data sets [21] , [25] . Genome-wide ancestry estimates were strongly correlated ( r2>0 . 99 ) with coordinates along the top principal component ( eigenvector with largest eigenvalue ) ( Figure 1 ) . We measured gene expression in lymphoblastoid cell lines from 60 CEU and 60 YRI samples from HapMap and 89 AA samples from Coriell , using the Affymetrix Genome Focus Array ( see Materials and Methods ) . Our basic approach was to validate observed differences between CEU and YRI ( differences between sample means ) by analyzing the correlation between the genome-wide proportion of European ancestry estimated from SNP genotyping and the gene expression levels we measured in the AA cell lines . A caveat is that the proportion of European ancestry in African Americans might in principle be correlated to environmental variables . However , such correlations would not affect our approach unless they specifically tracked environmental differences between CEU and YRI . An additional caveat is that the Coriell panel of AA samples is known to be sampled from several ( unknown ) cities in the United States; AA samples from different U . S . cities might differ systematically in both the average proportion of European ancestry [21] , [26] and in the preparation of cell lines . However , ancestry differences among AA populations in different U . S . cities are usually relatively small ( standard deviation of 1% in Table 2 of [21]; standard deviation of 6% in Figure 2 of [26] ) , and in any case would not affect our approach unless differences in cell line preparation specifically tracked differences between CEU and YRI . Using the ancestry estimates and expression data at 4 , 197 genes for CEU , YRI and AA samples , we fit a model in which the effect of ancestry on gene expression at gene g is equal to ag per unit of European ancestry for CEU and YRI samples ( so that ag is equal to the difference in mean expression level between CEU and YRI , which have ancestry 1 and 0 respectively ) , and equal to cag per unit of European ancestry for AA samples , where c is constant across genes ( see Materials and Methods ) . Thus , the global parameter c measures the extent to which observed gene expression differences between CEU and YRI are validated in AA , and therefore heritable . If systematic differences observed between CEU and YRI were entirely due to genetic factors , we would expect to see the same ancestry effects in AA samples , so that c = 1 . On the other hand , under the hypothesis that observed differences between CEU and YRI are entirely due to non-genetic factors , we would expect c = 0 . We note that our procedure for estimating c accounts for both experimental noise and sampling noise in the measurement of gene expression levels . Thus , assuming analogous normalizations for CEU , YRI and AA samples , our estimate of c is not dependent on the accuracy of our measurements; it is also independent of sampling effects . Fitting the above model , we obtained c = 0 . 43 , the slope of the regression line in Figure 2 . With 4 , 197 genes analyzed , this estimate of c is different from zero with overwhelming statistical significance ( P-value<10−25; 95% confidence interval [0 . 38 , 0 . 47] ) . Thus , gene expression differences among AA samples of varying ancestry strongly confirm that heritable differences contribute to observed gene expression differences between CEU and YRI . Performing the analogous computation with genotype data , we obtained c = 0 . 96 , confirming that c is close to 1 for genetic effects ( see Figure 3 ) and that modeling AA as a mix of CEU and YRI is appropriate for our analyses . The deviation between c = 0 . 96 and the expected value of 1 is discussed in Text S1 . We investigated whether the correspondence between observed CEU vs . YRI gene expression differences and expression differences due to ancestry among AA is concentrated in genes with large differences between CEU and YRI . If only a fraction of genes were truly differentiated , as suggested by previous studies , then genes with large observed CEU vs . YRI differences would be more likely to be truly differentiated and would show stronger validation in AA . For example , when we simulated a mixture model in which c = 0 . 43 for the set of all genes but only 50% of genes are truly differentiated between CEU and YRI , we obtained a larger value of c = 0 . 53 for genes in the top 10% of observed CEU vs . YRI differences ( see Text S1 ) . However , Figure 2 shows no evidence of nonlinear effects . Indeed , we recomputed c using only genes in the top 10% of the magnitude of observed CEU vs . YRI differences , and obtained c = 0 . 44 , which is similar to the value of 0 . 43 using all genes . These results suggest that population differences in gene expression are not restricted to a fraction of genes but in fact are widespread across genes , mirroring population differences at the DNA level [6] . We considered whether the alternative approach of analyzing the AA data independently , without regard to differences between CEU and YRI , would be informative about differences in gene expression due to ancestry . We determined that the AA data analyzed separately contains too much sampling noise for that approach to be useful here ( see Text S1 ) . A related observation is that efforts to estimate the proportion of genes with population differences in gene expression , for example using the previously described [27] lower bound statistic 1–π0 , may produce substantial underestimates in the case of data sets affected by sampling noise ( see Text S1 ) . The effect of ancestry on gene expression in African Americans may be due either to variation in regulatory variants proximal to the gene ( cis ) or to variants elsewhere in the genome ( trans ) . We inferred the local ancestry of each AA sample at each location in the genome ( see Materials and Methods ) . A description of how local ancestry varies across the genome ( either across or within samples ) is provided in Text S1 . We quantified the extent to which the validation of CEU-YRI expression differences in AA was attributable to cis or trans effects in AA by computing validation coefficients ccis and ctrans ( see Materials and Methods ) . We obtained ccis = 0 . 05 and ctrans = 0 . 38 . As expected , the sum ccis+ctrans is very close to the validation coefficient c that was obtained using genome-wide ancestry only ( see Text S1 ) . Both ccis ( P-value = 6×10−6; 95% confidence interval [0 . 03 , 0 . 07] ) and ctrans ( P-value<10−25; 95% confidence interval [0 . 33 , 0 . 43] ) were significantly different from zero . Thus , only a small fraction of the effect of ancestry on gene expression is due to ancestry at the cis locus . On the other hand , performing the analogous computation with genotype data , we obtained ccis = 0 . 99 and ctrans = –0 . 03 , indicating as expected that the effect of ancestry on genotype is entirely due to ancestry at the cis locus , and confirming the high accuracy of our estimates of local ancestry . We estimate the proportion πcis of heritable gene expression variation between Europeans and Africans that is due to cis variants as ccis/ ( ccis+ctrans ) = 12% , with a standard error of 3% . An important question is whether our estimate of πcis can be extended to all heritable variation in human gene expression . If the relative magnitude of cis vs . trans effects were different for all variation as compared to population variation—equivalently , if the relative magnitude of population variation relative to all variation were different for cis vs . trans effects—then the answer to this question would be no . To evaluate whether this is the case , we computed FST ( CEU , YRI ) for ∼3 , 000 unique cis eQTL SNPs and ∼700 unique trans eQTL SNPs identified in a recent study of gene expression in human liver [20] . We obtained FST values of 0 . 158 for cis eQTLs and 0 . 154 for trans eQTLs , which were not significantly different from 0 . 159 for all HapMap SNPs ( P-values = 0 . 79 and 0 . 51 respectively ) , based on standard errors computed using the EIGENSOFT software [6] , [24] . Although this analysis involved eQTLs for liver tissue rather than lymphoblastoid cell lines , a reasonable assumption is that the same result holds for other tissue types . Thus , population variation does not appear to differ for cis vs . trans effects , implying that our estimate of πcis = 12±3% applies to all heritable variation in human gene expression . We estimated both the proportion of gene expression variation attributable to population differences , which we call F , and the quantity QST = F/ ( 2−F ) which is analogous to FST for genetic ( allele-frequency ) data ( see Materials and Methods ) . We obtained a mean F = 0 . 20 and median F = 0 . 12 , similar to the median F = 0 . 15 from a previous analysis of CEU and YRI gene expression [8] . A jackknife calculation indicated that the standard error in our estimate of mean F was 0 . 02 , corresponding to a 95% confidence interval of [0 . 15 , 0 . 25] . In our initial calculation of F , we ignored the possibility of non-genetic contributions to population differences . However , the fact that c is smaller than 1 implies that not all of the observed CEU vs . YRI differences are reflected in differences due to ancestry among AA . Some of these differences must reflect non-genetic factors . We therefore adjusted our estimates of F by multiplying them by c = 0 . 43 ( see Materials and Methods ) . After this adjustment , we obtained a mean F = 0 . 09 and median F = 0 . 05 . These estimates of F are substantially lower than those reported previously [8] . Our mean F corresponds to a QST value of 0 . 05 , which is lower than the FST of 0 . 16 that is observed in genetic data [6] . The lower value of QST as compared to genetic data is unsurprising since QST represents a proportion of total gene expression variation , which is expected to include both genetic and non-genetic components . We also note that if measurement variation is substantial , then the use of technical replicates to correct for the effects of measurement variation would lead to a higher value of QST .
We have shown how phenotypic variation in an admixed population can be coupled with variation in ancestry to shed light on differences between ancestral populations; our approach makes no assumptions about the population histories underlying the differences between the ancestral populations . We have applied this approach to gene expression in African Americans and shown that observed population differences ( differences in sample means ) between CEU and YRI in gene expression correspond , with overwhelming statistical significance , to differences among African Americans of varying ancestry , implying a substantial heritable component to the population differences . In reaching this conclusion via analysis of an admixed population , we eliminate confounding with non-genetic contributions to observed differences between the ancestral populations , which could result from differences in environment , differences in preparation of cell lines , or batch effects . The value of 0 . 43 for the “validation coefficient” c implies that both genetic and non-genetic effects contribute to observed population differences between CEU and YRI . Interestingly , the validation coefficient c did not vary appreciably as a function of the magnitude of observed gene expression differences between CEU and YRI . This suggests that the effects of ancestry on gene expression are widespread across genes , as opposed to affecting only a fraction of genes . Although there exist genes for which the observed effect of ancestry on expression levels is close to zero ( Figure 2 ) , this does not rule out small ancestry effects at these genes , as similar results are observed in genetic data ( Figure 3 ) in which it is commonly believed that ancestry affects 100% of common SNP frequencies . Indeed , if ancestry affects genotype and genotype affects gene expression ( as indicated by previous studies reporting a substantial heritable component to gene expression [16] , [17] ) , then the presence of ancestry differences at almost all expressed genes seems a not unreasonable hypothesis , and one with which our results are entirely consistent . However , just as with DNA variation , it is clear that population differences in gene expression represent only a small fraction of the overall variance , most of which is due to variation within populations . In addition to validating the aggregate effects of ancestry on human gene expression , we were able to partition heritable variation into cis and trans effects , which would not be possible in a simple comparison of continental populations . Our admixture approach was fruitful despite the small magnitude of differences between human subpopulations . Our distinction between cis and trans effects is somewhat imprecise , due to the extended length ( >10 Mb ) of segments of continental ancestry in African Americans , but this has little effect on our conclusions , since a 10 Mb region represents a proportion of the genome that is much smaller than the 12% proportion of heritable variation in gene expression that we attribute to variation at the cis locus . Comparing our results to results obtained in other species , we note that two recent studies of gene expression in Drosophila also reported that cis effects explain a small fraction of heritable variation [28] , [29] , although previous Drosophila studies had suggested a larger role for cis effects [30] , [31] . Our results have broad ramifications for future efforts to map the genetic regulation of gene expression . However , conclusions drawn from gene expression measured in lymphoblastoid cell lines do not necessarily extend to other tissue types , motivating further investigation . Going forward , admixed populations will continue to be useful for understanding and mapping gene expression and other phenotypes .
|
Variation in gene expression is a fundamental aspect of human phenotypic variation , and understanding how this variation is apportioned among human populations is an important aim . Previous studies have compared gene expression levels between distinct populations , but it is unclear whether the differences that were observed have a genetic or nongenetic basis . Admixed populations , such as African Americans , offer a solution to this problem because individuals vary in their proportion of European ancestry while the analysis of a single population minimizes nongenetic factors . Here , we show that differences in gene expression among African Americans of different ancestry proportions validate gene expression differences between European and African populations . Furthermore , by drawing a distinction between an African American individual's ancestry at the location of a gene whose expression is being analyzed ( cis ) versus at distal locations ( trans ) , we can use ancestry effects to quantify the relative contributions of cis and trans regulation to human gene expression . We estimate that 12±3% of all heritable variation in human gene expression is due to cis variants .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"genetics",
"and",
"genomics/population",
"genetics",
"genetics",
"and",
"genomics/gene",
"expression"
] |
2008
|
Effects of cis and trans Genetic Ancestry on Gene Expression in African Americans
|
In the juvenile brain , the synaptic architecture of the visual cortex remains in a state of flux for months after the natural onset of vision and the initial emergence of feature selectivity in visual cortical neurons . It is an attractive hypothesis that visual cortical architecture is shaped during this extended period of juvenile plasticity by the coordinated optimization of multiple visual cortical maps such as orientation preference ( OP ) , ocular dominance ( OD ) , spatial frequency , or direction preference . In part ( I ) of this study we introduced a class of analytically tractable coordinated optimization models and solved representative examples , in which a spatially complex organization of the OP map is induced by interactions between the maps . We found that these solutions near symmetry breaking threshold predict a highly ordered map layout . Here we examine the time course of the convergence towards attractor states and optima of these models . In particular , we determine the timescales on which map optimization takes place and how these timescales can be compared to those of visual cortical development and plasticity . We also assess whether our models exhibit biologically more realistic , spatially irregular solutions at a finite distance from threshold , when the spatial periodicities of the two maps are detuned and when considering more than 2 feature dimensions . We show that , although maps typically undergo substantial rearrangement , no other solutions than pinwheel crystals and stripes dominate in the emerging layouts . Pinwheel crystallization takes place on a rather short timescale and can also occur for detuned wavelengths of different maps . Our numerical results thus support the view that neither minimal energy states nor intermediate transient states of our coordinated optimization models successfully explain the architecture of the visual cortex . We discuss several alternative scenarios that may improve the agreement between model solutions and biological observations .
In the primary visual cortex of primates and carnivores , functional architecture can be characterized by maps of various stimulus features such as orientation preference ( OP ) , ocular dominance ( OD ) , spatial frequency , or direction preference [1]–[21] . Many attempts have been made to explain and understand the spatial organization of these maps as optima of specific energy functionals the brain minimizes either during development or on evolutionary timescales [22]–[38] . In part ( I ) of this study we presented an analytical approach to study the coordinated optimization of interacting pairs of visual cortical maps where maps are described by real and complex valued order parameter fields [39] . We used symmetry considerations to derive a classification and parametrization of conceivable inter-map coupling energies and identified a representative set of inter-map coupling terms: a gradient-type and a product-type coupling energy which both can enter with different power in the dynamics . Examining this set of inter-map coupling energies was further motivated by the experimentally observed geometric relationships between cortical maps [5] , [7] , [15] , [19] , [26] , [40] , [41] . We examined the impact of these coupling energies in a system of coupled Swift-Hohenberg equations . These were constructed such that without coupling stripe patterns emerge for the complex valued order parameter field . We found that these types of inter-map coupling energies can induce the formation of defect structures , so-called pinwheels , in the complex order parameter field describing the OP map . For solutions that can become optima of the model , pinwheels are arranged on regular periodic lattices such as rhombic pinwheel crystals ( rPWCs ) or hexagonal pinwheel crystals ( hPWCs ) . These analyses focused on the optimization of a single pair of feature maps in which the complex valued map represented the OP map and the real map the OD map . For this case we presented a complete characterization of the stable OP and OD patterns , stripe-like solutions , rhombic and hexagonal crystalline patterns predicted by the coordinated optimization models . In all analyzed models pinwheel crystallization required a substantial bias in the response properties of the co-evolving real-valued map . The pinwheel crystals we obtained , although beautiful and easy to characterize , qualitatively deviate from the spatially irregular layout observed for OP maps in the visual cortex [2]–[4] . Large scale empirical studies of the arrangement of pinwheel positions and spatial densities in the visual cortex of four species widely separated in mammalian evolution recently showed that orientation maps although spatially irregular precisely conform with apparently species insensitive quantitative layout rules [42] , [43] . In particular , it was found that not only the mean density of pinwheels but also number fluctuations over a wide range of spatial scales and local next neighbor arrangements within individual hypercolumns agree across species with an accuracy in the range of a few percent [42] , [43] , see also Fig . 1 . In contrast to the large variability of local map layouts in experimentally observed maps [1]–[21] , the pinwheel crystals found in the coordinated optimization models introduced in part ( I ) show a regular and stereotyped structure . Quantitatively , all PWC solutions that we found exhibit a large pinwheel density of about 3 . 5 or even 5 . 2 pinwheels per hypercolumn . For experimental OP maps the average pinwheel density was found to be between 3 . 1 and 3 . 2 and statistical indistinguishable from the mathematical constant up to a precision of 2% [42] , [44] , [45] . Our previous analytical results thus raise the question of whether and how our coordinated optimization models can be reconciled with the experimentally observed layout rules of orientation maps . From a biological perspective , one might suspect that the crystalline layouts of local minima and optima results from the restrictions of the applied perturbation method which allowed us to study optima analytically but might be biased towards particular solution classes . Furthermore , results might change substantially if one would consider the coordinated optimization of more than two feature maps . Examining this aspect is also demanded because of the presence of multiple feature maps in the visual cortex of primates and carnivores . Furthermore geometrical rules coordinating map layout might in general be the harder to satisfy the more maps are simultaneously optimized . Finally , when studying optima predicted by a particular optimization principle we disregarded transient states that could in principle dominate developmental optimization on biologically relevant timescales . Such transient solutions are expected to be more irregular than the final attractor states . Analytical results were obtained using a perturbative treatment close to the pattern forming threshold . This perturbative treatment , however , gives no information on the speed with which singularities will crystallize into highly ordered arrays . It is conceivable that this process may occur on very long timescales . If this was the case , developmental optimization may lead to long-lived spatially irregular states that are transients towards regular patterns that would be reached after very long times or potentially never . To assess this possibility it is critical to examine model predictions over a wide range of timescales and compare biological developmental phases to different stages in numerical model simulations . In the current study we propose a systematic procedure for such comparisons that is based on a wide array of development experiments and theoretical analyses . Numerical studies complementing the analyses presented in part ( I ) are also demanded for various theoretical reasons . In part ( I ) we showed that one can neglect the backreaction of the OP map onto the OD map if the OD map is ‘dominant’ i . e . its amplitude is much larger than that of the OP map . This can be achieved for a sufficiently small ratio of their distances to threshold . This finding raises questions that cannot be easily addressed perturbatively . Do the observed local minima and optima of the optimization principles persist when taking the backreaction into account or when considering map formation further from the pattern formation threshold ? Besides the influence of the backreaction , the full dynamical system receives additional corrections . There are higher order corrections to the uncoupled amplitude equations which can become important for finite bifurcation parameters but were neglected in part ( I ) [39] . In part ( I ) of this study we also assumed equal periodicities of the two interacting maps . Systematic differences of OD and OP wavelengths have been observed for instance in macaque monkey visual cortex [7] , [46] . In case of cat visual cortex different OP and OD wavelength have been observed within the same animal [47] although the average wavelength of the OD and OP pattern appears similar on average [48] , [49] . Experiments suggest that the different periodicities in the layout of OP and OD maps can have an impact on the map layout [7] , [48] , [49] . It is thus also interesting to explore whether and how a detuning of typical periodicities affects optimal layouts and whether it can lead to spatially irregular maps . To assess these issues we generalized the field dynamics to describe the coordinated optimization of coupled complex valued and several real valued scalar fields . From a practical point of view , the analyzed phase diagrams and pattern properties indicate that the higher order gradient-type coupling energy is the simplest and most convenient choice for constructing models that reflect the correlations of map layouts in the visual cortex . For this coupling , intersection angle statistics are reproduced well , pinwheels can be stabilized , and pattern collapse cannot occur . In the current study we thus numerically analyzed the dynamics of coordinated optimization focusing on the high order gradient-type inter-map coupling energy . We use a fully implicit integrator based on the Crank-Nicolson scheme and a Newton-Krylow solver . In numerical simulations we characterize the kinetics and conditions for pinwheel crystallization and the creation of pinwheels from a pinwheel-free initial pattern . We assessed layout parameters of OP maps throughout all stages of optimization . To aid comparison with developmental timescales all results are represented with time normalized to the time required for maturation of orientation selectivity . Creation of pinwheels from a pinwheel-free initial pattern is a sufficient although not a necessary criterion for systems in which a pinwheel-rich state is energetically favored . As we point out this criterion can be easily assessed in models of arbitrary complexity that otherwise evade analytical treatment . We further explored the impact of inter-map wavelength differences , as observed in certain species , on the structure of the resulting solutions . Finally , we extended the models to explore the coordinated optimization of more than two feature maps . To examine whether the observed quantitative properties can be reproduced in models for the coordinated optimization of maps we calculated various pinwheel statistics during optimization . We find that spatially irregular patterns decay relatively fast into locally crystalline arrays . Further long-term rearrangement mainly leads to the emergence of long-range spatial alignment of local crystalline arrangements . We showed that our previous finding that OD stripes are unable to stabilize pinwheels generalizes to the case of detuned wavelengths . The observation that the coordinated optimization of two interacting maps leads to spatially perfectly periodic optima is also robust to detuned typical wavelengths and to the inclusion of more than two feature maps . Our results suggest that the coordinated optimization of multiple maps that would in isolation exhibit spatially perfectly periodic optimal layouts on its own does not offer a simple explanation for the experimentally observed spatially irregular design of OP maps in the visual cortex and its quantitative aspects . We consider alternative scenarios and propose ways to incorporate inter-map relations and joint optimization in models in which the optimal OP map layout is intrinsically irregular already for vanishing inter-map coupling .
We model the response properties of neuronal populations in the visual cortex by two-dimensional scalar order parameter fields which are either complex valued or real valued [50]–[52] . We consider inter-map coupling between a complex valued map and one or several real valued maps . The complex valued field can for instance describe OP or direction preference of a neuron located at position . A real valued field can describe for instance OD or spatial frequency preference . Although we consider a model for the coordinated optimization of general real and complex valued order parameter fields we view as the field of OP throughout this article to aid comparison to the biologically observed patterns . In this case , the pattern of preferred stimulus orientation is obtained by ( 1 ) The modulus is a measure of orientation selectivity at cortical location . OP maps are characterized by so-called pinwheels , regions in which columns preferring all possible orientations are organized around a common center in a radial fashion [50] , [53]–[55] . The centers of pinwheels are point discontinuities of the field where the mean orientation preference of nearby columns changes by 90 degrees . Pinwheels can be characterized by a topological charge which indicates in particular whether the orientation preference increases clockwise or counterclockwise around the pinwheel center , ( 2 ) where is a closed curve around a single pinwheel center at . Since is a cyclic variable in the interval and up to isolated points is a continuous function of , can only have values ( 3 ) where is an integer number [56] . If its absolute value , each orientation is represented only once in the vicinity of a pinwheel center . In experiments , only pinwheels with a topological charge of have been observed , which are simple zeros of the field . In case of a single real valued map the field can be considered as the field of OD , where indicates ipsilateral eye dominance and contralateral eye dominance of the neuron located at position . The magnitude indicates the strength of the eye dominance and thus the zeros of the field corresponding to the borders of OD domains . If visual cortical maps are described by optima of an energy functional , a formal time evolution of these maps that represents the gradient descent of this energy functional can be used to obtain predicted map layouts . The field dynamics thus takes the form ( 4 ) where and are nonlinear operators given by , . The system then relaxes towards the minima of the energy . The convergence of this dynamics towards an attractor is assumed to represent the process of maturation and optimization of the cortical circuitry . Various biologically detailed models can be cast into the form of Eq . ( 4 ) [22] , [52] , [57] . To dissect the impact of inter-map coupling interactions we split the energy functional into single field and interaction components . All visual cortical maps are arranged in roughly repetitive patterns of a typical wavelength that may be different for different maps . We chose to obtain , in the absence of coupling , a well studied model reproducing the emergence of a typical wavelength by a pattern forming instability , the Swift-Hohenberg model [58] , [59] . This model has been characterized comprehensively in the pattern formation literature and mimics the behavior of for instance the continuous Elastic Network or the Kohonen model for orientation selectivity ( see [22] ) . We note that many other pattern forming systems occurring in different physical , chemical , and biological contexts ( see for instance [60]–[63] ) have been cast into a dynamics of the same form . Its dynamics in case of the OP map is of the form ( 5 ) with the linear Swift-Hohenberg operator ( 6 ) , and the Laplace operator . In Fourier representation , is diagonal with the spectrum ( 7 ) The spectrum exhibits a maximum at , see Fig . 2A . For , all modes are damped since and only the homogeneous state is stable . This is no longer the case for when modes on the critical circle acquire a positive growth rate and grow , resulting in patterns with a typical wavelength . This model exhibits a supercritical bifurcation where the homogeneous state looses its stability and spatial pattern emerge . While the linear part of the dynamics establishes a typical wavelength , the nonlinear term in the dynamics leads to the selection of the final pattern [64] , [65] . Considering the time evolution following Eq . ( 5 ) initialized with a random OP map and low selectivity ( small ) several different stages of the dynamics can be distinguished . The linear part forces modes on the critical circle to grow with rate while strongly suppressing modes off the critical circle when starting from small amplitude white noise initial conditions , see Fig . 2A . The OP map becomes more ordered in this linear phase as one dominant wavelength emerges . The total power of the field is given by ( 8 ) where denotes spatial average . The time dependence of the power reflects the different growth rates among modes . The time evolution of the power is depicted in Fig . 2B . Initially , the power decreases slightly due to the suppression of modes outside the circle of positive growth rate . At there is a rapid increase followed by the saturation of the power . The amplitudes of the Fourier modes reach their stationary values and . At this stage of the evolution the influence of the nonlinear part becomes comparable to that of the linear part . Once the modes saturate the phase of nonlinear competition between the active modes along with a reorganization of the structure of the OP map starts . The competition between active modes leads to pattern selection i . e . the convergence toward one of the in principle infinitely many periodic and aperiodic fixed points of the evolution equations . The final pattern then consists of distinct modes in Fourier space [59] , [64] . Once the active modes are selected a relaxation of their phases takes place . These stages thus represent an initial process of selectivity maturation and a process of convergence to a stationary layout . This suggests to compare the first stage to the biological developmental period in which neurons reach adult-like levels of orientation selectivity and the later convergence stage to the following period of developmental juvenile plasticity e . g . until the closure of the developmental critical periods . To aid a detailed comparison we are presenting all maps and layout parameters as a function of time . In such displays time , during gradient descent optimization , is represented in two different ways . Firstly , following conventions in the pattern formation literature , time is rescaled with the largest growth rate of the OP map , . Secondly , to aid comparison with biological observations , we also graph all calculated layout properties as a function of ‘developmental time’ where is the time for which the OP power reaches its peak value or , if there is no peak in the OP power , reaches 90% of its final value . In these units represents the time when orientation selectivity is essentially mature and later times correspond to subsequent convergence processes . Inter-map coupling can influence the time evolution on all stages of the development depending on whether this coupling affects only the nonlinear part or also the linear one . When incorporating additional maps into the system in all cases we rescaled the dynamics by the bifurcation parameter of the OP map i . e . . The coupled dynamics we considered is of the form ( 9 ) where , and is a constant . To account for the differences in the dominant wavelengths of the patterns we chose two typical wavelengths and . In the sections ‘Final states’ and ‘Kinetics of pinwheel crystallization’ we assume i . e . the Fourier components of the emerging pattern are located on a common circle . In the subsequent sections we also consider a potential detuning of the typical wavelength . The dynamics of and are coupled by interaction terms which can be derived from a coupling energy . Many optimization models of the form presented in Eq . ( 4 ) have been studied [22]–[38] . The concrete dynamics in Eq . ( 9 ) is the simplest which in the uncoupled case leads to pinwheel-free OP stripe patterns and to a stripe-like or patchy layout of the co-evolving real valued fields . As revealed by the symmetry-based classification of coupling energies ( 10 ) parametrizes a representative family of biologically plausible coupling energies for a single real valued map , see part ( I ) , [39] . The numerical integration scheme to solve Eq . ( 9 ) is detailed in the Methods part . For numerical analysis we focused on the high order gradient-type inter-map coupling energy . This energy can reproduce all qualitative relationships found between OP and OD maps , does not suffer from potential OP map suppression , and leads to a relatively simple phase diagram for two interacting maps near threshold . In part ( I ) we calculated phase diagrams for different inter-map coupling energies [39] . In all cases , hexagonal PWCs can be stabilized only in case of OD hexagons . We tested these results numerically . Numerical simulations of the dynamics Eq . ( 9 ) with the coupling energy ( 11 ) are shown in Fig . 3 . All remaining inter-map coupling energies in Eq . ( 10 ) are assumed to be zero . Initial conditions for the OD map were chosen as spatially irregular patterns or stripe patterns with saturated power plus Gaussian white noise . Initial conditions for the OP map are either pinwheel-free OP stripes or band-pass filtered Gaussian white noise for which the average pinwheel density is bounded from below by the constant [22] . The initial conditions and final states are shown for different bias terms and inter-map coupling strengths . We observed that for a substantial contralateral bias and above a critical inter-map coupling pinwheels are preserved for all times or are generated if the initial condition is pinwheel-free . Without a contralateral bias the final states were pinwheel-free stripe solutions irrespective of the strength of the inter-map coupling . To characterize the process of pinwheel annihilation , preservation , and creation during progressive map optimization we calculated the pinwheel density as well as various other pinwheel statistics ( see Methods ) during the convergence of patterns to attractor states . The time evolution of the pinwheel density is shown in Fig . 4 . Initial conditions for the OD map were chosen as hexagonal patterns plus Gaussian white noise . Initial conditions for the OP map are either pinwheel-free OP stripes or band-pass filtered Gaussian white noise . Note the logarithmic time scales . Pinwheel densities rapidly diverge from values near 3 . 1 as soon as the map exhibits substantial power . In the uncoupled case ( ) most of the patterns decayed into a stripe solution and their pinwheel density dropped to a value near zero . At small coupling strengths ( ) the pinwheel density converged either to zero ( stripes ) , to values near 3 . 5 for the rPWC ( see Fig . S6 in part ( I ) , [39] ) , or to approximately 5 . 2 for the contra-center hPWC ( see Fig . S7 in part ( I ) , [39] ) . At high inter-map coupling ( ) pinwheel free stripe patterns formed neither from pinwheel rich nor from pinwheel free initial conditions . In this regime the dominant layout was the contra-center hPWC . When starting from OD and OP stripes , see Fig . 4C ( green lines ) , the random orientation between the stripes first evolved towards a perpendicular orientation ( ) . This lead to a transient increase in the pinwheel density . At the time ( ) where the OD stripes dissolve towards OD hexagons hPWC solutions formed and the pinwheel density reached its final value . Regions of hPWC layout can however be inter-digitated with long lived rPWC patterns and stripe domains . Figure 4D shows the time course of the normalized power , where denotes spatial average . The field is obtained from the solution of the amplitude equations ( see [39] ) while is the field obtained from the simulations . Starting from a small but nonzero power the amplitudes grew and saturated after . When the amplitudes were saturated the selection of the final pattern started . Quantitatively , we found that with weak backreaction the critical coupling strengths were slightly increased compared to their values in the limit . Snapshots of the simulation leading to the hPWC solutions at three time frames are shown in Fig . 5 . Already at a substantial rearrangement of the pattern took place and one can identify different domains in the pattern that are locally highly stereotyped . For the time evolution of the maps we also calculated the distributions of pinwheel next-neighbor distances , measured in units of the column spacing . The distributions of distances for simulations leading to rhombic and hPWC solutions are shown in Fig . 6 . They are characterized by three stages in the evolution of the pinwheel distances . At early stages of the evolution ( ) there is a continuous distribution starting approximately linearly from . At the time where the amplitudes saturated ( ) the distribution of pinwheel distances became very inhomogeneous . Different domains with stripe-like , rhombic , or hexagonal patterns appeared until for the rhombic or hexagonal pattern took over the entire system . As pinwheels carry a topological charge we could divide the distributions according to distances between pinwheels of the same charge or according to distances between pinwheels of the opposite charge . In Fig . 7 we present pinwheel distances for the final states of the dynamics . In case of the rhombic solutions there is only a single pinwheel to pinwheel distance with . In numerical simulations small variations in the amplitudes lead to a slightly larger distance between pinwheels of equal charge than between pinwheels of opposite charge . Therefore their distance distributions do not collapse exactly , see Fig . 7A . In case of the hPWC there are three peaks at and in the pinwheel distance distribution of arbitrary charge , see Fig . 7B . These three peaks all result from distances between pinwheels carrying the opposite charge while the distance between pinwheels of the same charge shows two peaks at and in the distribution . The origin of the peaks is indicated in Fig . 7C and Fig . 7D . These results confirm that inter-map coupling can induce the stabilization of pinwheels in the OP pattern . This however does not mean that the pinwheels initially generated by spontaneous symmetry breaking will be preserved during convergence of the map . To what extent are the pinwheels in the crystalline OP maps preserved from pinwheels of the initial OP pattern ? To answer this question we calculated the pinwheel annihilation and creation rate during time evolution , see Methods . The time evolution of these rates , averaged over 20 simulations leading to a hPWC , is shown in Fig . 8A . We observe that both rates were fairly similar throughout development , with a slightly higher creation rate in the later stage of development . During the initial stages of time evolution creation and annihilation rates decay algebraically . At both rates deviate from this algebraic decay . From thereon annihilation and creation rates increase , reflecting the nonlinear rearrangement of the pattern . After no pinwheels are created or annihilated anymore and the pinwheels of the final pattern are present . Pinwheels are created and annihilated until a first crystal-like pattern is formed . How many pinwheels of the initial pattern are still present in the final pattern ? For a given set of pinwheels at an initial time we further calculate the fraction of those pinwheels surviving until time . The fraction of pinwheels present at time that survive up to the final time is given by . Both fractions are shown in Fig . 8B for and in Fig . 8C for , a time where the power has almost saturated , see Fig . 4D . We observed that about 20% of the initial pinwheels are preserved until the final time and therefore most of the pinwheels of the crystal pattern are created during development . From those pinwheels which are present when the power saturates about 65% are also present in the final pattern . The analytical results obtained in [39] as well as the previous numerical results ( see Fig . 3B ) predict that OD stripes do not lead to spatially complex patterns and are not capable of stabilizing pinwheels . In case of gradient-type inter-map couplings the OP map consists of stripes which run perpendicular to the OD stripes . In case of the product-type inter-map coupling high gradient regions of both maps avoid each other by producing again OP stripes but now oriented parallel to the OD stripes . In numerical simulations we also investigated the case of OD stripes of larger wavelength than OP columns , as is the case in macaque monkey primary visual cortex [7] , [46] . In case of a gradient-type inter-map coupling we find that the OD bands are perpendicular to the OP bands independent of the ratio , see Fig . 9C , D and Fig . S1 . In case of the product-type inter-map coupling , if the ratio , the orientation representation does not collapse as it would be the case for , see [39] . The system , however , again finds a way to put zero contours ( Re and Im ) along lines of maximal OD resulting in an orientation fracture line , see Fig . 9E , F and Fig . S1 . The angle between the active OP and OD modes is given by corresponding to the resonance condition , see Fig . 9A . The time evolution of all pinwheel statistics and its comparison to the case of equal wavelengths is shown in Fig . 10 and Fig . S2 . Initial conditions are band-pass filtered Gaussian white noise with initial power a few percent of the final power . Note , the pinwheel statistics are shown for the timescale and which relates the pinwheel statistics to the rise and saturation of the orientation selectivity . The power of the OP map reached about 90 percent of its final value earliest at . A non-monotonic time dependence of OP power can result from inter-map coupling . In particular , the rise of OD power leads to OP suppression . This suppression is absent if the pattern arranges such that the inter-map coupling energy is zero i . e . for perpendicular stripe patterns . In all cases , at the average pinwheel density clearly deviates from the experimentally observed value . Furthermore , pinwheel densities are for all cases outside of the confidence interval of the species grand average pinwheel density obtained in [42] . Note , in case of equal wavelength and a product-type inter-map coupling energy the OP map develops towards orientation scotoma solutions which are selective to only two preferred orientations , see Fig . 9G and part ( I ) . In case of identical wavelengths strong interaction with a system of hexagonal OD patches leads to hPWC solutions . For these solutions pinwheel positions are correlated with OD extrema . For instance in case of the higher order gradient-type inter-map coupling energy , for which the contra-center PWC corresponds to the energetic ground state , half of the pinwheels are located at OD extrema while the remaining half are located near OD borders ( see Fig . S7 of part ( I ) , [39] ) . If , however , the typical wavelengths of OD and OP patterns are not identical such a precise relationship cannot be fulfilled in general . We therefore studied whether a detuning of typical wavelengths can lead to spatially irregular and pinwheel rich OP patterns . In numerical simulations which lead to OD hexagons with a fixed wavelength we varied the OP wavelength using as initial conditions band-pass filtered Gaussian white noise with power a few percent of the final power . Wavelength ratios were chosen such that each pattern exhibited an integer aspect ratio . Wavelength ratios were with an integer . Examples of final patterns of such simulations are shown in Fig . 11 and Fig . S3 using the high order gradient-type inter-map coupling energy . In all studied cases the final patterns are spatially regular . The observed patterns are either fractured stripe patterns with two active modes Fig . 11C or rPWC solutions ( two modes plus the corresponding opposite modes ) . We also studied map interactions with wavelength ratios were with . In this case , however , we found only pinwheel-free stripe patterns as final states . Much larger domains than used in the current simulations would be needed to simulate values intermediate to the wavelength ratios used here . Our results , nevertheless , clearly establish that OD induced pinwheel stabilization can occur also with detuned wavelengths . They furthermore confirm that wavelength detuning does not by itself generates irregular stable maps in the considered model . The time evolution of all pinwheel statistics and its comparison to the case of equal wavelengths is shown in Fig . 12 and Fig . S4 . The pinwheel density appears to exhibit a complex dependence on the wavelength ratio . The power of the OP map reached about 90 percent earliest at . At , however , the average pinwheel density in all conditions clearly deviates from the experimentally observed value . Furthermore , at the pinwheel density for all conditions is outside of the confidence interval of the species grand average pinwheel density as obtained in [42] . For three conditions , the mean pinwheel density transiently reentered the confidence interval at a later stage for a short period of time . For no condition , however , there was a robust and stable convergence of the predicted pinwheel density to the confidence interval for . Note , in case of equal wavelength the OP map develops towards contra-center PWCs , see also Fig . 3C , D and part ( I ) . The inclusion of more feature dimensions into the dynamics was performed as in Eq . ( 9 ) , Eq . ( 10 ) as the geometric correlations between the different types of maps seem to be qualitatively similar [9] , [10] , [14] , [26] , [40] . We used the higher order gradient-type inter-map coupling with three and four maps which are mutually coupled , see Fig . 13 and Fig . S5 . Initial conditions for all maps were band-pass filtered Gaussian white noise with the initial power a few percent of the final power , see Fig . 14 . Whereas in the case of two maps the coupling energy is zero if the two stripe solutions are perpendicular to each other the interactions between more maps could potentially lead to frustration as not all of the individual coupling energies can simultaneously vanish . Using the gradient coupling energy ( 12 ) no OD bias ( ) , and equal coupling strengths we observed two types of stationary solutions , see Fig . 13 . When all bifurcation parameters were equal , the OP map consisted of stripes . Also the two real fields consisted of stripes , both perpendicular to the OP stripes i . e . ( 13 ) The energy in this case is given by , which is minimal for , i . e . the energy is minimized by shifting one real field by one quarter of the typical wavelength . When the bifurcation parameter of the OP map was smaller than that of the two real fields we obtained PWC patterns , see Fig . 13B . The pinwheels were arranged such that they are in the center of a square spanned by the two orthogonal real fields and the resulting pinwheel density is . All intersection angles between iso-orientation lines and borders of the real fields were perpendicular . When extending the system by a third real field we observed a similar behavior . Figure 13C , D shows the stationary states of a complex field coupled to three real fields . In case of equal bifurcation parameters the stationary patterns were OP stripes , perpendicular to stripe and wavy real patterns . When the bifurcation parameter of the OP map was smaller than the other bifurcation parameters we again observed pinwheel crystallization . Note , that in this case all pinwheels were located at the border of one of the three real fields . In summary , pinwheel crystallization was only observed when the OP map is driven by the real field i . e . when the OP amplitudes are small . In all observed cases the final patterns were spatially perfectly periodic . The time evolution of all pinwheel statistics is shown in Fig . 14 and Fig . S6 . The power of the OP map reached about 90 percent of its final power earliest at . At the average pinwheel density in all cases clearly deviates from the experimentally observed value . Furthermore , at the pinwheel density in all cases is outside of the confidence interval of the species grand average pinwheel density as obtained in [42] .
In this and the accompanying analytical study , we presented a dynamical systems approach to the coordinated optimization of maps in the visual cortex such as orientation preference ( OP ) and ocular dominance ( OD ) maps . In part ( I ) we examined in particular the predicted optima of various candidate energy functionals [39] . We calculated phase diagrams for different energy functionals showing that for strong inter-map coupling pinwheel crystals are optima of the system . In the current study , we numerically analyzed the dynamics of two representative examples of these coordinated optimization models . We focused on the high order gradient-type inter-map coupling energy that can reproduce all qualitative relationships found experimentally between OP and OD maps , does not suffer from potential OP map suppression , and has a relatively simple phase diagram near the symmetry breaking threshold . The main phenomenon characterizing the considered models , crystallization induced by coordinated optimization and inter-map coupling , was confirmed numerically . This phenomenon was found to be robust to the influence of a weak backreaction of the OP map on the OD map , to detuning of the typical wavelengths , and was found to persist in models with higher feature space dimensionality . We characterized the complex dynamics during crystallization and calculated all pinwheel statistics known for the common design of OP maps found experimentally . The crystalline periodic layout of pinwheel-rich solutions persisted in all studied conditions . Characterizing the behavior of transients , we found that spatially irregular transient states decayed relatively fast into locally ordered patterns during optimization . The optimal layouts predicted by the models considered in the current studies deviate qualitatively and quantitatively from experimentally observed map layouts . It is therefore not reasonable to assume that a genetically encoded pattern of cortical columns has been optimized on evolutionary timescales following the optimization principles formalized by our models . Two alternative scenarios , however , are raised by our results . First , visual cortical maps could be considered as optimized with respect to principles qualitatively distinct from those examined here ( see e . g . [66] , [67] ) . Second , visual cortical maps might be incompletely optimized by a developmental dynamics that reduces an energy functional such as the ones considered here but does not reach optimized states due to a finite duration of the period of juvenile plasticity . In the following we first discuss the incomplete optimization scenario and propose quantitative criteria for testing its plausibility . We then discuss likely ingredients of fundamentally different optimization principles that appear better suited to explain visual cortical architecture . In our simulation studies , we examined the sequence of stages predicted , for the maps under the assumption of developmental optimization . Our results consistently show that our coordinated optimization models exhibit a complex dynamics that persistently reorganizes maps over different timescales before attractors or optima are reached . As can be predicted from symmetry principles [22] , [68] , at early stages of development maps must be spatially irregular if they develop from weakly tuned random initial conditions . Such OP patterns are essentially random exhibiting a model insensitive , universal spatial organization throughout the initial emergence of orientation selectivity . The average pinwheel density in these early maps is bounded from below by the mathematical constant and the distributions of nearest pinwheel distances are continuous and broad . As soon as orientation selectivity started to saturate the patterns typically reorganized towards one of a few crystalline spatial patterns . This early phase of local crystallization rapidly leads to the occurrence of different spatial domains within the pattern , with a locally stereotypical periodic layout . Even in the cases exhibiting the slowest decay of irregular patterns , this process was complete after ten intrinsic timescales . The slower dynamics that characterizes further development progressively aligns these domains leading to a long-range ordered perfectly periodic crystalline array . This long-ranged reorganization of patterns lasts substantially longer than the intrinsic timescale . Similar behavior was also observed when starting near spatially irregular unstable fixed points of the orientation map dynamics . Pinwheel crystal ( PWC ) solutions represent attractor states and we found no other , spatially irregular , long-living states in the dynamics . The overall progression of states observed in our models has been found previously in numerous pattern forming dynamics both highly abstract as well as in detailed ab initio simulations [59] , [64] , [65] . Does the observed rapid decay of irregular OP layouts into crystalline patterns speak against the biological plausibility of an optimization dynamics of the type considered here ? Can one reasonably expect that a similar crystallization process could also unfold relatively rapidly during the development of the brain ? Or is it more likely that what seems rapid in our numerical simulations would take very long in a biological network - potentially so long that the cortical circuitry has already lost its potential for plastic reorganization before substantial changes have occurred ? To answer these questions it is important ( 1 ) to examine whether secondary reorganization processes subsequent to the initial establishment of selectivity are occurring during biological development , ( 2 ) to delimit the fundamental timescales of the postnatal development of visual cortical circuits subserving orientation preference and ocular dominance and ( 3 ) to discuss how these timescales can be compared to the formal timescales that appear in dynamical models of map formation and optimization . In the following we address these issues . We will first summarize available evidence for ongoing pattern reorganization subsequent to the initial emergence of feature selectivity . We will then discuss the theoretically predicted properties of the fundamental time scale of the map dynamics and finally discuss how to empirically estimate it relative to the duration of visual cortical critical period plasticity . For comparing simulation results to developmental stages in the biological system the most important quantity is the relative duration of the period of juvenile plasticity; the ratio of the absolute duration of juvenile plasticity and the fundamental time scale of the map dynamics . Secondary map rearrangement has been experimentally found by several studies [69]–[73] . It is expected if this ratio is substantially larger than one . How far developmental reorganization can be expected to progress towards attractor states during the period of juvenile plasticity is determined by its absolute value . Current empirical uncertainties do not permit to determine the relative duration of the period of juvenile plasticity with great accuracy . It is however , possible to estimate a conservative lower bound and a worst case estimate upper bound . We argue that plausible candidate models should correctly predict map layouts in adult visual cortex when reaches the lower bound . In general model predictions should be compared to biological observations throughout the range delimited be the lower and upper bounds for a systematic assessment of the robustness of model behavior . Current data implies a conservative lower bound to the duration of the period of juvenile plasticity of about . Accumulating evidence suggests that juvenile plasticity supports an ongoing pattern reorganization [69]–[73] . For cat visual cortex , Kaschube and coworkers have demonstrated that the spatial organization of orientation columns in striate cortex is progressively reorganized between the sixth and the 14th postnatal week such that the organization of orientation columns that are reciprocally connected to extra-striate visual cortex and contra-lateral hemisphere striate cortex are better matched [70] . A second line of evidence is related to the fact that the surface area of cat striate cortex substantially increases postnatally [71] , [74]–[76] . The spatial periodicity of both orientation as well as OD columns , however , remains basically unaffected during this period [70] , [71] , [77] . Keil and coworkers reported that this areal growth in the presence of maintained mean column spacing induces a specitifc kind of spatial reorganization of the layout of OD columns within cat striate cortex [71] . Independently , growth related rearrangement of orientation columns has also been suggested previously by Kiorpes and coworkers from observations on a smaller data set from juvenile macaques [46] . Perhaps the most striking demonstration that the functional preferences of visual cortical neurons can reorganize over long time scales during the period of juvenile plasticity has emerged from studies of the mouse visual cortex . In the mouse , as in cat , visual cortical neurons first develop orientation selectivity around the time of first eye opening in the second postnatal week [69] , [72] , [73] . Similar to the developmental time course in the cat , the duration of the period of juvenile plasticity in the mouse is quite long and extends beyond the third postnatal month [78] , [79] . At a duration of more than 10 weeks , it is thus substantially longer than required for the expression of adult-like single neuron selectivities . Wang and coworkers demonstrated that neurons in the binocular segment of mouse visual cortex change their preferred orientations during this period [69] . Neurons in the binocular segment of mouse striate cortex were found to first exhibit widely different preferred orientations in the left and right eye . The two different preferred orientations then underwent secondary reorganization and became matching at an age of 5 weeks postnatally after the peak of the OD critical period [69] . In the monocular segment of mouse visual cortex , Rochefort and coworkers found substantial changes in the complement of preferred orientations and preferred directions represented during the first postnatal month [72] . It is noteworthy that a substantial long-term reorganization of cortical preferences has also been demonstrated for the preferred direction for whisker deflection in the rat barrel cortex [80] . Here Kremer and coworkers found that preferences for the direction of whisker deflection reorganize over the course of the first three postnatal months . Long-term reorganization might thus potentially constitute a general feature of sensory cortical representations . Experimental evidence thus clearly supports that circuits in sensory cortical areas remain in a state of flux for weeks and months after the initial emergence of sensory responsiveness and stimulus selectivity . To judge how far the rearrangement of cortical circuitry can progress towards a stationary optimized state one has to relate the duration of the period of juvenile plasticity to the fundamental timescale of the map dynamics . This time scale essentially is the duration of the process of establishing mature levels of response selectivity and in our models is the time . Let us first discuss the determinants of this time scale from a theoretical perspective . All models for the development of visual cortical functional selectivity from an unselective or weakly selective initial condition are known to exhibit a distinct intrinsic time scale for the emergence of stimulus selectivity in individual neurons , see e . g . [36] , [50] , [52] , [81]–[83] . In the abstract order parameter models used here this time scale is set by the inverse of the maximum eigenvalue . It is important to note that this time scale represents an effective parameter describing a collective circuit property . Consequently this parameter is not rigidly related to any particular cellular or synaptic time constant such as e . g . the characteristic times required for the expression of LTP , spine growth , homeostatic plasticity or other functional or morphological synaptic changes . Theoretical studies of microscopic models in which the effective time scale was explicitly calculated established that the intrinsic time scale depends ( 1 ) on the ensemble of activity patterns driving development and also ( 2 ) on characteristics of the local cortical circuits [57] , [83] , [84] . For instance in a representative , analytically solvable microscopic model for the emergence of OD patterns the maximum eigenvalue is given bywhere is the characteristic time scale for synaptic changes , the spatial extend of co-activated neuron groups in the model cortex and , and are auto- and cross-correlation functions of the activity patterns in the left and right eye layers of the LGN [57] . The effective time scale for the emergence and saturation of response selectivity is thus expected not to be faster than the fundamental processes of synaptic change . A broad range of time constants , however , is in principle consistent with Hebbian models of sensory cortical development depending on details of circuit interactions . An empirical estimate for the relative duration of juvenile plasticity can be obtained by comparing the characteristic time scale of initial map emergence and the duration of the period from map emergence to the closure of critical periods for visual cortical plasticity . How long does the emergence of stimulus selectivity in the visual cortex take under normal conditions ? For orientation selectivity in the primary visual cortex this information has been experimentally obtained for cat and ferret . In both species orientation selectivity is established starting from an initial condition in which cells are only weakly orientation biased . Data indicate a period between a few days and at most one week to reach mature levels of single cell orientation selectivity [11] , [17] , [21] , [85]–[88] . Similar time scales are sufficient for substantial morphological changes of thalamo-cortical axonal structure [89] . A conservative estimate for the intrinsic time scale of map dynamics is thus that is about one week . As mentioned above the period of juvenile cortical plasticity is known to last substantially longer . Best established experimentally is the period of susceptibility to monocular deprivation in the cat that lasts for several months of postnatal life [86] , [90]–[95] . The primary visual cortex of the cat is maximally susceptible to monocular deprivation in kitten of four weeks of age [90]–[92] . An initial establishment of OP and OD maps in kittens occurs around the time of eye opening and is complete at the end of the second postnatal week [85] , [87] . A maximal degree of plasticity is thus reached two weeks after the emergence of maps and the onset of natural vision . After the first postnatal month susceptibility to monocular deprivation gradually declines back to levels comparable to those present at the onset of vision and initial map emergence . The closure of the period of developmental plasticity was estimated by three independent studies to occur between the 14th and 18th postnatal week [90]–[92] . This is 12 to 16 weeks after single neurons first exhibit adult like levels of orientation selectivity and eye dominance . The reported durations for the period of juvenile plasticity thus are 12 to 16 fold of our conservative estimate for the fundamental time scale . A lower bound for the relative duration of juvenile plasticity is thus . According to this estimate , plausible candidate models should thus predict map layout consistent with biological observations in the adult at . The simulations presented in the current studies were designed to assess the convergence of model maps to final attractor states . We presented our results in a way that enables comparison of predicted and biologically observed layouts throughout a broad time span including but also extending beyond this stage . This approach enables to assess the dynamical stability and robustness of the layout obtained . Considering this robustness is not only interesting for theoretical reasons . Experimentally it cannot be excluded that the relative duration of the period of juvenile plasticity is substantially longer that the lower bound estimated above . The fact that the maximal level of plasticity is observed not at eye opening but two weeks later means that similar size changes will unfold on shorter time scales as the peak of the critical period is approached . Several studies have attempted to assess the fundamental time scale for the establishment of stimulus selectivity near the peak of the critical period . Classical studies by Blakemore and Mitchell [96] , and Imbert and coworkers [97]–[99] examined how much visual experience is needed to achieve selectivity from an unselective initial condition at the time of peak OD plasticity . To this end they examined the newly generated selectivity of neurons in kitten dark-reared until the peak of the OD critical period and then given short epochs of normal visual experience . These studies indicate that 6 hours of visual experience are sufficient to induce a substantial degree of selectivity in visual cortical neurons . Recent studies provide further evidence for relevant time scales of visual cortical plasticity on such an accelerated time scale . Mitchell , Sengpiel and coworkers examined how many hours of normal visual experience are sufficient to prevent gross neuronal changes of selectivity and impairment of perceptual abilities under visual deprivation [100]–[103] . They report that two hours of normal visual experience per day are sufficient to completely prevent deprivation induced impairments of visual function . Directly imaging the emergence of direction preference columns in a network initially lacking such columns has been achieved by Li et al . in juvenile ferrets [104] . This study found that even under anesthesia , balanced visual stimulation over 3–6 hours was sufficient to drive the de novo formation of a system of columns . These novel experiments fundamentally differ from previous pairing studies [105]–[108] in that visual cortical neurons were not artificially trained to adopt a particular stimulus preference but were stimulated by a set of opposing motion patterns . Neurons were not artificially activated and were free to develop preference and anti-preference for any stimulus from a balanced set . It appears unlikely that this stimulation paradigm artificially accelerates changes . A corresponding process in the brain of an awake and attending animal is thus expected to be orders of magnitude slower . These studies thus substantially extend prior observations in anesthetized animals in which visual cortical neurons were artificially activated . Also under these more artificial conditions visual cortical preferences for stimulus orientation , direction , or preference for one eye were found to undergo substantial activity induced changes within a few hours [105]–[108] . Various studies thus confirm the basic expectation that around the peak of the critical period the typical time scale for the emergence of selectivity from unselective network states and for changes of selectivity is substantially shorter than one week . In particular , in light of these results it is required to assess the robustness and dynamical stability of model predictions beyond . A pessimistic estimate for the resulting prolongation of realistic simulations can be obtained by assuming that the accelerated time scale is effectively relevant for the most of the 10–14 week period of juvenile plasticity . Assuming h , larger that the times reported in the above experiments , the duration of the period of juvenile plasticity would correspond to 280–370 . The above considerations are a useful reminder that that the current understanding and experimental characterization of the timescales of circuit dynamics are substantially limited . New experimental approaches that provide a more direct and certain assessment of what one might call circuit turnover times , would in fact be very informative for calibrating dynamical and optimization studies . In the absence of such information , it appears also useful to estimate a maximal upper bound to the relative duration of juvenile plasticity that is very unlikely to be ever overturned by future improvements in experimental technology . Such a worst-case estimate is obtained by the ratio of the longest duration ever reported for critical period plasticity in the visual cortex to the low end of the accelerated time scales . To our best knowledge the longest estimate for a critical period in cat visual cortex was obtained by Jones , Spear and Tong [93] . These authors examined juvenile cats deprived at older ages than in the classical studies cited above in an attempt to determine whether visual areas higher in the cortical processing hierarchy exhibit a delayed or extended period of developmental plasticity . They reported that substantial modifications of responses could still be induced up to the 35th postnatal week [93] . Remarkably their results indicate that the period of susceptibility extends longer in the primary visual cortex than in areas higher in the visual processing hierarchy . This would amount to an entire duration of the period of juvenile plasticity of 33 weeks . Assuming a fundamental timescale in the lower range of the experimentally reported peak critical period time scales i . e . 3h one estimates an absolute upper bound for the relative duration of the period of juvenile plasticity in the cat of 1850 . Even in light of the most conservative considerations presented above it appears of limited value to compare maps from a simulation obtained when selectivity first reaches mature levels to biological patterns present in the adult cortex . Maps in the adult visual cortex of the cat have been subject to more than ten weeks of ongoing plasticity . They are thus better viewed as dynamic equilibria that are largely maintained under a continuous process of ongoing activity-driven synaptic turnover . Current experimental evidence , nevertheless , indicates that the maps emerging initially over the first days of normal vision exhibit many layout properties that are preserved throughout the juvenile period of plasticity and into adulthood [11] , [42] . Taking the long duration of the period of juvenile plasticity into account , this is likely to mean that these properties have been actively maintained by an ongoing dynamics . The requirement to generate and maintain a realistic column layout is a more selective criterion for the identification of appropriate candidate models than the mere ability to initially generate good looking maps as demonstrated by our current results as well as many prior theoretical studies ( reviewed in [42] , supplement ) . It is thus a more stringent test of a models explanatory power to compare the maps obtained at later stages , e . g . to the biologically observed functional organization of the visual cortex . Our estimates suggest that it is reasonable to require of a biologically plausible model that states which resemble the adult functional architecture are predicted at least one order of magnitude later than the maturation of average selectivity . Using this criterion , the states observed in our simulations in fact suggest that the considered models are not capable of explaining the biological organization in a satisfying fashion: Patterns observed after 10 intrinsic timescales are dominated by crystalline local arrangements that are distinctly different from the spatially irregular layout of orientation maps observed in both juvenile and adult visual cortex . Quantitatively layout parameters at this time substantially deviate from biological observations . The numerical studies presented here further elucidate the conditions for pinwheel stabilization by map interactions . The analytical results presented in part ( I ) showed that in several models OD stripes are not able to stabilize pinwheels near symmetry breaking threshold and for only one real-valued scalar field . This result appeared to be insensitive of the specific type of inter-map interaction [39] . Our numerical results show that this result is also insensitive to a detuning of typical wavelengths . For different ratios of the typical wavelengths of OP and OD , pinwheel-rich patterns either decay into pinwheel-free OP stripes or patterns with OP fracture lines when interacting with OD stripes . These findings support the conclusion that in models for the joint optimization of OP and OD maps a patchy OD layout is important for pinwheel stabilization by crystallization . In the current study we also generalized our dynamical systems approach to include any additional number of columnar systems . One reason to consider additional visual cortical maps originates from the finding that the removal of the OD map in experiments does not completely destabilize pinwheels [29] . Moreover , in tree shrews , animals which completely lack OD columns , OP maps contain pinwheels and exhibit a pinwheel arrangement essentially indistinguishable from species with columnar OD segregation [13] . This might reflect the influence of additional columnar systems such as spatial frequency columns that can be expected to interact with the OP map in a similar fashion as OD columns [40] . From a theoretical perspective , one might suspect that couplings between more than two systems that promote a mutually orthogonal arrangement are harder to satisfy the more maps are considered . In principle this could lead to the emergence of irregular patterns by frustration . In numerical simulations we examined coordinated optimization with three and four columnar systems . In these cases pinwheel stabilization is possible even without an OD bias . The resulting stationary OP patterns are , however , still either stripes or PWC solutions . For more than two feature maps , asymmetry of one feature dimension is thus not a necessary condition for pinwheel stabilization by coordinated optimization . We also characterized the dynamics of pinwheel crystallization from pinwheel-free initial conditions . With the analytical approach presented in part ( I ) we were able to show that pinwheel-rich solutions correspond to the energetic ground state of our models for large inter-map coupling [39] . This can be confirmed by simulations in which pinwheels are created even when starting from an initial OP stripe pattern . Assessing pinwheel creation from pinwheel-free initial conditions could more generally serve as a simple test for the existence of a pinwheel-rich attractor state in models of OP development that can be applied to models of arbitrary complexity . One should note , however , that the production of pinwheels from a pinwheel-free initial condition provides only a sufficient but not a necessary criterion to verify the existence of a pinwheel-rich attractor state . This criterion may be violated if pinwheel-free and pinwheel-rich attractor states coexist . Nevertheless , the pinwheel production criterion can be used to demonstrate that pinwheels are not just a remnant of random initial conditions . Pinwheel crystals have been previously found in several abstract [109]–[111] as well as in detailed synaptic plasticity based models [81] , [112] , [113] . Remarkably , in a model of receptive field development based on a detailed dynamics of synaptic connections the resulting OP map showed a striking similarity to the hPWC presented above , compare Fig . ( 7 ) and [39] , [112] . These observations indicate that pinwheel crystallization is not an artifact of the highly idealized mathematical approach used here . In fact , the first OP map predicted ever by a synaptically based self-organization model presented by von der Malsburg in 1973 exhibited a clearly hexagonal column arrangement [81] . Von der Malsburgs calculations as well as those presented in [112] utilized a hexagonal grid of cells that may specifically support the formation of hexagonal patterns . Our numerical and analytical results clearly demonstrate that patterns of hexagonal symmetry do not critically depend on the use of a hexagonal grid of cells as our simulations can generate hexagonal patterns also for square lattices of cells . To determine whether the hexagonal layout in Von der Malsburgs model is intrinsically stable it should be implemented for other grids both of square symmetry as well as for irregular positions of cells . In our study , we examined whether the non-crystalline layout of visual cortical maps could result from a detuning of OP and OD wavelengths . However , while destabilizing hPWC solutions , wavelength detuning leads to spatially regular solutions in all studied cases . This suggests that a spatially regular layout is not an exceptional behavior in models for the coordinated optimization of visual cortical maps that would require fine tuning of parameters . Recently , Paik and Ringach have argued that a roughly hexagonal arrangement of iso-orientation domains would provide evidence for a defining role of retinal ganglion cell mosaics for the spatial arrangement of orientation columns [114]–[116] . It is interesting to consider this claim in view of our results as well as in view of the wealth of activity-dependent models that predict hexagonal arrangements irrespective of the arrangement of retinal ganglion cells [66] , [67] , [81] , [112] , [117]–[119] . Since all of these distinct models are known to generate hexagonal arrays of orientation columns it seems questionable to view evidence for a hexagonal arrangement as evidence for a particular activity-independent mechanism . Our characterization of the dynamics of crystallization , however , enables to identify more selective predictions of a retinal ganglion cell mosaic based formation of hexagonal iso-orientation domains . If the pattern of OP columns was seeded by retinal ganglion cell mosaics , as initially proposed by Soodak [120] and recently re-articulated by Paik and Ringach [114] , hexagonal structures should be detectable from the very beginning of development , i . e . already at stages when orientation selectivity is still increasing . Hexagonal PWCs in self-organizing models , in contrast , form from an initially irregular and isotropic state . Thus the time dependence of hexagonal-like column arrangements can distinguish in principle between self-organized as opposed to retinal ganglion cell mosaic imprinted hexagonal arrangements . As we found in all models examined hexagonal arrangements are frequently of rather high pinwheel density of about 5 pinwheels per hypercolumn . Also the hexagonal pattern constructed by Paik and Ringach appear to exhibit relatively high pinwheel density of . Thus both theories appear inconsistent with observed pinwheel densities . A mixed scenario in which retinal ganglion cell mosaics seed the initial pattern of iso-orientation domains and later activity-dependent refinement drives a rearrangement of OP maps towards the experimentally observed design therefore predicts a substantial degree of net pinwheel annihilation . Kaschube et al . presented evidence for essentially age independent pinwheel densities in ferret visual cortex between week 5 and 20 . No indication of substantial pinwheel annihilation is visible in this data [42] . One should note that in ferret visual cortex orientation columns first arise in the fifth postnatal week [11] . The relation of the analysis by Paik and Ringach and the statistical laws described by Kaschube and coworkers ask for further analysis and comparison . The reason for the substantial differences to experimentally observed maps might thus be the presence of biological factors neglected in the models examined here . Candidate factors are a greater distance from the pattern formation threshold , different kinds of biological noise , or the presence of long-range neuronal interactions . Vinals and coworkers demonstrated for the case of stripe patterns that a Swift-Hohenberg model sufficiently far from the bifurcation point can exhibit stable disclination defects [121] . Although the results indicate only a spatially sparse set of stabilized defects it will be interesting to examine whether this also applies to the case of multidimensional coupled models and to establish which properties the model solutions develop very far from the bifurcation point . Theoretically , it is well understood that in principle so called ‘nonadiabatic effects’ can induce the pinning of grain boundaries in pattern forming media [122]–[125] . In one spatial dimension and for models with several interacting order parameter fields similar mechanisms may even lead to the emergence of spatially chaotic solutions [126]–[128] . These studies suggest to examine whether spatial incommensurability far from threshold can induce spatially chaotic patterns in one and two dimensional coordinated optimization models . Such studies may uncover a completely novel scenario for explaining the emergence of spatially irregular states in models of cortical map optimization . A second interesting direction will be the inclusion of frozen spatial disorder in models for the self-organization of multiple cortical maps . Such disorder could represent a temporally fixed selectivity bias that favors particular feature combinations at different position in the cortical sheet . For OP , the proposals of Waessle and Soodak recently revisited by Ringach and coworkers that retinal ganglion mosaics might constrain and seed orientation column patterns would represent a specific mechanisms for such a fixed local bias [114]–[116] , [120] , [129] . Experimental evidence that retinal organization can impose local biases was revealed by Adams and Horton's finding that the pattern of retinal blood vessels can specifically determine the layout of OD columns in squirrel monkey visual cortex ( [130] , [131] , for a modeling study see also [132] ) . Spatial disorder terms in dynamical models can also be designed to represent randomness in the interactions between neurons at different positions . This might result from heterogeneities in lateral interactions within the cortical sheet . In particular this later type of modification has been examined in simple examples of order parameter equations and was found to qualitatively change the type of the bifurcation and the nature of the unstable modes [133]–[136] . It will be important to investigate how different types of spatial disorder modify the behavior of models derived from biologically meaningful energy functionals . We hope that for such studies of the influence of ‘biological noise’ a thorough understanding of the properties of perfectly homogeneous and isotropic systems as achieved here will provide a solid basis for disentangling the specific contributions of randomness and self-organization . Finally , a third promising direction for modifying the type of models considered here is the inclusion of long-ranging intra-cortical interactions in the equations for the individual order parameter fields . The impact of long-ranging intra-cortical interactions has been studied previously both with respect to the properties of patterns emerging during the phase of initial symmetry breaking [32] , [137]–[139] as well as for its influence on long-term pinwheel stabilization and pattern selection [66] , [67] , [140] . Models for orientation maps that include orientation-selective long-ranging interactions exhibit a good quantitative agreement of both attractors and transient states to the biological organization of OP maps in the visual cortex [42] . Including orientation-selective long-ranging interactions in models for the coordinated optimization of multiple cortical maps could provide a transparent route towards constructing improved models for the coordinated optimization of column layouts matching the spatial structure of orientation maps . Viewed from a practical perspective , the presented theoretical approach offers also convenient ways to model the impact of spatial inhomogeneities in the visual cortex on OP map structure . For this purpose , the co-evolving field does not represent a feature map but would be designed to describe a real or artificial areal border or a disruption of local circuitry . To this end , the OP map would be coupled , using low order coupling energies , to a fixed field describing the areal border such that its values are for instance one inside and minus one outside of the area with a steep gradient interpolating between the two . Outside the areal border a strong coupling to such a field can lead to complete suppression of orientation selectivity . Using a gradient-type inter-map coupling energy inter-map coupling can also be used to favor a perpendicular intersection of iso-orientation lines with the areal borders as observed in some experiments [13] , [141] . Artificial heterogeneities and areal borders have been induced by local ablation or other local interventions [141] , [142] . Viral approaches such as the silencing of cortical regions by transfection with hyperpolarizing ion-channels now make it possible to impose such heterogeneities with minimal intervention and potentially in a reversible fashion [143] , [144] . In the current studies we focused on a particular hierarchy of visual cortical maps . In the analytical calculations and most simulations the OD map was assumed to be dominant which corresponds to a choice of control parameters that satisfy . That maps form a hierarchy under such conditions can be seen from the limiting case in which inter-map interactions become effectively unidirectional . In this case the dynamics of the OP map is influenced by OD segregation while the OD dynamics is effectively autonomous . This limit substantially simplifies the analysis of map-interactions and the identification of ground states . The effect of a backreaction on the OD map can be studied within the presented approach either by solving amplitude equations numerically or by solving the full field dynamics . We observed that , although the presented optima persist , with increasing backreaction on the OD map the minimum inter-map coupling strength necessary for the stability of hexagonal pinwheel crystals increases . By solving the full field dynamics numerically we confirmed this conclusion . In the presented numerical simulations the backreaction , however , was relatively small . The simulations nevertheless establish that our results are not restricted to the limit i . e . that limit is not singular . A comprehensive analysis of the effect of strong backreaction is beyond the scope of the current study . One should note that the decoupling limit does not lead to completely unrealistic OD patterns . In particular , compared to the architecture of macaque visual cortex the uncoupled OD dynamics has stationary patterns which qualitatively resemble the layout of observed OD maps . Macaque primary visual cortex appears to exhibit essentially three different kinds of OD patterns: Fairly regular arrays of OD stripes in most of the binocular part of the visual field representation , a pattern of ipsilateral eye patches in a contralateral background near the transition zone to the monocular segment and of course a monocular representation in the far periphery . These qualitatively correspond to the three fundamental solutions of the OD equation: stripes , hexagons , and a constant solution , which are stable depending on the OD bias ( see Fig . 16 of [39] ) . In cat visual cortex the observed OD layout is patchy throughout V1 [49] , [145]–[148] . The presented models for the coordinated optimization of maps in the visual cortex , that were studied analytically in part ( I ) and numerically in part ( II ) , lead in all studied conditions to spatially regular energy minima . In local regions on the order of a few hypercolumn areas , column layout rapidly converges to one of a few types of regular repetitive layouts . Because of this behavior the considered models cannot robustly explain the experimentally observed spatially irregular common design of OP maps in the visual cortex . As expected from these qualitative differences all pinwheel statistics considered , exhibit substantial quantitative deviations from the experimentally observed values . These findings appear robust with respect to finite backreaction , detuning of characteristic wavelengths , and the addition of further feature space dimensions . Recent work demonstrated that the spatially irregular pinwheel-rich layout of pinwheels and orientation columns in the visual cortex can be reproduced quantitatively by models that represent only the orientation map but include long-range interactions [42] , [66] , [67] . In the visual cortex of species widely separated in mammalian evolution we previously found virtually indistinguishable layout rules of orientation columns that are quantitatively fulfilled with a precision of a few percent [42] . In view of these findings the current results suggest that models in which pinwheel stabilization is achieved solely by coordinated optimization and strong inter-map coupling are not promising candidates for explaining visual cortical architecture . In order to achieve a quantitatively more viable coordinated optimization theory one might consider taking additional ‘random’ factors into account . One should however not conclude that coordinated optimization does not shape visual cortical architecture . Using the general approach developed here it is possible to construct a complementary type of models in which the complex OP map is dominant . Such models , using non-local terms in the energy functional of the OP map , can be constructed to reduce to the model in [42] , [66] , [67] in the weak coupling limit . Because long-range interaction dominated models can reproduce the spatially irregular layout of OP maps , one expects from such models a better reproduction of the observed architecture for weak coupling . Alternative scenarios might emerge from the inclusion of quenched disorder or very far from the pattern formation threshold . Because of its mathematical transparency and tractability the approach developed in the present studies will provide powerful tools for examining to which extend such models are robust against coupling to other cortical maps and to disentangle the specific contribution of coordinated optimization to visual cortical architecture .
During the evolution of OD and OP maps we monitored the states from the initial time to the final time using about 150 time frames . To account for the various temporal scales the dynamics encounters the time frames were separated by exponentially increasing time intervals . Pinwheel centers were identified as the crossing of the zero contour lines of the real and imaginary parts of . During time evolution we tracked all the pinwheel positions and , as the pinwheels carry a topological charge , we divided the pinwheels according to their charge . The pinwheel density is defined as the number of pinwheels per unit area . By this definition , the pinhweel density is independent of the spacing of columns and dimension-less . The distribution of pinwheel distances indicates the regularity and periodicity of the maps . Therefore we calculated the minimal distance between pinwheels , measured in units of the column spacing , during time evolution . In simulations we used periodic boundary conditions . In order to correctly treat pinwheels close to map borders we periodically continued the maps . Nearest neighbors of pinwheels are thus searched also in the corresponding periodically continued maps . To calculate pinwheel density variability in subregions of size we sampled for each map circular shaped regions of various size and placed their centers at random locations of the map . Sizes of circular regions were uniformly distributed . To calculate pinwheel density variability for a given area , we randomly selected from all regions in the set up to 1000 regions with size in the interval where , and calculated the standard deviation SD of pinwheel densities . To characterize density variability as a function of area size we estimated the variability coefficient and the exponent by fitting the function to the SD ( A ) -curves . The rearrangement of OP maps leads to annihilation and creation of pinwheels in pairs . Between two time frames at and we identified corresponding pinwheels if their positions differed by less than and carry the same topological charge . If no corresponding pinwheel was found within it was considered as annihilated . If a pinwheel at could not be assigned to one at it was considered as created . We define the pinwheel creation and annihilation rates per hypercolumn as ( 14 ) where and are the numbers of created and annihilated pinwheels . Creation and annihilation rates were confirmed by doubling the number of time frames . To what extend are the pinwheels of the final pattern just rearrangements of pinwheels at some given time ? To answer this question for a given set of pinwheels at an initial time we further calculated the fraction of those pinwheels surviving until time . Finally , the fraction of pinwheels present at time that survive up to the final time is given by . As the Swift-Hohenberg equation is a stiff partial differential equation we used a fully implicit integrator [149] . Such an integration scheme avoids numerical instabilities and enables the use of increasing step sizes when approaching an attractor state . The equation ( 15 ) is discretized in time . Using a Crank-Nicolson scheme this differential equation is approximated by the nonlinear difference equation ( 16 ) This equation is solved iteratively for with the help of the Newton method which finds the root of the function ( 17 ) The field is discretized . For a grid with meshpoints in the -direction and meshpoints in the -direction this leads to an dimensional state vector . Discretization is performed in Fourier space . The Newton iteration at step is then given by ( 18 ) with the Jacobian of . Instead of calculating the matrix explicitly a matrix free method is used , where the action of the matrix is approximated using finite differences . To solve the linear system with , we used the Krylov subspace method [149] . The Krylov subspace of dimensionality is defined as ( 19 ) In the Generalized Minimum Residual ( GMRES ) algorithm the Krylov subspace is generated by with , and an initial guess , see [149] . After iterations , the refined solution is given by ( 20 ) where the matrix has the base vectors of the Krylov subspace as its columns . The vector is chosen by minimizing the residuum ( 21 ) where denotes the Euclidean norm . For this procedure an orthonormal basis of the Krylov subspace is generated with an Arnoldi process . With the use of the similarity transformation ( 22 ) where is an upper Hessenberg matrix , , and the orthogonality of , the optimality condition Eq . ( 21 ) becomes ( 23 ) with the first unit vector of dimension . For a that minimizes this norm the approximate solution is given by . To improve the convergence of this iterative method preconditioning was used . A preconditioner is multiplied to such that is close to unity . A preconditioner suitable for our model is the inverse of the linear operator in Fourier space with a small shift in order to avoid singularities i . e . ( 24 ) The convergence of Newton's method is only guaranteed from a starting point close enough to a solution . In the integration scheme we use a line search method to ensure also a global convergence [150] . Newton's method Eq . ( 18 ) is thus modified as ( 25 ) where the function ( 26 ) is iteratively minimized with respect to . This integrator was implemented using the PetSc library [151] . As the dynamics converges towards an attractor an adaptive stepsize control is very efficient . The employed adaptive stepsize control was implemented as described in [152] . The described integration scheme has been generalized for an arbitrary number of real or complex fields . The coupling terms are treated as additional nonlinearities in . As a common intrinsic timescale we choose with the bifurcation parameter of the OP map . Due to the spatial discretization not all points of the critical circle lie on the grid . Thus , the maximal growth rate on the discretized circle is not exactly equal to , the theoretical growth rate . In particular , some modes may be suppressed or even become unstable . Due to this we expect deviations from analytical solutions . To minimize such deviations the size of the critical circle was chosen such that this disbalance between the active modes was minimized . Periodic boundary conditions were applied to account for the translation invariance of the spatial pattern .
|
Neurons in the visual cortex of carnivores , primates and their close relatives form spatial representations or maps of multiple stimulus features . In part ( I ) of this study we theoretically predicted maps that are optima of a variety of optimization principles . When analyzing the joint optimization of two interacting maps we showed that for different optimization principles the resulting optima show a stereotyped , spatially perfectly periodic layout . Experimental maps , however , are much more irregular . In particular , in case of orientation columns it was found that different species show apparently species invariant statistics of point defects , so-called pinwheels . In this paper , we numerically investigate whether the spatial features of the stereotyped optima described in part ( I ) are expressed on biologically relevant timescales and whether other , spatially irregular , long-living states emerge that better reproduce the experimentally observed statistical properties of orientation maps . Moreover , we explore whether the coordinated optimization of more than two maps can lead to spatially irregular optima .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"physics",
"general",
"physics",
"computational",
"neuroscience",
"interdisciplinary",
"physics",
"biology",
"computational",
"biology",
"sensory",
"systems"
] |
2012
|
Coordinated Optimization of Visual Cortical Maps (II) Numerical Studies
|
Intestinal Listeria monocytogenes infection is not efficient in mice and this has been attributed to a low affinity interaction between the bacterial surface protein InlA and E-cadherin on murine intestinal epithelial cells . Previous studies using either transgenic mice expressing human E-cadherin or mouse-adapted L . monocytogenes expressing a modified InlA protein ( InlAm ) with high affinity for murine E-cadherin showed increased efficiency of intragastric infection . However , the large inocula used in these studies disseminated to the spleen and liver rapidly , resulting in a lethal systemic infection that made it difficult to define the natural course of intestinal infection . We describe here a novel mouse model of oral listeriosis that closely mimics all phases of human disease: ( 1 ) ingestion of contaminated food , ( 2 ) a distinct period of time during which L . monocytogenes colonize only the intestines , ( 3 ) varying degrees of systemic spread in susceptible vs . resistant mice , and ( 4 ) late stage spread to the brain . Using this natural feeding model , we showed that the type of food , the time of day when feeding occurred , and mouse gender each affected susceptibility to L . monocytogenes infection . Co-infection studies using L . monocytogenes strains that expressed either a high affinity ligand for E-cadherin ( InlAm ) , a low affinity ligand ( wild type InlA from Lm EGDe ) , or no InlA ( ΔinlA ) showed that InlA was not required to establish intestinal infection in mice . However , expression of InlAm significantly increased bacterial persistence in the underlying lamina propria and greatly enhanced dissemination to the mesenteric lymph nodes . Thus , these studies revealed a previously uncharacterized role for InlA in facilitating systemic spread via the lymphatic system after invasion of the gut mucosa .
L . monocytogenes are facultative intracellular bacteria that cause food borne disease in humans ranging in severity from mild , self-limiting gastroenteritis to life-threatening sepsis and meningoencephalitis [1]–[3] . The factors that determine host resistance to intestinal infection and subsequent systemic spread of L . monocytogenes are not well understood , primarily due to the lack of a suitable small animal model . Oral infection of mice , for example , is widely perceived to be inefficient , requiring an inoculum of 109–1011 bacteria , and is typically not as reproducible as intravenous ( i . v . ) infection . The low infectivity of L . monocytogenes in the gut has long been attributed to a weak interaction between the bacterial surface protein internalin A ( InlA ) and E-cadherin , a cell adhesion protein expressed on intestinal epithelial cells . InlA has a high affinity for human , rabbit and guinea pig E-cadherin , but does not interact strongly with E-cadherin in rodents [4] . L . monocytogenes can directly invade intestinal epithelial cells in vitro using a “zipper mechanism” triggered by the binding of InlA to E-cadherin [5] , [6] . Pentecost et al . showed that basolaterally expressed E-cadherin was transiently exposed at the tips of intestinal villi as dying cells were extruded from the epithelium , and that L . monocytogenes preferentially bound at the multicellular junctions where this occurred [7] . More recently , Nikitas et al . showed that E-cadherin is also luminally accessible near mucus expelling goblet cells [8] . However , other routes of invasion in the gastrointestinal tract are also possible . Many pathogens are transcytosed across the epithelium by M cells located both in macroscopically visible Peyer's Patches and scattered elsewhere throughout intestinal villi [9] , [10] . L . monocytogenes were shown to associate with murine M cells both in vivo and in vitro [11]–[15] and internalin B ( InlB ) was implicated in this process [16] . The bacterial adhesins LAP and Vip have also been implicated in translocation across the gut mucosa [17] , [18] . Two approaches have been used to improve the efficiency of oral infection in mice , each focused on modeling the interaction between human E-cadherin and InlA . In one approach , transgenic mice expressing both murine and human E-cadherin were generated . In that study , InlA had the greatest effect on colonization in the cecum and colon of the humanized mice , but importantly , intragastric ( i . g . ) inoculation of an InlA deletion mutant still resulted in significant colonization of intestinal tissues [19] . As an alternate approach , Wollert et al . generated mouse-adapted L . monocytogenes expressing a modified InlA ( InlAm ) that bound mouse E-cadherin with the same affinity as for the wild type InlA::human E-cadherin interaction [20] . Infection with 107 CFU ( a dose 100-fold lower than typically used ) was possible with this strain; however , in that study , no significant difference in intestinal colonization was observed for the InlAm-expressing bacteria compared to wildtype L . monocytogenes until 72 hours after i . g . inoculation . Both of these approaches suggested that a high affinity interaction between InlA and E-cadherin was not required to breach the intestinal barrier , and hinted at a possible role for InlA during the later stages of intestinal infection . However , the high degree of variability in bacterial loads after i . g inoculation of L . monocytogenes InlAm ( up to 1000-fold difference within the same experimental group ) made it difficult to distinguish clear phenotypes [20] . The fate of microbes delivered by oral gavage , a process that puts organisms suspended in saline directly into the stomach via a feeding needle , is not well understood , despite its widespread use in models of oral infection [21] . In many reports , i . g administration of L . monocytogenes resulted in rapid spread to the spleen and liver within 4–12 hours of inoculation , regardless of the bacterial isolate or mouse strain used [20] , [22]–[25] . However , in some studies , no systemic spread was observed until 48 hours post-infection ( hpi ) [14] , [26] , [27] . The reason for this variable rate of systemic spread is not known , but could be related to the invasive nature of i . g . inoculation , since minor trauma in the esophagus or stomach could facilitate a mechanism of direct bloodstream invasion . In support of this idea , Kinder et al . recently showed that mice fed from a syringe in the mouth were able to generate oral tolerance against ovalbumin , but mice treated by gavage with a feeding needle were not tolerant and instead generated an ovalbumin-specific systemic antibody response ( Heather Bruns , personal communication ) . Rapid spread to the bloodstream is problematic because L . monocytogenes does not need a prolonged period of incubation in the host to begin intracellular replication , and the high inoculum typically used to promote intestinal colonization ( 107–1011 CFU ) is several orders of magnitude greater than the systemic lethal dose for mice . It is not currently known how long it takes L . monocytogenes to translocate across the intestinal mucosa and spread to peripheral tissues after natural ingestion of contaminated food in either mice or humans . To clearly delineate the role of InlA during intestinal infection , a better model of oral transmission was needed that relied solely on translocation across the gut mucosa , without the complications that arise from a rapid direct bloodstream invasion . In this paper , we report the development of a food borne model of murine listeriosis that consistently results in a 36–48 h period of infection only in the gastrointestinal tract , followed by varying degrees of systemic spread in susceptible BALB/c/By/J ( BALB ) versus resistant C57BL/6/J ( B6 ) mice . Using this non-invasive natural feeding model , we showed that InlA was not required for early colonization of the murine intestines . However , the mouse-adapted InlAm did promote bacterial persistence in the underlying lamina propria and enhanced dissemination to both the mesenteric lymph nodes and spleen , but not the liver .
The first goal of this study was to develop an improved murine model that could be used to clarify the role of InlA in establishing intestinal infection following oral transmission of L . monocytogenes . The most important criterion for the model was a reproducible phase of gastrointestinal infection that preceded systemic spread , to ensure that all bacteria from the initial inoculum that reached the spleen or liver had translocated across the gut mucosa . Our initial strategy was to use i . g . infection at doses lower than previously reported to avoid overwhelming innate resistance mechanisms in the gut and inadvertently facilitating rapid , direct bloodstream invasion by L . monocytogenes . Groups of BALB/c/By/J ( BALB ) and C57BL/6J ( B6 ) mice were infected with 103 , 104 , 105 , or 106 CFU of mouse-adapted L . monocytogenes that expressed a modified InlA protein ( InlAm ) [20] . The total number of L . monocytogenes present in either the small intestine ( Fig . 1A ) or the large intestine ( not shown ) 24 hours post infection ( hpi ) was proportional to the inoculum given , with a lower limit of approximately 104 CFU for establishing infection in either mouse strain . However , when mice were inoculated with only 104 Lm InlAm , very few cell-associated bacteria ( adherent or intracellular organisms not removed by extensive flushing of the lumen ) could be recovered from these tissues ( Fig . 1C ) . Using higher inocula , approximately equal amounts of luminal and cell-associated L . monocytogenes were recovered from each tissue . A dose of at least 106 CFU of Lm InlAm was required to achieve consistent intestinal infection in all inoculated mice ( Fig . 1A ) . This inoculum was 10-fold lower than used in the original study published by Wollert et al . [20] . However , i . g . challenge with 106 CFU of Lm InlAm resulted in rapid dissemination to both the spleen and liver in all mice tested ( Fig . 1D ) . The route of systemic spread was not likely to have occurred via the lymphatic system since L . monocytogenes were found in the draining mesenteric lymph nodes ( MLN ) of only a few mice ( Fig . 1B ) . Therefore , lowering the challenge dose did not prevent rapid spread of L . monocytogenes after i . g . inoculation . Since we suspected that i . g . inoculation with a feeding needle was facilitating a direct mechanism of bloodstream invasion that might not be physiologically relevant to human food borne disease , we next set out to develop a less invasive means of orally inoculating mice . To do this , small pieces of Lm InlAm-contaminated bread were placed in empty cages , and each mouse was allowed to pick up the food and eat it voluntarily . Using this method , at least 108 CFU were required to see cell-associated L . monocytogenes in the small intestines of B6 mice ( Fig S1A ) . Importantly , at 24 hpi , there were no bacteria in the spleens or livers of mice fed L . monocytogenes ( Fig . S1B ) . Thus , infection by natural feeding resulted in a distinct gastrointestinal phase of infection prior to systemic spread of L . monocytogenes . However , intestinal infection was not uniformly observed in all of the mice fed contaminated food , so further optimization of the natural feeding protocol was required . The composition of gastric secretions varies depending on the type of food ingested , so it was possible that bacterial survival in the gut could vary depending on nature of the contaminated food used for oral transmission . To test this , fecal shedding was monitored in mice fed bread saturated with Lm InlAm suspended in either PBS , a glucose solution , or butter . Three hours after ingestion , 50-fold more L . monocytogenes had survived passage through the stomach when butter or PBS was used compared with glucose ( Fig . S1C ) . Furthermore , there was a significantly greater number of cell-associated L . monocytogenes in the colon 24 hpi , and less variation among mice when butter was used ( Fig S1D ) . Thus , for all subsequent experiments , mice were fed bread saturated with L . monocytogenes-contaminated butter . Mice are typically denied food 16–24 hours prior to i . g . infection to ensure that the stomach is empty enough to receive a 200–500 µl bacterial inoculum . Although this was not necessary for infection by natural feeding , in our pilot studies , mice were fasted to facilitate optimal comparison to the i . g . infection route . To find out if fasting was required for food borne transmission of L . monocytogenes , BALB mice were denied food for either 0 , 4 , or 16 hours . With 0–4 hours of food restriction , only a few of the mice had cell-associated L . monocytogenes in the intestines ( Fig . S1E ) . In contrast , when food was denied overnight ( 16 h ) , the number of bacteria shed in the feces 3 hpi increased 10 to 100-fold , and importantly , cell-associated Lm InlAm were present in the majority of small intestines and in the colons of all mice tested . Thus , in all subsequent experiments , mice were denied food overnight prior to ingestion of Listeria-contaminated bread . Using the optimized parameters , the course of food borne Lm InlAm infection was followed in both BALB and B6 mice , strains that are known to have significantly different susceptibility to i . v . challenge with L . monocytogenes [28] . Similar loads of cell-associated bacteria were found in the intestines of both mouse strains 24 h after ingestion of L . monocytogenes ( Fig . 2A ) . In B6 mice , clearance initiated rapidly , with both cell-associated bacteria and L . monocytogenes shed in the feces ( Fig . 2C ) completely eliminated within 5 to 8 dpi . In contrast , L . monocytogenes grew exponentially in the intestines of BALB mice ( Fig . 2A ) . The organisms persisted in the colon , and fecal shedding of L . monocytogenes was still detected 8 dpi in BALB mice ( Fig . 2C ) . Thus , resistant B6 mice had a mild , self-limiting gastrointestinal infection while susceptible BALB mice had a more severe infection that persisted in the colon . By 48 hpi , L . monocytogenes had disseminated from the gut to both the spleen and liver ( Fig . 2B ) . The bacterial loads were identical in the spleen and only slightly higher in the livers of BALB mice compared with B6 mice . However , as seen in the intestines , the infection was rapidly cleared in B6 mice , while L . monocytogenes continued to replicate exponentially in the organs of BALB mice . The peak of infection occurred 5 days after ingestion of contaminated food , when BALB mice had 100 , 000-fold more Lm InlAm in the liver compared with B6 mice ( Fig . 2B ) . These results suggested that B6 mice had innate resistance mechanisms that could rapidly inhibit the growth of L . monocytogenes , and that these mechanisms appeared to be lacking or deficient in BALB mice . Therefore , BALB mice were the preferred strain to test for virulence of L . monocytogenes following food borne transmission . During the feeding sessions , the B6 mice were generally more receptive to eating contaminated bread pieces than BALB mice . In any given experiment , up to 50% of the BALB mice would not pick up the bread and eat it within 1 hour , while all of the B6 mice ate it within 5–10 minutes . Prior studies showed that BALB mice have a strong food anticipatory rhythm and feed primarily at night [29] , [30] , but all of our infections had occurred at approximately noon , a time point midway through the 14 hour light cycle for the animals . Reasoning that BALB mice might be more receptive to feeding at night , L . monocytogenes-contaminated bread pieces were offered just after the onset of the dark cycle ( ∼9:30 PM ) . As expected , both mouse strains readily ate the contaminated food within several minutes during the night feedings . To find out if night feeding altered the course of L . monocytogenes infection , groups of BALB and B6 mice were fed contaminated bread at either noon or 9:30 PM , and bacterial loads were determined 5 dpi . This time point was chosen because it represented the peak of bacterial growth in susceptible BALB mice after noontime feedings , and resistant B6 mice had typically cleared the infection by this point . Night feeding did not significantly alter intestinal infection in BALB mice ( Fig . 2D ) . However , increased bacterial loads were observed in both the spleen and liver of BALB mice infected at night . In contrast , B6 mice were uniformly less resistant to infection in all tissues examined when fed L . monocytogenes-contaminated food at night ( Fig . 2D ) . The increased susceptibility of B6 mice was not related to initial colonization rates , as mice infected during the day ( Fig . 2A ) and mice infected at night ( Fig . 2E ) both had approximately 102 CFU of Lm InlAm in either the small intestine or the colon 24 hpi . The key difference for B6 mice infected at night was that the number of cell-associated L . monocytogenes in the gut increased between days 1 and 3 post-infection , prior to the onset of clearance that initiated by 5 dpi ( Fig . 2E ) . This suggested that innate resistance mechanisms in the B6 gut normally capable of inhibiting the rapid exponential growth of L . monocytogenes were either delayed or not activated when the food borne transmission of infection occurred at night . Growth curves in the spleen were similar whether mice were infected during the day or at night , with the greatest difference between BALB and B6 mice occurring 5 dpi ( Fig . 2F ) . In the liver , the largest difference between mouse stains was delayed until 7 dpi , when L . monocytogenes had been cleared from the livers of B6 mice and BALB mice had an average of 2 . 34×105 CFU per liver ( Fig . 2F ) . Prolonged growth in the spleen and liver is thought to lead to a secondary wave of bacteremia and further systemic spread of L . monocytogenes to tissues such as the brain [31] . We were unable to detect L . monocytogenes in the brains of most mice fed 108 CFU . However , infection with 109 CFU did result in spread of L . monocytogenes to the brain following either noontime ( not shown ) or night ( Fig . 2I ) feeding . Preliminary studies indicated that a feeding dose of 5×109 CFU was near the LD50 for BALB mice ( Fig . S2 ) . We concluded from these results that B6 mice could readily be infected at any time of day , but were most resistant to infection when fed L . monocytogenes-contaminated food at approximately midday . For susceptible BALB mice , time of day did not significantly alter the course of infection , but night infection was preferable since the animals were more receptive to feeding during their dark cycle . Hardy et al . recently showed that extracellular L . monocytogenes accumulated in the gall bladders of BALB mice infected by the i . g route , and that the bacteria could be excreted back into the intestines upon subsequent feeding [32] , [33] . Using the food borne model of infection , we also found significant numbers of L . monocytogenes in the gall bladders of BALB/c/By/J mice ( Fig . 2G & 2H ) . However , we were unable to detect L . monocytogenes in the gall bladders of resistant B6 mice that were fed L . monocytogenes during the day ( Fig . 2H ) . Following night feeding , a few bacteria were found in the B6 gall bladder beginning at 5 dpi , but the bacterial load did not increase significantly over the next two days ( Fig . 2G ) . In contrast , L . monocytogenes increased more than 1000-fold from 3 to 5 dpi in BALB gall bladders . These data suggest that B6 mice have innate resistance mechanisms that can prevent dissemination and possibly extracellular growth of L . monocytogenes in the gall bladder . Furthermore , the rapid exponential growth of L . monocytogenes in the gall bladders of BALB mice may contribute to the persistence of intestinal infection that we observed following either day or night feeding . Pasche et al . previously showed that female mice were more susceptible to infection than males , as measured by both survival and colony counts after i . v . injection of L . monocytogenes [34] . To test whether females were also more susceptible to food borne infection , Lm InlAm-contaminated bread was fed to groups of BALB and B6 mice and the number of CFU present in various tissues 5 days later was determined . In BALB mice , significantly greater numbers of L . monocytogenes were recovered from females , with at least 100-fold higher loads in the gut , spleen , liver , and brain ( Fig . 3 ) . The greatest difference was observed in the gallbladder , with at least 10 , 000-fold more L . monocytogenes found in female tissues compared with males . In B6 mice , slightly higher numbers of Lm InlAm were recovered from female tissues; however , the only significant difference occurred in the spleen ( Fig . 3 ) . Thus , gender was a key factor influencing the susceptibility of BALB mice , but did not contribute significantly to resistance in B6 mice . Having established that female BALB mice fed at night were most prone to infection , we then tested whether expression of the mouse adapted InlAm was required for intestinal colonization in mice following food borne transmission of L . monocytogenes . To do this , we assessed the ability of the mouse-adapted strain ( expressing the modified InlAm ) to compete with either wild type Lm EGDe ( a strain that expressed an InlA protein with a low affinity for murine E-cadherin ) or Lm ΔinlA ( a deletion mutant strain that lacked InlA ) in mice co-infected with two L . monocytogenes strains at a 1∶1 ratio . The number of cell-associated CFU in the gut was determined at both an early ( 16 h ) and later ( 60 h ) time point during the infection . Only the terminal third of the small intestine ( approximating the ileum ) was examined for these experiments , because a pilot study showed very little colonization of either the duodenum or jejunum using the natural feeding model ( Fig . S3 ) . Since InlA is proposed to enhance the efficiency of intestinal infection by promoting rapid invasion of enterocytes and goblet cells [8] , [24] , we predicted that significantly more Lm InlAm would be recovered at the early time point . Instead , we found that the mouse-adapted L . monocytogenes strain had only a slight advantage in colonization of the intestines at 16 hpi . In the colon , only 5-fold fewer wild type Lm EGDe were recovered compared to Lm InlAm and the inlA deletion mutant had just a 2-fold defect ( Fig . 4 ) . The greatest difference was observed in the ileum , where on average , the mouse adapted InlAm strain outcompeted the deletion mutant by 10-fold . It is unlikely that co-infection with InlAm-expressing bacteria promoted invasion of the other strains because similar bacterial titers were observed when mice were infected singly with only Lm ΔinlA or wt EGDe ( data not shown ) . These results indicated that L . monocytogenes lacking a high affinity ligand for E-cadherin could readily establish intestinal infection following ingestion of contaminated food . Therefore , the mouse adapted InlAm protein was not an essential factor needed for food borne transmission of L . monocytogenes . By 60 hpi , the colonization defect for wildtype Lm EGDe had increased four- to five-fold in both the ileum and the colon ( Fig . 4 ) . The greatest difference was observed in the colon , where the InlAm strain outcompeted the wild type by 27-fold . Surprisingly , the colonization defect for the inlA deletion strain did not change significantly from 16 to 60 hpi ( Fig . 4 ) . These data suggested that the mouse-adapted L . monocytogenes strain did not have an intrinsic growth advantage , but rather that bacteria expressing the wild type InlA were impaired for either growth or persistence in the intestines . Because the intestinal colonization defect for wild type L . monocytogenes increased over time , we hypothesized that expression of the low affinity ligand for E-cadherin impaired the ability of the bacteria to gain access to an intracellular niche that would allow for both replication and dissemination . To find out if L . monocytogenes expressing the mouse-adapted InlAm had a competitive advantage for systemic spread from the gut , female BALB mice were co-infected with Lm InlAm and Lm EGDe and the bacterial loads in the MLN , spleen , liver , and gall bladder were quantified at various time points post-infection . By 36 hpi , small numbers of both L . monocytogenes strains had trafficked to the MLN in each mouse tested ( Fig . 5B ) . One day later , at 60 hpi , InlAm-expressing L . monocytogenes outnumbered wild type Lm EGDe by an average of 34-fold ( Fig . 5A ) . A closer examination of the bacterial loads 60 hpi revealed a strikingly consistent number of Lm InlAm in the MLN of each mouse tested ( Fig . 5A ) . In contrast , wild type L . monocytogenes had a bi-modal distribution . In 6 out of 23 mice tested ( 26% ) , little or no Lm EGDe was recovered , while the remaining 17 mice had bacterial loads that were 10 to 100-fold lower than Lm InlAm . In mice co-infected with Lm InlAm and the ΔinlA mutant , the deletion strain also had a significant colonization defect ( Fig . 5A ) , and showed the same bi-modal distribution ( Fig . 5C ) in the MLN . These data strongly suggested that the mouse-adapted InlAm promoted , but was not required , for passage through a key bottleneck to exit the intestinal lamina propria and traffic to the draining lymph nodes . After reaching the MLN , bacteria can transit further via the lymphatic system , eventually draining into the blood , where they are rapidly filtered in either the spleen or liver . In the spleens of the co-infected mice , InlAm-expressing L . monocytogenes had a 13-fold advantage over wild type Lm EGDe ( Fig . 5A ) , and the wild type bacteria had bi-modal distribution similar to that observed in the MLN ( Fig . 5B ) . Wollert et al . previously showed that InlAm-expressing L . monocytogenes had no growth or survival advantage compared to wild type bacteria in the spleen following i . v . inoculation [20] , thus , this result suggests that the low affinity InlA expressed by wild type L . monocytogenes impaired dissemination from the MLN to the spleen . In contrast , the inlA deletion mutant showed no significant colonization defect in the spleen compared to InlAm expressing bacteria ( Fig . 5A , C ) . InlA had little impact on spread to the liver . The competitive indexes for co-infected mice showed only a two-fold ( ΔinlA ) or three-fold ( wt EGDe ) colonization defect ( Fig . 5A ) and there was no significant difference in the actual number of CFU recovered from the liver ( Fig . 5B , C ) . Similar patterns of spread to the MLN , spleen , and liver were also observed in resistant B6 mice ( Fig S4C ) . Routes of dissemination to the gall bladder are not well understood , and extensive colonization of this organ typically does not occur until 5 days after infection of susceptible BALB mice ( Fig . 2 ) . However , even at 60 hpi , we found a significantly greater number of InlAm expressing L . monocytogenes in the gall bladder compared with wild type Lm EGDe ( Fig . 5 ) . Together , these results suggested that expression of wildtype InlA impaired systemic spread to tissues of the lymphatic system ( MLN and spleen ) and the gall bladder , but did not impact efficient dissemination to the liver . The striking distribution of L . monocytogenes strains in the MLN 60 hpi suggested that the mouse-adapted InlAm protein could enhance the ability of the bacteria to disseminate from the gut to the draining lymph nodes . L . monocytogenes could traffic extracellularly in the lymph directly to the MLN , or be carried inside migratory phagocytes such as dendritic cells . This type of “stealth transport” is thought to augment dissemination by protecting the bacteria from mechanisms of immune clearance [35]–[37] . If expression of InlAm helped to promote transit of L . monocytogenes inside a migratory phagocyte , then one would expect to find a larger number of InlAm-expressing bacteria residing in cells within the lamina propria underlying the intestinal epithelium . To test this , we needed to be able to quantify the total number of bacteria in each compartment of the intestinal tissue . Since microscopic approaches would only identify foci of infection , and not the total number of bacteria , we chose to modify enzymatic digestion methods routinely used for the isolation of intestinal lymphocytes [38] , [39] to separate gut tissues into three fractions: the mucus layer , epithelial cells ( EC ) , and lamina propria ( LP ) cells ( Fig . 6A ) . Single cell suspensions of either EC or LP cells were treated with 25 µg/ml gentamicin and then lysed to define the total number of intracellular L . monocytogenes in each cell type . During the processing of each fraction , all washes were collected and centrifuged so the total number of extracellular bacteria present in the supernatant could also be determined . As shown in Fig . 6B , the number of InlAm-expressing L . monocytogenes present in the mucus layer of female BALB mice increased over time in both the ileum and the colon . At 24 hours post-infection , very few intracellular InlAm L . monocytogenes were detected in any of the gut tissues ( Fig . 6C ) . Two days later , however , CFU counts had increased in each intestinal fraction , with the majority of the bacterial load present in the colon . By 5 days post-infection , the number of intracellular Lm InlAm had decreased in the both the ileal and colonic epithelium , while the bacterial load in the lamina propria was maintained ( Fig . 6C ) . Surprisingly , an equal or greater number of extracellular L . monocytogenes were found in both tissues at all time points tested . Together , these results suggested that the bulk of L . monocytogenes replication occurred in the colon , and that persistence of infection beyond three days was a result of growth or survival in the lamina propria and the mucus layer , but not the epithelium . Having developed a method that facilitated quantification of the entire bacterial load as L . monocytogenes translocated across the gut mucosa , we next asked whether a high affinity interaction between InlA and E-cadherin was needed either for invasion of the epithelium or for persistence in the underlying lamina propria . Female BALB mice were co-infected with a 1∶1 mixture of wild type and InlAm-expressing L . monocytogenes , and the ileum and colon from each mouse was fractionated and plated at 16 , 36 and 60 hpi . At 16 hpi , the total CFU for each strain in mucus and the intracellular CFU in either EC or LP cells was below the limit of detection ( data not shown ) . As expected , very few Listeria were detected in the ileum at either 36 or 60 hpi , and there was no significant difference in the number of Lm EGDe or Lm InlAm recovered ( Fig . 7 ) . In the colon , similar numbers of the two bacterial strains were recovered from the mucus , EC and LP fractions 36 hpi , but the bacterial load remained below the limit of detection in many mice ( Fig . 7B , C , D ) . By 60 hpi , intracellular L . monocytogenes were found in the EC fraction of most mice , however , there was not a substantial difference in the number of wild type or InlAm-expressing bacteria isolated ( Fig . 7A , C ) . Thus , the mouse-adapted InlAm was not essential for invasion of the colonic epithelium , and expression of the high affinity E-cadherin ligand did not enhance the intracellular replication or survival of L . monocytogenes over time in epithelial cells . In contrast , by 60 hpi , InlAm-expressing L . monocytogenes outcompeted wild type Lm EGDe by an average of 39-fold in the colonic lamina propria ( Fig . 7A , D ) . Thus , the mouse- adapted high affinity ligand for E-cadherin promoted either the growth or persistence of L . monocytogenes inside cells of the colonic lamina propria .
Oral transmission of L . monocytogenes is not highly efficient in mice , and this has been attributed largely to a species specificity for the interaction between the bacterial surface protein InlA and E-cadherin expressed on intestinal epithelial cells . In this study , we developed a novel model of food borne listeriosis in mice and showed that expression of an InlA protein that could serve as a high affinity ligand for E-cadherin was not required for colonization of the murine gut . We propose that the species barrier for InlA is not the major factor responsible for inefficient oral transmission of L . monocytogenes in small animal models . Instead , other parameters of the gastric environment are likely to play a much larger role in blocking infection in mice . McConnell et al . showed that the pH of both the stomach and the intestinal tract was lower in mice than in humans [40] , and the increased acidity could result in greater bacterial killing . In fact , in this study , very few ingested L . monocytogenes survived passage through the murine stomach as evidenced by both CFU counts recovered from the intestinal lumen 24 hpi , and the amount of live L . monocytogenes shed in feces 3 hpi . However , prolonged exposure to either the acidic milieu or high osmolarity of the stomach may be essential for L . monocytogenes virulence , since these stresses trigger sigmaB-dependent changes in gene transcription that result in increased invasion of enterocytes and growth in macrophages [41]–[43] . Thus , the small number of L . monocytogenes that survive passage through the murine stomach are likely to be better adapted for intestinal colonization . Nonetheless , invasion of the intestinal epithelium appears to be an infrequent event , even when a high affinity interaction between InlA and E-cadherin is possible . Melton-Witt et al . recently estimated that only 1 in 106 L . monocytogenes invaded intestinal villi following oral inoculation of guinea pigs [44] and we found a similar frequency of cell-associated bacteria in both the ileum and the colon using a food borne infection model in mice . L . monocytogenes is commonly thought of as an organism that infects the small intestine; however , the colon appeared to be the primary site for bacterial replication in mice following ingestion of contaminated food . In many of the previously published reports of oral listeriosis in mice , the large intestine was not examined . However , our results are consistent with a previous study by Disson et al . that showed increased invasion of L . monocytogenes in the colon compared to the small intestine in transgenic mice expressing human E-cadherin [19] . Furthermore , Nikitas et al . recently identified goblet cells as a primary site of intestinal invasion using a ligated jejunal loop model [8] , and goblet cells are both more numerous and larger in size in the colon . In that study , the authors used a microscopic approach to show that L . monocytogenes lacking InlA were unable to mediate rapid invasion ( within 30–45 minutes ) of ligated jejunal loops in transgenic mice expressing both murine and human E-cadherin . Intestinal infection was not assessed at later time points , so the ability of luminal bacteria to translocate across the mucosa using other , possibly slower routes , was not determined . Only one other study by Wollert et al . has quantified the amount of InlAm-expressing L . monocytogenes in the gut at multiple time points following oral transmission , and they also found no difference in the number of wild type or mouse-adapted intracellular bacteria in the small intestine for the first 48 hours after i . g . inoculation [20] . In this study , bacteria completely lacking inlA showed a defect only in the ileum , but colonized the colon as efficiently as InlAm-expressing L . monocytogenes . Although InlA-mediated uptake may be faster , these studies clearly indicate that L . monocytogenes can readily use alternate routes , such as passage through M cells , to translocate across the gut mucosa . Furthermore , our data suggest that the invasion mechanisms used by L . monocytogenes may differ significantly in the small and large intestines . The efficiency of enterocyte invasion is not the only factor that determines the net rate of intestinal colonization . Bacterial growth rates , the ability to avoid immune clearance mechanisms , and the rate of dissemination to other tissues , all influence the number of CFU present in the gut at any given time point during infection . However , the route of intestinal invasion may influence the subsequent localization into intestinal compartments with varying degrees of innate resistance against bacterial growth or survival . For example , M cells overlying Peyer's patches deliver phagocytosed bacteria directly to an underlying lymphoid follicle comprised of B cells , T cells , macrophages and dendritic cells . Although the phagocytes in these follicles could provide a replicative niche for L . monocytogenes , the close proximity to other immune cells that can rapidly produce IFN-gamma and TNF-alpha may quickly lead to activation of the macrophages , so they no longer support intracellular replication of the bacteria [45] , [46] . Resident CD11b ( + ) CD11c ( + ) CX3CR1 ( + ) cells in the subepithelial dome of Peyer's patches were recently shown to express significantly higher levels of lysozyme compared with phagocytes found in intestinal villi [47] . L . monocytogenes are not killed by lysozyme alone [48] , but this observation suggests that there may be subsets of macrophage-like cells in Peyer's patches that have enhanced bactericidal activity and thus , do not support efficient replication of intracellular bacterial pathogens . In contrast , InlA-mediated uptake occurs primarily at villus tips or near goblet cells and promotes rapid transcytosis of L . monocytogenes directly to the underlying lamina propria [7] , [8] . Once in the lamina propria , Listeria can infect macrophages or dendritic cells , or re-infect epithelial cells by binding to E-cadherin expressed on the basolateral surface . In agreement with Nikitas et al . [8] , we found that prolonged infection of intestinal epithelial cells did not occur using the food borne model . This is likely because InlAm-expressing bacteria that entered epithelial cells from the apical surface were quickly transcytosed to the lamina propria , and bacteria that infected from the basolateral side were rapidly shed back into the lumen in extruded enterocytes [44] . Interestingly , expression of the low affinity ligand for murine E-cadherin ( native InlA ) appeared to be deleterious in mice later in the infection , 60 h after ingestion , when wild type Lm EGDe was beginning to be cleared from the colon , but Lm InlAm persisted . Likewise , Wollert et al . began to observe differences in colonization of the small intestine 72 h after i . g . inoculation of either wildtype or InlAm-expressing L . monocytogenes [20] . This was not the result of an intrinsic growth or survival advantage for InlAm-expressing bacteria , because a deletion mutant lacking inlA persisted in the colon equally as well as the mouse-adapted strain . One explanation for these results could be that InlA with a low affinity for E-cadherin may act as a decoy receptor that causes the bacteria to engage non-productively with E-cadherin on the basolateral surface of the epithelium . If the bacteria do not find an intracellular niche in either enterocytes or phagocytes in the lamina propria they would be vulnerable to clearance by innate immune mechanisms . Although InlA was not required for dissemination to the MLN , InlAm-expressing L . monocytogenes had a clear advantage in spread from the gut to the draining lymph nodes . In about 20% of the animals we examined , bacteria that lacked a high affinity ligand for E-cadherin ( wt EGDE or ΔinlA ) did not spread to the MLN . This suggests that InlA helps promote passage through a bottleneck in the gut that leads to systemic spread . We presume that this bottleneck is entry into a migratory phagocyte such as a dendritic cell . Although in vitro studies with bone marrow-derived cells suggest that L . monocytogenes does not replicate efficiently in dendritic cells , the migratory nature of dendritic cells could serve an important function to promote dissemination of intracellular bacteria via the lymphatic system [49] , [50] . In support of this idea , Siddiqui et al . recently identified a minor subset of intestinal dendritic cells that expresses E-cadherin . These monocyte-derived CD103 ( + ) CX3CR1 ( − ) cells accumulated in the intestinal lamina propria during both T cell-mediated colitis and Trichuris muris infection , and then migrated to the mesenteric lymph nodes [51] , [52] . Indeed , in preliminary studies , we have been able to identify a subset of CD11c ( + ) E-cadherin ( + ) cells in the MLN that increased in number during food borne listeriosis ( data not shown ) . InlAm-expressing bacteria would thus have an advantage in dissemination , because they would not be limited solely to uptake by phagocytosis and could use InlA-mediated uptake to gain access to an additional subset of migratory phagocytes . InlAm-expressing bacteria had only a slight advantage in reaching the spleen , and no competitive advantage in reaching the liver . Previous studies using tagged strains of either Yersinia or Listeria in oral inoculation models suggested that there are two possible routes of bacterial dissemination from the gut to the spleen and liver [44] , [53] . In the direct pathway , bacteria travel via the portal vein to the liver , Kupffer cells efficiently remove the majority of the bacterial load , and unfiltered organisms continue through the peripheral blood system to reach the spleen . L . monocytogenes may access intestinal blood vessels by direct invasion of endothelial cells , a process that is independent of InlA in vitro [54] , [55] . A second , indirect pathway occurs when bacteria spread via the lymphatic system , first to the draining lymph nodes , then through efferent lymphatic vessels to the thoracic duct , and then on to peripheral tissues via the bloodstream . Our data differ from the findings of Monk et al . who reported that InlAm promoted spread to both the spleen and the liver [27] . However , that study was performed with mice infected by the i . g . route , so it is possible that physical trauma facilitated direct bloodstream invasion and the large number of bacteria inoculated ( 1000× more than the i . v . LD50 ) resulted in significant seeding of the spleen as well as the liver . Another recently described reservoir of L . monocytogenes replication in mice is the gall bladder [32] . In this study , we confirmed the observations that L . monocytogenes can be recovered from the gall bladders of BALB mice a few days after infection , and that an exponential increase in bacterial load occurred in this tissue . However , very few bacteria reached the gall bladder in B6 mice , and strikingly , there was little increase in the number of L . monocytogenes recovered from B6 gallbladders from 3 to 7 dpi . It is possible that the B6 gall bladder is not a permissive site for bacterial replication , or alternatively , continuous spread from other infected tissues such as the spleen or the liver may be a more important factor in determining the overall bacterial load in the gall bladder . The efficiency of gall bladder colonization may also greatly impact the bacterial load in the gut . Upon ingestion of food , the large number of L . monocytogenes in the BALB gall bladder could be excreted back into the intestines [33] , contributing to the persistent colon infection observed in BALB , but not B6 mice . During the development of the food borne model of listeriosis , three factors were shown to greatly influence susceptibility to infection: gender , time of day , and food restriction . In susceptible BALB mice , females had the highest bacterial burdens , but the innate resistance of B6 mice did not appear to be gender-dependent . Pasche et al . previously reported increased lethality in female mice infected intravenously with L . monocytogenes; however , in that study both BALB and B6 females were more susceptible than males [34] . In another study , B6×C3H F1 mice pre-treated with estrogenic compounds were more susceptible to L . monocytogenes [56] . This suggests that estrogen levels in female mice may significantly alter innate resistance to infection . Time of day-dependent changes in immune cell number or function have been reported previously [57]–[59] , so it is possible that a circadian rhythm triggered by exposure to light controls the expression of genes needed to rapidly clear L . monocytogenes . However , peripheral oscillators that respond to other cues , such as feeding activity , can also establish independent rhythms of gene expression in specific tissues . In that regard , it is notable that a period of food restriction enhanced susceptibility to food borne listeriosis in both B6 ( Fig . S1 ) and BALB ( not shown ) mice . McConnell et al . showed that the intestines of fasted mice had a higher pH than mice given free access to food [40] . Since decreased acidity would also promote L . monocytogenes survival , it is not yet clear exactly how food restriction impacts innate resistance to infection . However , the preliminary data presented here indicate that food borne listeriosis in B6 mice will be a useful model to better understand how circadian rhythms and diurnal variations affect the innate immune system . The food borne model of L . monocytogenes infection has several advantages over the conventional i . g . inoculation model . Transmission of the bacteria occurs by natural feeding , and thus , is not invasive and does not cause unintended minor trauma in the esophagus or stomach . No specialized skills are required for infection , so this method can be widely used by many different laboratories , and may not result in as much lab-to-lab variation as was observed with i . g . inoculation . The model is ideal for studying host response to infection since any mouse strain can be used , including the multitude of knockout and transgenic animals that currently exist , offering an important advantage over use of the recently described guinea pig model [44] . As reported here , susceptible BALB and resistant B6 mice represent the two ends of the spectrum of human disease ranging from mild , self-limiting gastroenteritis to potentially lethal systemic and brain infection . Importantly , it is the first small animal model that can be readily adapted to study the role of particular types of food in transmission of listeriosis . There is a large body of data in the literature examining the growth and survival rates of Listeria found in various types of foods , but very little information regarding the infectivity of Listeria isolates propagated in different food types or stored at different temperatures [60]–[63] . A large percentage of human listeriosis outbreaks have been associated with foods that are high in fat composition , including one linked to contaminated butter [64] . Ingestion of fatty foods is likely to induce a different profile of gastric secretions that could influence both bacterial survival and the ability to colonize the intestines . Finally , the approach used here should be widely applicable to many other orally transmitted bacterial pathogens such as Salmonella spp , Yersinia enterocolitica and Escherichia coli .
This work was performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health . All procedures were approved by the Institutional Animal Care and Use Committee ( IACUC ) at the University of Kentucky . Wildtype L . monocytogenes EGDe and the modified internalin A derivative ( Lm InlAm ) [20] were provided by Wolf Dieter Schubert ( Braunschweig , Germany ) . An inlA deletion mutant on the EGDe background ( Lm ΔinlA ) was the gift of Cormac Gahan ( University College Cork , Ireland ) . Antibiotic resistant versions of each L . monocytogenes were generated for co-infection studies using site-specific integrative plasmids pAD1-cGFP and pAD1-cYFP [65] ( for chloramphenicol resistance; kindly provided by Pascale Cossart , Pasteur , France ) or pIMC3 plasmids [66] ( for erythromycin , kanamycin and tetracycline resistance; provided by Cormac Gahan ) . Each strain was intestinally passaged by oral infection of a BALB mouse . Bacteria recovered from the small intestine were grown to early stationary phase in Brain Heart Infusion ( BHI ) broth shaking at 37°C , and then aliquots were prepared and stored at −80°C . To infect mice , an aliquot was thawed on ice , cultured standing in BHI broth for 1 . 5 h at 30°C , washed once in PBS , and then suspended in the inoculation solution . Tissue samples were plated on BHI agar ( Difco ) supplemented with 15 g/L LiCl and 10 g/L glycine ( BHI/L+G ) , a selective medium that inhibited the growth of most intestinal microbiota . Colony growth was monitored after 48 h incubation at 37°C; suspect colonies were confirmed to be L . monocytogenes by plating on CHROMagar Listeria plates . Male and female C57BL/6/J ( B6 ) and BALB/c/By/J ( BALB ) mice were purchased from The Jackson Laboratory ( Bar Harbor , ME ) at 5 weeks of age and used in experiments when they were 6–9 weeks old . All mice were maintained in a specific-pathogen free facility at the University of Kentucky with a 14 h light cycle ( 7 AM–9 PM ) and a 10 h dark cycle ( 9 PM–7 AM ) . Mice were denied food , but given unrestricted access to water 16 h prior to infection . The Lm InlAm inoculum was suspended in PBS without bicarbonate and 200 µl was placed directly into the stomachs of non-anesthetized mice [67] using a 20 g straight feeding needle . Food was returned 1 h post-infection . Mice were placed in cages with raised ( 1 inch ) wire flooring ( #3 mesh ) to prevent coprophagy and food was removed 16–24 h prior to infection unless otherwise indicated . The L . monocytogenes inoculum was suspended in 5 µl of either PBS , 2% glucose in PBS , or melted salted butter ( Kroger ) and used to saturate a 2–3 mm piece of white bread ( Kroger ) in a microcentrifuge tube . In pilot experiments , blue food coloring was also added to the inoculum to facilitate visual monitoring of the food particle , however , this step was later determined to be unnecessary . At the time of infection , each mouse was placed in an empty cage ( no bedding ) and the contaminated bread piece was placed on the bottom of the cage . Typically , mice picked up the bread and ate all of it within 5–10 minutes . After eating the bread , mice were returned to their original cages and normal mouse chow was replenished within 30–45 min . Small intestines were either processed whole ( Figs . 1–3 ) or detached and cut into thirds ( Figs . 4 , 6 , 7 ) approximating the duodenum ( proximal ) , jejunum ( middle ) , and ileum ( distal ) . Colon and cecum sections of the large intestine were processed separately . Intestinal contents were removed by squeezing with sterile forceps , and then each section was flushed with a total of 8–10 ml of PBS through a 25 g needle . To quantify the number of bacteria in the lumen , the pooled contents and flushes were centrifuged for 20 min . at 12 , 000 x g . The bacterial pellet was suspended in 0 . 5–1 . 0 ml sterile water and serial dilutions were plated on BHI/L+G agar . Washed intestinal tissues were cut longitudinally with a sterile scalpel blade , placed in 2 ml of sterile water , and then homogenized for 1 minute using a PowerGen 1000 homogenizer ( Fisher ) at 80% power . The total number of cell-associated ( adherent extracellular plus intracellular ) bacteria was determined by plating serial dilutions on BHI/L+G agar . Spleen , liver and brain were harvested aseptically and homogenized in sterile water for 30 seconds . Gall bladders were collected into microcentrifuge tubes containing 1 ml sterile water , ruptured with sterile scissors , and vortexed for 30 seconds . Mesenteric lymph nodes were mashed through a sterile mesh screen into 0 . 5 ml of sterile water and each screen was rinsed with an additional 1 ml of water . Dilutions of each tissue sample were prepared in sterile water and plated on BHI/L+G agar . For tissues harvested 16 hours post-infection ( hpi ) , the homogenates were centrifuged for 20 min . at 12 , 000 x g and suspended in 0 . 1–0 . 25 ml PBS to lower the limit of detection . Fecal pellets were collected at the indicated time points , weighed , and then suspended in sterile water ( 150 mg/ml ) . Typically , 2–3 pellets with an average total weight of 40 mg were collected from each animal . The pellets were mashed with a sterile toothpick , and then vortexed for 30 seconds before diluting and plating on BHI/L+G agar . Colonies were counted after 24 h growth at 37°C . The limit of detection for L . monocytogenes in these samples was 0 . 13 CFU per mg of feces . For co-infections , bacterial suspensions were mixed prior to saturation of a single bread piece . In early experiments , one strain was tagged with chloramphenicol resistance and tissue homogenates ( Fig . 4 , 5 ) or single cell suspensions of intestinal fractions ( Fig . 7 ) were plated on BHI/L+G with or without the presence of chloramphenicol ( 7 µg/ml ) . The number of chloramphenicol sensitive ( CmS ) CFU was determined by subtracting the number of ( CmR ) colonies from the total CFU found on plates without antibiotic . In later experiments , both of the strains used for co-infection were marked with antibiotic resistance genes and homogenates were differentially plated on BHI with 1 mM IPTG plus 50 µg/ml kanamycin ( to detect wildtype Lm EGDe ) , 5 µg/ml erythromycin ( to detect Lm InlAm ) , or 10 µg/ml tetracycline ( to detect Lm ΔinlA ) . Competitive index ( CI ) ratios were determined by dividing the number of either wild type Lm EGDe or ΔinlA Lm CFU by the number of Lm InlAm CFU recovered from each tissue . If only one strain was recovered , a value equal to the limit of detection was used for the other strain; if no CFU were recovered , then a CI value was not calculated . Flushed ileum and colon sections were cut longitudinally and treated with N-acetylcysteine ( NAC; Sigma ) using a variation of a previously described protocol [38] to remove the mucus layer without damaging the underlying epithelium . Each tissue was washed three times by incubating for 2 min . in a tube containing 3 ml of 6 mM NAC , then shaken vigorously before transferring to a fresh tube . The pooled washes were centrifuged for 20 min . at 12 , 000 x g , suspended in sterile water , and vortexed for 30 sec . prior to dilution and plating . No eukaryotic cells ( viable or dead ) were found in the mucus fractions , as determined by trypan blue staining . The epithelium was removed using standard protocols [39] modified as follows to give higher cell yield and increased cell viability . After mucus extraction , each tissue was cut into small pieces and incubated for 10–20 min . shaking at 37°C in three successive washes ( 5 ml ) of RPMI ( Invitrogen # 21870 ) containing 5% FBS ( RP-5 ) , 5 mM EDTA , and 1 mM DTT . The pooled washes ( referred to as the EC fraction ) were centrifuged at 1 , 200 x g and then the cells and the supernatant were processed separately as described below . The remaining intestinal pieces were rinsed with PBS to remove excess DTT/EDTA , and then digested with collagenase IV ( 1 mg/ml; Worthington ) and DNAse I ( 40 µg/ml; Worthington ) to release the lamina propria ( LP ) cells . The tissue pieces were incubated for 40 min . shaking at 37°C in 2–3 successive changes of digestion solution ( 5 ml ) until visible pieces of tissue disappeared . The pooled LP fractions were centrifuged at 1 , 200 x g and the cells and supernatant were processed separately . Supernatants from the EC and LP fractions were centrifuged for 20 min . at 12 , 000 x g , suspended in sterile water , and diluted and plated to determine the total number of extracellular L . monocytogenes . The cellularity of single cell suspensions of EC and LP was confirmed by Diff-Quik staining . The cells were incubated for 30 min . at 37°C with 7% CO2 in RP-5 containing 25 µg/ml gentamicin to kill any adherent or remaining extracellular L . monocytogenes . After gentamicin treatment , the cells were washed twice in PBS , lysed in sterile water , then diluted and plated to determine the number of intracellular L . monocytogenes . A comparison of total intracellular plus extracellular CFU recovered from intestinal sections before and after collagenase treatment indicated that the bacteria did not replicate significantly during in vitro processing ( not shown ) . All statistical analysis was performing using Prism5 for Macintosh ( Graph Pad ) . P values less than 0 . 05 were considered significant and are indicated as follows: * , P<0 . 05; ** , P<0 . 01; *** , P<0 . 001; **** , P<0 . 0001 .
|
Ingestion of Listeria monocytogenes-contaminated food can be life-threatening for immune compromised individuals , can cause severe gastroenteritis in otherwise healthy people , and is also thought to occur frequently with little consequence . The factors that determine susceptibility to this infection are unknown , due to the lack of an appropriate animal model that closely mimics this wide range of human disease . Mice are highly resistant to oral L . monocytogenes infection , and the prevailing view has been that a low affinity between the bacterial surface protein InlA and E-cadherin expressed on the gut mucosa was largely responsible for limited invasion of the murine intestines . We used a novel food borne model of listeriosis to show that a mouse-adapted InlA offers little advantage over wild type InlA for initial colonization of the gut , and indeed , even bacteria lacking InlA can establish intestinal infection in mice . Thus , other aspects of the murine gastrointestinal environment appear to be the key to innate resistance against oral transmission . Surprisingly , our study uncovered a novel function for InlA later in the infection , when the bacteria begin to spread systemically . The natural feeding model presented here using susceptible and resistant strains of mice should be very useful for future studies investigating both mechanisms of microbial pathogenesis and host responses to oral infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"models",
"model",
"organisms",
"biology",
"microbiology",
"bacterial",
"pathogens",
"pathogenesis"
] |
2012
|
InlA Promotes Dissemination of Listeria monocytogenes to the Mesenteric Lymph Nodes during Food Borne Infection of Mice
|
From Darwin's study of the Galapagos and Wallace's study of Indonesia , islands have played an important role in evolutionary investigations , and radiations within archipelagos are readily interpreted as supporting the conventional view of allopatric speciation . Even during the ongoing paradigm shift towards other modes of speciation , island radiations , such as the Lesser Antillean anoles , are thought to exemplify this process . Geological and molecular phylogenetic evidence show that , in this archipelago , Martinique anoles provide several examples of secondary contact of island species . Four precursor island species , with up to 8 mybp divergence , met when their islands coalesced to form the current island of Martinique . Moreover , adjacent anole populations also show marked adaptation to distinct habitat zonation , allowing both allopatric and ecological speciation to be tested in this system . We take advantage of this opportunity of replicated island coalescence and independent ecological adaptation to carry out an extensive population genetic study of hypervariable neutral nuclear markers to show that even after these very substantial periods of spatial isolation these putative allospecies show less reproductive isolation than conspecific populations in adjacent habitats in all three cases of subsequent island coalescence . The degree of genetic interchange shows that while there is always a significant genetic signature of past allopatry , and this may be quite strong if the selection regime allows , there is no case of complete allopatric speciation , in spite of the strong primae facie case for it . Importantly there is greater genetic isolation across the xeric/rainforest ecotone than is associated with any secondary contact . This rejects the development of reproductive isolation in allopatric divergence , but supports the potential for ecological speciation , even though full speciation has not been achieved in this case . It also explains the paucity of anole species in the Lesser Antilles compared to the Greater Antilles .
Speciation generates biodiversity and is therefore a key process in evolution and ecology , and the relative importance of factors contributing to speciation in sexually reproducing animals , such as genetic drift in spatial isolation , natural selection , sexual selection and mutation-order , remains an active area of research [1]–[8] . Since neo-Darwinism [9] the most conventional view of speciation in sexually reproducing animals has been by the accumulation of differences by genetic drift and selection in allopatry . While there has been growing paradigm shift towards models [10] and processes such as ecological speciation [1] , [5] , [7] , [8] that are not dependent on allopatry , there have been few critical tests of allopatric speciation in systems which are regarded as exemplifying the process , such as island archipelagos [9] , [11]–[17] . This is primarily because , from a contemporary perspective , genetic isolation cannot be assessed in spatially isolated populations . However , a historical perspective allows us to test the genetic isolation of anole species isolated for a very substantial time before their islands coalesced . Anolis ( small insectivorous lizards ) is the most speciose amniote genus ( circa 400 species ) [18] and show little inter-specific hybridization [19] . Just two colonizations of the Caribbean islands have resulted in 150 species , so they may be thought of as exemplifying allopatric speciation in island archipelagos [11]–[14] , [18] , [20] . These anole radiations appear to have inhabited the Lesser Antilles since the origin of the younger island arc , or just before ( i . e circa 8–9 mybp ) with a southern and a northern series [21] . On what is currently recognized as the island of Martinique ( southern series ) , the paraphyletic anole Anolis roquet has deep phylogeographic divisions , with Anolis extremus from Barbados nested within it [21] . Geology [22]–[23] , molecular phylogeography and molecular clock analysis [21] reveals that four precursor islands of Martinique ( Figure 1 ) are associated with four mtDNA lineages of ‘A . roquet’ . The island ages , molecular clock and geographic distribution of the lineages link closely to suggest that the precursor islands of Martinique ( together with Barbados ) had separate anole allospecies for up to about 8mybp , before central uplifting joined the Martinique precursors to form a single island ( with Barbados remaining independent ) [21] . This gave three secondary contact zones in Martinique ( Figure 2 ) between previously allopatric forms ( south-central , SW-central , NW-central ) that:- 1 ) are phylogenetically deeper than the species-level split between A . extremus ( consistently regarded as a valid species [18] ) and its sister clade within A . roquet [21]; 2 ) have diverged a substantial time ago ( 6–8 mybp ) and have a level/time of phylogenetic divergence that is comparable to other Lesser Antillean anole species [11]–[12]; 3 ) may show distinct mtDNA lineages with almost no haplotype inter-digitation [21]; and 4 ) may show a prima facie case for parapatric bimodality in multivariate quantitative traits at some points of contact [24] ( Figure S1 ) . Hence , this is an appropriate test of allopatric speciation . Martinique anoles also provide a test for ecological speciation , or isolation by adaptation [7] . The quantitative traits of Lesser Antillean anoles adapt by natural selection to environmental zonation , as shown by common garden and natural selection experiments [25]–[26] , parallels among island species , and correlation studies that take phylogenetic history into account [27] . In Martinique , the montane rainforest and coastal xeric woodland are distinctly different habitats with pronounced differences in the environmental conditions across the ecotone between them . As with other Lesser Antillean anoles , the Martinique anole adapts to these conditions and their populations show marked habitat-related differences in quantitative traits such as morphology ( shape , color , pattern and scalation ) and dewlap hue [21] , [24] , resulting in distinct ecotypes . The ecotone between these coastal xeric and montane rainforest habitats provides a test for ecological speciation for comparison with secondary contact zones . Hence , with the Martinique anole there is the potential for speciation to occur in accordance with both an allopatric model ( where the different lineages on precursor islands speciate ) , and an ecological model ( where the different ecotypes speciate ) . Preliminary analysis of a single transect suggested that under specific circumstances there may be greater restriction of genetic exchange between habitat types than previously allopatric forms in secondary contact [21] . However , this analysis examined just one of the three pairs of coalescing islands ( northwest vs . central ) , under only one set of selection regimes ( strong convergent selection for montane rainforest on both lineages where they met along that transect ) . Furthermore , this study was not replicated and no control was used , limiting the capacity to generalize , and raising several questions . Specifically , do the other pairs of coalescing precursor islands populations ( southwest-central and south-central ) show evidence of genetic isolation or not; is the pattern of inter-digitation of the mtDNA lineages and introgression of the neutral nDNA consistent along the length of each of the three secondary contact zones , or does it vary dependant on other factors; how do the selection regimes along the transect influence the extent of genetic isolation among previously allopatric forms; under what ecological conditions ( extent and abruptness of habitat change ) is there restricted genetic exchange among habitat types and how long does it take to develop ? To answer these questions we investigated the xeric/rainforest ecotone and all three cases of island coalescence , each with two to four replicate transects , together with a control transect ( Figure 2 , Table 1 , Table 2 ) . By measuring nuclear genetic structure , mtDNA lineage , quantitative traits and climate variation along these replicated transects , across both geological and habitat contact zones , we are able to critically test the role of these two factors , and their interaction , in the differentiation of island anoles . We show that , although there is always a signature of past allopatry in the nuclear genetic structure , and this can be quite strong dependent on the comparative selection regimes across the secondary contact zone , there is no complete allopatric speciation for any of the three allopatric pairs . Instead , if there is sufficient magnitude and abruptness of habitat change , then there is even greater differentiation across the ecotone , and this can develop over a brief period of time . Although the ecological speciation is not complete , it has reached what has been characterized as a “later stage” in the speciation continuum [7] . This supports a relatively important role for ecological speciation under the appropriate circumstances [1] , [5] , [7] , [8] .
Recent island-wide phylogenetic studies identified four main mtDNA lineages within A . roquet whose geographical limits correspond very closely to the geological junctions between precursor islands [21] , [28] , with the timing of divergence between these lineages compatible with the age of the different precursor islands . This supports the scenario illustrated in Figure 1B , which suggests that the individual lineages evolved in allopatry for about 6mybp ( central-south ) to 8 mybp ( central southwest and central northwest ) until the precursor islands merged to form present day Martinique . Here , we use a large sample per site , with sites along transects focussed on the contact zones . Estimating the frequency of mtDNA lineages at localities along these transects enables us to test for any inter-digitation of the lineages and the fit between the distribution of the lineages and the precursor islands at this fine scale . With the exception of transect VIII ( φ = 0 . 61 ) , we observed a very close association between the precursor islands and the mtDNA lineages ( 0 . 71<φ<0 . 95 ) and little , or almost no ( transects I , IV ) , inter-digitation , even at this fine spatial scale ( Table 1 , Figure 3 , Figure 4 , Figure 5 ) . This absence of substantial inter-digitation , despite a relatively long period of contact ( the precursor islands merged about 1 Mya [21] ) , implies the absence of extensive female-driven gene flow [29] between these previously allopatric lineages . Climate is a strong determinant of habitat and can be objectively measured and quantified . The results ( Table 2 ) show strong climatic variation along the transects ( III , IV ) that run from the xeric coast to the montane rainforest with a sharp transition ( ecotone ) between these habitats ( Figure 3 ) . Other transects generally run within habitat types and show more subtle climatic variation , i . e . , are without abrupt changes in habitat of a high magnitude . A wide-ranging multivariate profile of the quantitative traits ( QTs ) of individuals was taken ( these include both spectrometric dewlap hues [21] and morphological traits such as colour pattern , body dimensions and scalation ) to estimate the change in QTs along a transect in relation to habitat type , ecotone and lineage . Lesser Antillean anole quantitative traits ( QTs ) are generally tightly linked to the habitat and have generally been shown to adapt to environmental conditions and reflect selection regime rather than phylogeographic lineage [24]–[28] . Hence , as predicted , the large magnitude of habitat variation in transects III and IV is matched by a high magnitude of QT variation ( 16–17 within group standard deviations , Table 2 ) with highly divergent rainforest and xeric ecotypes ( Figure 2 and Figure 3 ) , and a close correlation between QTs and climate variation along the transect ( r = 0 . 96 to 0 . 97 , Table 2 ) . The large magnitude and close association of the climate and QT variation indicates the potential importance of the ecotone in determining population structure . Elsewhere , where the magnitude of climatic variation along a transect is less ( because they largely run within habitat types ) , such a very high correlation between climate and QTs is not predicted or observed . Even so , the correlation is only insignificant in one non-control transect ( Table 2 ) , once again suggesting the general importance of habitat type in determining quantitative traits . The correlation between quantitative traits and lineage frequency is significant along four transects ( Table 1 ) , and is particularly high in transect I , where , on this spatial scale , there is no overlap between multivariate morphology of morphs either side on the lineage contact zone ( Figure S1 ) . The allopatric model of speciation predicts that the four lineages that spent a substantial time in isolation on separate islands ( divergence at circa 6–8my ) should all be reproductively isolated entities . That is , there should be four species with very little ( if any ) gene exchange among them where they meet along all three contact zones ( northwestern/central , southwestern/central and southern/central ) on what is currently Martinique . The very close association ( with only one exception ) between the precursor islands and the mtDNA lineages along the transects does not contradict this ( Table 1 ) . This prediction was tested by estimating the population structure along transects I–VIII using neutral , hypervariable , nuclear microsatellite markers , primarily analysed by Bayesian assignment , with support from AMOVA and standardised FST′ values . Principal component analysis ( PCA ) provides an independent perspective on the population affinities . The analysis of these markers along the replicated transects across these three zones clearly rejects the presence of reproductively isolated ( or even partially isolated ) species . This is the case for all three precursor island contacts: the central/northwest contact ( Figure 3 ) , the central/southwest contact ( Figure 4 ) , and the central/south contact ( Figure 5 ) along the length of the contact for all replicates and various types of selection regimes ( but see transect I below ) . The Bayesian assignment method detects two clusters in most transects ( Table 3 , Table S1 ) , but the transition between the two clusters generally is not closely associated with the lineages and/or forms a smooth cline ( Figure 3 , Figure 4 , Figure 5 ) . The PCA ( Figure S2 ) supports the Bayesian clusters in transects I–VIII as the pattern of relative frequency of the Bayesian clusters ( where K = 2 ) is almost identical to the pattern of PC1 scores for each transect ( r = 1 . 0 with one exception ) . The association between nuclear genetic clusters and allopatric speciation model ( lineage categories ) is , with one exception , modest ( 0 . 32<φ<0 . 51 ) even if significant ( Table 3 ) and φ does not approach unity ( complete isolation ) . This pattern is supported by the AMOVA and standardized FST′ values . The AMOVA show sporadic significant structure associated with lineages ( transects I , II , V , VII ) , but all the ΦCT values are substantially less than unity and too low for reproductive isolation ( Table 4 ) . Similarly , mean standardized genetic differentiation between pairs of populations on each side of lineage boundaries is low to moderate ( 0 . 072<FST′<0 . 166 , Table 4 ) . This suggests high levels of nuclear gene exchange between lineages ( the equivalent unstandardized FST values are 0 . 014<Fst<0 . 043 ) . Of particular interest are transects III and IV where the nuclear genetic structure associated with the northwest and central lineages can be compared directly with that associated with habitats ( Table 3 and Table 4 ) . Here the Bayesian clusters show substantially poorer fit to the allopatric speciation model ( 0 . 32<φ<0 . 37 ) than the habitat categories ( 0 . 62<φ<0 . 71 ) . The AMOVA shows low ( −0 . 00064<ΦCT<0 . 00158 ) and insignificant ΦCT for the lineage categories , but higher ( 0 . 03481<ΦCT<0 . 01665 ) and significant ΦCT for the habitat categories . The mean standardized genetic differentiation is also much lower between lineage categories ( 0 . 072<FST′<0 . 075 ) than habitat categories ( 0 . 137<FST′<0 . 213 ) . In general , while there may be a nuclear genetic signature of past allopatry for all four mtDNA lineages associated with precursor islands , there is no allopatric speciation . The partial exception to this general trend is transect I , where the central and northwest lineages meet on the northeast coast . Here , there is almost no inter-digitation of mtDNA lineage markers ( φ = 0 . 95 ) , and a sharp stepped cline in quantitative traits at the junction of the precursor islands ( Figure 3B ) . There is some genetic isolation between the lineages as shown by the Bayesian assignment ( φ = 0 . 68 , Figure 3C ) , although neither AMOVA nor standardized FST′ values are exceptionally high ( Table 4 ) . If these lineages were equally isolated along their entire secondary contact zone there might have been a rather weak case for partial allopatric speciation and recognition of their status as separate species . However , they are not . Even along the adjacent transect ( II ) in the transitional forest , which is only 5km inland , the lineages show little genetic isolation ( no Bayesian clusters , Table 3 , Table S1 , Figure 3C ) and do not have distinct quantitative traits ( Figure 3B ) . Further along this secondary contact zone in the montane rainforest ( transects III , IV ) the quantitative traits are identical either side of the secondary contact zone with little nuclear genetic isolation estimated from Bayesian assignment , AMOVA or standardized FST′ . Direct experimental measures of selection in Lesser Antillean [25] , [26] , and other [18] , [30] anoles , as well as other studies of adaptation [21] , [24] , [27] , have shown strong selection intensity on anole quantitative traits , and the pattern of climate variation and QT variation along transects differs among transects I–IV . Hence , although the populations from transects I–IV may broadly share the same history ( particularly adjacent transects I and II ) , they differ in the pattern and intensity of selection along the transect . The similarity of the environment either side of this secondary contact zone in the rainforest ( transects III , IV , Figure 3B ) , and the remarkably parallel appearance of these northwestern and central lineages forms in the rainforest [28] ( Figure 2 , image 3 ) suggests strong convergent selection working on these populations . Along the coastal transect ( I ) there may be no such strong convergent selection , and indeed the environmental variables show a smooth cline along the transect so there may be some divergent selection . This suggests that the persistence of a strong genetic signal of past allopatry may be contingent on the pattern of selection regimes . In conclusion , even though there has been a substantial period of allopatric divergence between northwest/central ( 8 mybp ) , southwest/central ( 8 mybp ) and south/central ( 6 mybp ) lineages , and only restricted inter-digitation of the mtDNA , there is no evidence of complete allopatric speciation even though there may be a significant signal of past allopatry . This is consistent across all three pairs of putative allospecies and between the replicates along the length of all three contact zones , irrespective of the pattern of selection regimes . Nevertheless , if the pattern of selection allows , a stronger signal of past allopatry may be retained . Overall , the results are compatible with divergence in allopatry followed by substantial introgression on secondary contact due to a lack of reproductive isolation . The distinctly different habitats of the xeric coast and the montane rainforest , associated with strongly divergent quantitative traits , provide an opportunity to test for ecological speciation along transects III and IV ( Table 2 ) . Bayesian assignment indicates that both transects have restricted genetic exchanges across the xeric-montane ecotone , although this is stronger in transect IV ( φ = 0 . 71 ) than III ( φ = 0 . 62 ) . The populations of Anolis in the area of transect III ( Figure 2 ) were most likely severely impacted by the 1902 pyroclastic surge that destroyed St Pierre [31] . Although the reinstatement of the reduced gene exchange associated with the ecotone may have been facilitated by ecotypes colonizing the vacant area from adjacent populations of the same altitude , anoles can readily colonize adjacent areas of different attitudes . Consequently , perhaps to some extent , the signal of restricted genetic exchange may have to have developed in circa 100 years , which is likely to be much shorter than elsewhere along this ecotone , and may be too short even for ecological differentiation in these terrestrial amniotes [32] . The results of the AMOVA also support a reduction of gene exchange between habitats for the two transects . This test shows a significant structure when the sites are grouped according to their habitat , but not when they are grouped according to their lineage ( Table 4 , Table S1 ) . Similarly , the mean standardized genetic divergence ( FST′ ) is much higher between habitats than lineages ( see above ) . Even if this is not full reproductive isolation , the restriction of gene exchange between the habitats is very substantial , and along transect IV it is greater than any in this study . Moreover , ( with the above caveat regarding altitudinal restrictions on re-colonization ) it may be capable of developing rapidly as transect III shows greater isolation than associated with allopatric divergence with the AMOVA and ( with one exception ) the goodness of fit ( φ ) statistics . Nosil et al [7] recognise several stages in the continuum of ecological speciation: 1 ) population differentiation , 2 ) ecotype formation , 3 ) speciation and 4 ) post-speciation divergence . They suggest that increased genotypic clustering ( as evidenced here ) indicates a later stage of the speciation process , and the degree of genetic isolation here is as great , or greater , than that associated with their [7] example of the most reproductively isolated Pundamilia cichlid pairs . Moreover , the adjacent , and environmentally very comparable , island of Dominica also has distinct anole ecotypes . A study of microsatellite variation among anole populations on Dominica did not indicate genetic clustering of the ecotypes [27] , so Martinique anoles appear to be at a later stage than the stage 2 of the Dominican ecotypes . Hence , although it is clear that this there is no full ecological speciation here , it appears that the Martinique anoles are between the ecotype ( 2 ) and speciation ( 3 ) stages in the ecological speciation continuum . It may be that the situation is in equilibrium , or is a stage in a progression towards greater isolation . Moreover , even if progression to greater isolation was possible , it could be prevented by persistent volcanic disturbance of the ecotone and/or its spatial discontinuity . Both natural and sexual selection may play a role in this ecological pattern of gene exchange as predation pressure for crypsis [33] may interact with the need for conspecific communication . Substantial work on Lesser Antillean anole ecotypes , including natural selection experiments , indicates that a wide range of character systems , rather than just single characters , adapt these ecotypes to the specific biotope [24]–[27] . Hence , natural selection will be impacting many independent traits [7] . Moreover , sensory drive may be important [34] as these habitats have different light conditions which may impact on visual conspecific communication via secondary sexual traits , including dewlap hue . If assortative mating occurs , where a female preferentially chooses a male with the appropriate pattern and hue for that habitat , then this could result in reduced gene exchange among populations in different habitats . This replicated population genetic study robustly and consistently suggests that , across a range of opportunities and conditions , there is pronounced introgression after allopatry and that even a very substantial amount of time in spatial isolation does not , on its own , necessarily allow for the development of reproductive isolation and speciation . This is all the more notable as fertile , natural inter-specific hybrids are extremely rare in this large , well-studied , genus [18] , [19] , and this is a radiation that is generally regarded as exemplifying allopatric speciation [11]–[14] , [18] , [20] . Even though the habitat forms are partially , rather than completely , reproductively isolated , they can show greater isolation than the putative allospecies , and it may be that this can develop rapidly . In addition , the extent of the genetic signature of past allopatry may be dependent on the pattern of selection regimes across the secondary contact . These observations have implications for animal speciation in general and speciation in anoles in particular . While one could choose to emphasize the lack of complete ecological speciation in this case , we believe these observations reveal the potential importance of ecological divergence as a contributory factor in speciation , including in situations where ecological divergence initiates speciation , but does not complete it [7] , and where allopatry is important , but adaptation to environmental differences are also required , as recently suggested for speciation in birds [35] . Consequently , a role for ecology in speciation , including ecological speciation , or isolation by adaptation [1] , [5] , [7]–[8] , [32] , [36]–[38] , may be of widespread relevance , and non-allopatric models [10] should not be excluded from consideration . These implications are particularly relevant to the most speciose amniote genus , Anolis , including the large Greater Antillean communities , where sympatric and parapatric speciation have been regarded as not being an important phenomena in anole evolutionary diversification [14] , [18] , [20] . Finally , it contributes to an explanation of why there are so few species of Anolis in the Lesser Antilles compared to the Greater Antilles [14] . At the stage of the allopatric model where species number on an island is increased by colonization from other islands [14] , the colonizers interbreed with the species already on the island , because no reproductive isolation has developed while they are in allopatry . The genetic signal of this interbreeding is then lost because the number of overseas colonizers per unit time will be vanishingly small compare to the turnover in the large endemic population .
Replicate transects were taken across each precursor island junction ( Figure 2 ) ; northwest lineage to central lineage transects I , II , III and IV , southwest lineage to central lineage transects V and VI , south to central lineage transects VII and VIII , with a control transect ( IX ) within the central lineage . The number of sites per transect was 8 , 8 , 7 , 7 , 9 , 9 , 8 , 7 and 5 respectively for transects I to IX . At each site 48 naturally autotomized tail-tip biopsies were sampled for molecular analysis , while quantitative traits and dewlap hue were recorded from ten adult males . Where transects crossed the same lineages and were in broadly comparable habitats ( eg , III+IV , V+VI , VII+VIII ) samples were collected , and data was recorded and analysed in these transect pairs . The lineages were first investigated using complete cytochrome b sequence from the mtDNA . PCR-RFLP analyses were then designed to efficiently assign numerous individuals to a specific lineage ( northwest , southwest , south or central ) . The cyt b fragment used in the phylogeographic analysis was digested after amplification using the restriction enzyme SspI ( New England Biolabs ) for 3 hours at 37° . The digested products were run on a 2% agarose gel containing ethidium bromide . This enzyme distinguishes between the central lineage ( uncut by this enzyme ) , the southern lineage ( cut at position 598 ) and the clade comprising the southwestern and the northwestern lineages ( cut at position 166 ) . To further distinguish between southwest and northwest lineages , we digested the same fragment using the restriction enzyme DraI ( New England Biolabs ) that cuts the PCR products from the northwest lineage at position 227 , while those from SW lineage were uncut by this enzyme . The habitat type at each site was estimated from a multivariate climatic profile using nineteen climatic variables from Worldclim ( http://www . worldclim . org/ ) . These variables were annual mean temperature , mean diurnal range , isothermality , temperature seasonality , maximum temperature warmest month , minimum temperature coldest month , temperature annual range , mean temperature wettest quarter , mean temperature driest quarter , mean temperature warmest quarter , mean temperature coldest quarter , annual precipitation , precipitation wettest month , precipitation driest month , precipitation seasonality , precipitation wettest quarter , precipitation driest quarter , precipitation warmest quarter , and precipitation coldest quarter . Logarithm ( natural ) transformed data was subjected to principal component analysis . The component defining the climatic trend along the transect was plotted and the magnitude of climatic change in this trend can be taken as the range between maximum and minimum component scores . If there was an ecotone the cut-point between habitat types was defined as the midpoint between these maximum/minimum component scores . A multivariate suite of 21 morphological characters ( colour pattern , trunk hue , scalation , body dimensions ) were recorded [21] , [39] . The hue of the anterior and posterior dewlap was recorded using reflectance spectrometry [21] and the spectrum of each was divided into 6 independent hues following a multiple-group eigenvector procedure [21] , [40] . The morphological and spectrometric characters were then subjected to canonical analysis with the CVs scaled so that the pooled within-group standard deviation was unity . Heteroscedasticity was a problem with transect III so , as an alternative , a principal component analysis was also run on normalized site means for this transect . The samples were genotyped at nine nuclear microsatellite loci ( AAE-P2F9 , ABO-P4A9 , AEX-P1H11 , ALU-MS06 , ARO-035 , ARO-062 , ARO-065 , ARO-120 , ARO-HJ2 ) [41]–[43] in a single multiplex using a Qiagen Multiplex PCR kit with the annealing temperature at 55° . PCR products were then analysed on an ABI 3130xl genetic analyser and the genotypes scored using Genemapper v4 . 0 ( Applied Biosystems ) . Hardy-Weinberg equilibrium and linkage disequilibrium were tested for using Genepop v3 . 4 [44] . After Bonferoni correction , there were no consistent departures from Hardy-Weinberg equilibrium , or linkage disequilibrium . Only one locus in one population showed a significant departure from Hardy-Weinberg equilibrium ( transect I , site 8 for locus ARO-HJ2 ) , and there was only one significant association between loci ALU-MS06 and ARO-035 in one population ( transect IV site 3 ) . The primary genetic structure along each transect was studied using Bayesian clustering performed by the program STRUCTURE v2 . 1 [45] . We defined the number of populations ( K ) from 1 to 9 and 10 independent runs were performed for each value of K using the admixture model , a burn-in of 100 , 000 steps followed by 400 , 000 post burn-in iterations . We determined the optimal number of populations using the maximum value of the posterior probability of the data [45] . We also used AMOVA , performed by Arlequin v3 . 11 [46] , to test for genetic differentiation predicted by alternative speciation models . Within each transect populations were grouped by modal lineage , or , where appropriate , by habitat . For two transects ( III , IV ) , where both types of speciation could have occurred , this allowed direct comparison of competing speciation hypotheses . Finally , the mean genetic differentiation among populations either side of a lineage , or habitat , boundary along a transect was estimated by calculating the mean standardized pairwise FST′ using RecodeData v0 . 1 [47] and FSTAT v2 . 9 . 3 [48] . To give an independent perspective on the population affinities revealed by the Bayesian clustering we performed principal component analysis ( PCA ) of transect site gene frequencies using PCAGEN [49] . For each transect the PC1 site scores were compared to Bayesian site frequencies ( where K = 2 ) by correlation . The relationship between lineage , genetic isolation , past allopatry , ecotone , climate , and adaptive quantitative traits was investigated at sites along a series of replicated transects ( Figure 2 ) across the secondary contact zones ( transects I to VII ) and ecotone ( transects III and IV ) . Transect IX did not cross any lineage boundary or ecotone and was used as a control transect .
|
Over the last 150 years , since Darwin's study of islands and his “Origin of Species , ” island archipelagos have played a central role in the understanding of evolution and how species multiply ( speciation ) . Islands epitomise the conventional view of geographic ( allopatric ) speciation , where genomes diverge in isolation until accumulated differences result in reproductive isolation and the capacity to coexist without interbreeding . Current-day Martinique in the Lesser Antilles is composed of several ancient islands that have only recently coalesced into a single entity . The molecular phylogeny and geology show that these ancient islands have had their own tree lizard ( anole ) species for a very long time , about six to eight million years . Now they have met , we can genetically test for reproductive isolation . However , when we use selectively neutral markers from the nuclear genome , on this naturally replicated system , we can see that these anoles are freely exchanging genes and not behaving as species . Indeed , there is more genetic isolation between adjacent populations of the same species from different habitats than between separate putative allospecies from the ancient islands . This rejects allopatric speciation in a case study from a system thought to exemplify it , and suggests the potential importance of ecological speciation .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"evolutionary",
"biology",
"genetics",
"and",
"genomics/population",
"genetics",
"evolutionary",
"biology/animal",
"genetics",
"evolutionary",
"biology/evolutionary",
"ecology"
] |
2010
|
Genetic Tests for Ecological and Allopatric Speciation in Anoles on an Island Archipelago
|
The central event underlying prion diseases involves conformational change of the cellular form of the prion protein ( PrPC ) into the disease-associated , transmissible form ( PrPSc ) . PrPC is a sialoglycoprotein that contains two conserved N-glycosylation sites . Among the key parameters that control prion replication identified over the years are amino acid sequence of host PrPC and the strain-specific structure of PrPSc . The current work highlights the previously unappreciated role of sialylation of PrPC glycans in prion pathogenesis , including its role in controlling prion replication rate , infectivity , cross-species barrier and PrPSc glycoform ratio . The current study demonstrates that undersialylated PrPC is selected during prion amplification in Protein Misfolding Cyclic Amplification ( PMCAb ) at the expense of oversialylated PrPC . As a result , PMCAb-derived PrPSc was less sialylated than brain-derived PrPSc . A decrease in PrPSc sialylation correlated with a drop in infectivity of PMCAb-derived material . Nevertheless , enzymatic de-sialylation of PrPC using sialidase was found to increase the rate of PrPSc amplification in PMCAb from 10- to 10 , 000-fold in a strain-dependent manner . Moreover , de-sialylation of PrPC reduced or eliminated a species barrier of for prion amplification in PMCAb . These results suggest that the negative charge of sialic acid controls the energy barrier of homologous and heterologous prion replication . Surprisingly , the sialylation status of PrPC was also found to control PrPSc glycoform ratio . A decrease in PrPC sialylation levels resulted in a higher percentage of the diglycosylated glycoform in PrPSc . 2D analysis of charge distribution revealed that the sialylation status of brain-derived PrPC differed from that of spleen-derived PrPC . Knocking out lysosomal sialidase Neu1 did not change the sialylation status of brain-derived PrPC , suggesting that Neu1 is not responsible for desialylation of PrPC . The current work highlights previously unappreciated role of PrPC sialylation in prion diseases and opens multiple new research directions , including development of new therapeutic approaches .
Prion disease is a family of lethal , neurodegenerative maladies that can be sporadic , inheritable or transmissible in origin [1] . The key molecular event underlying prion diseases involves conformational change of the normal , cellular form of the prion protein denoted PrPC into the disease-associated , self-propagating , transmissible form denoted PrPSc [2] . Upon expression in the endoplasmic reticulum , PrPC undergoes posttranslational modifications , including attachment of up to two N-linked carbohydrates to residues Asn-181 and Asn-197 and of glycosylinositol phospholipid anchor ( GPI ) to the C-terminal residue Ser-231 ( residue numbers are given for hamster PrPC ) [3]–[5] . These posttranslational modifications are intact upon conversion of PrPC into PrPSc [4] , [6] , [7] . Since the discovery that the PrPSc and PrPC glycans are sialylated more than 25 years ago [6] , [8] , the potential role of sialylation in PrPC function , prion replication or its pathogenesis remains uncertain . The two N-linked carbohydrates can carry from zero to four terminal sialic acid residues each [8] , [9] . While the PrP polypeptide has a strong positive charge , the isoelectric points ( pI ) of PrPC can vary significantly in part due to variation in sialylation of the glycans [10]–[15] . In glycans sialic acid residues are linked to the galactose residues at the C-6 or C-3 positions [8] . The detailed site-specific characterization of mouse PrPC revealed that the majority of glycans at Ans-180 have bi- and triantennary structures and are sialylated to a lesser degree than the glycans at Ans-196 , a majority of which are tri- and tetraantennary structures [16] . While the relative proportion of bi- , tri- , and tetra-antennary glycans appears to differ slightly in PrPC and PrPSc , the relative proportions of sialylated glycans was found not to be statistically different between PrPC and PrPSc [9] . Due to diverse structure and composition of oligosaccharides , PrPC primary structure consists of more than 400 different glycoforms . In addition to sialylation of both glycans , a single sialic acid was also found on a GPI anchor of PrPC and PrPSc [3] . The ratio of di- , mono- , and unglycosylated PrPC glycoforms was found to change in favor of di-glycosylated forms in the course of neuronal differentiation , as well as upon an increase in the density of cells cultured in vitro [17] , [18] . While diglycosylated PrPC is the dominant glycoform in adult brain , the ratio of di- , mono- , and unglycosylated PrPC glycoforms was found to vary in different brain regions [19] . 2D-gel electrophoresis analysis revealed variations in isoelectric points ( pI ) of PrPC isoforms expressed in different brain regions [10] , a variation that could presumably be attributed , at least in part , to the region-specific differences in sialylation status of glycans . Moreover , as probed by binding of 19 lectins specific to different sugars including sialic acid residues , the composition of PrPC glycans was found to change with normal aging [20] . In the last decade , numerous studies illustrated the essential role of protein sialylation in immunity including its role in cell signaling , cell activation , differentiation , and pathogen recognition ( reviewed in [21]-[23] ) . While sialylation of cell surface proteins is also involved in a number of functions of central nervous system , including cell differentiation , adhesion and neuronal plasticity , a big gap in understanding the role of sialylation in the normal and pathological function of PrPC exists . Sialylation of PrPC glycans was shown to prevent binding of PrPC to selectins , a family of cell surface proteins that interact with carbohydrates in a Ca2+-dependent manner and participate in cell adhesion and migration [24] . A recent study examined the role of Siglec-1 , a sialic acid-binding immunoglobulin-type lectin , expression of which is restricted to mononuclear phagocytes , in prion diseases [25] . While mononuclear phagocytes are known to be important for prion uptake and trafficking to/within lymphoid tissue and possibly prion clearance , no effect of Siglec-1 knockout on peripheral prion disease pathogenesis was observed [25] . Another study examined possible involvement of GPI sialylation in neurodegeneration and found that a dense clustering of sialic acid-containing GPI anchors in the plasma membrane resulted in alteration of membrane composition and synapse damage [26] , [27] . The presence of sialic acid in the GPI was a requirement for the toxic effect expressed by clustering of PrPC molecules on cell surface [26] , [27] . In the current work , we examined the role of sialylation of PrPC on prion replication , a topic that has not been addressed in previous studies . We showed that charge heterogeneity in brain-derived PrPC and PrPSc was due to sialylation and that undersialylated PrPC molecules ( sialylated less than the statistical average for PrPC ) were a preferable substrate for prion amplification in PMCAb . As a result , PrPSc produced in PMCAb was less sialylated than brain-derived PrPSc and also showed longer incubation time to disease . Consistent with the idea that sialylation of PrPSc is important for prion infectivity , PMCAb material produced using desialylated PrPC was not infectious . Nevertheless , in support of the hypothesis that PrPC sialylation controls prion replication rate , de-sialylation of PrPC was found to speed up considerably PrPSc amplification in PMCAb , with the magnitude of this effect found to be strain-dependent . Moreover , de-sialylation of PrPC reduced or eliminated a species barrier of prion amplification in PMCAb . Furthermore , 2D analysis suggested that sialylation status of brain-derived PrPC was different from that of spleen-derived PrPC . Surprisingly , the PrPSc glycoform ratio was found to be controlled by the sialylation status of PrPC , with a decrease in PrPC sialylation levels resulting in a higher percentage of the diglycosylated glycoform in PrPSc presumably due to a decrease in density of negatively charged groups on PrPSc surface . The current study exposes the previously underappreciated role of PrPC sialylation in a number of key aspects of prion diseases , including its role in controlling prion replication rate , its infectivity , species barrier and PrPSc glycoform ratio .
In the absence of posttranslational modifications the prion protein has a strong positive charge at physiological pH . Theoretical calculation of the isoelectric point for full-length Syrian hamster PrP predicts a value 9 . 58 . However , due to posttranslational modifications and , primarily , sialylation of N-linked glycans and the GPI anchor , the actual isoelectric points of PrPC molecules could be substantially lower than 9 . 58 [14] . In fact , because each of the two N-linked glycans contains up to four terminal sialic acids ( Figure 1E ) , brain-derived PrPC molecules are heterogeneous with respect to their charge as confirmed by 2D gel-electrophoresis ( Figure 1A ) . To test whether sialic acid residues indeed account for broad charge heterogeneity , Syrian hamster normal brain homogenate ( NBH ) was treated with A . ureafaciens sialidase ( sialidase-treated NBH will be referred to as dsNBH ) , an enzyme that cleaves off terminal α2 , 3- and α2 , 6-linked sialic acid residues . While in non-treated NBH PrPC molecules are spread between pI 3 and 10 ( Figure 1A , top ) , enzymatic desialylation resulted in a substantial shift of PrPC towards pI 10 . A relatively intense spot at acidic pH in dsNBH appears to be due to aggregation of PrPC at low pH . Nevertheless , this experiment illustrates that PrPC charge heterogeneity is attributable at least in part to its variable sialylation status . Consistent with a previous study [20] , the diglycosylated form of PrPC in dsNBH migrated slightly faster on a 1D gel than that in non-treated NBH ( Figure 1D ) . Brain-derived , proteinase K ( PK ) -treated scrapie material from animals infected with the strains of natural or synthetic origin 263K or SSLOW [28] , respectively , also showed broad charge heterogeneity on 2D gels ( Figure 1B , C ) . When compared to the 2D profile of PrPC , the charge distributions of PK-treated 263K and SSLOW were found to shift toward pI 10 , despite an expected shift toward acidic pH due to proteolytic cleavage of the positively charged N-terminal region . The reason behind such a shift is difficult to explain . There is a possibility that PrPSc is less sialylated than PrPC , although no notable differences in sialylation status of PrPC and PrPSc were found in previous study [9] . Alternatively , a fraction of PrPC molecules could be subjected to posttranslational modifications including deamidation of Asn and Gln to Asp and Glu [29] , [30] , respectively , phosphorylation of serine 43 [15] , or modification of amino groups of Lys and Arg by reducing sugars resulting in advanced glycation end-products [5] , [31] . Such modification would increase PrPC charge heterogeneity and account for spreading PrPC to acidic pH on 2D . Attempts to remove sialic acid in PrPSc by treatment with sialidase were not successful , presumably due to high aggregation status of PrPSc ( data not shown ) . Because previous studies showed that properties of PrPSc change during PMCA [32] , [33] , we decided to compare charge distribution of PMCAb-derived and brain-derived PrPSc . To rule out any interference of the initial brain-derived PrPSc seeds , twenty four serial PMCAb rounds were performed with a dilution 1∶10 between rounds to produce PMCAb-derived material . Surprisingly , both 263K and SSLOW PMCAb-derived materials showed a considerable shift towards basic pI when compared to that of brain-derived PrPSc ( Figure 1B , C ) . Moreover , consistent with the previous study [32] , the percentage of monoglycosylated glycoforms decreased in PMCAb-derived material comparing to those of brain-derived PrPSc . These results suggest that ( i ) undersialylated PrPC molecules are selected during in vitro amplification at the expense of overersialylated PrPC ( sialylated more than the statistical average for PrPC ) and ( ii ) a decrease in PrPSc sialylation level reduces the negative charge on PrPSc surfaces that might lead to an increase in percentage of diglycosyated molecules incorporated into PMCAb-derived material . To provide independent support that undersialylated PrPC is a preferable substrate for PrPSc amplification in vitro , we tested whether desialylation of PrPC increases the rate of amplification using two alternative PMCAb formats . dsNBH prepared by treatment of NBH with A . ureafaciens sialidase ( Figure 1A ) was used as a substrate in PMCAb along with 10% non-treated NBH . In the first format , increasing dilutions of brain-derived 263K , Hyper , Drowsy and SSLOW materials were subjected to a single round of PMCAb conducted in dsNBH or NBH . The range of seed dilutions was chosen individually for each strain according to the previously published results [34] . For the hamster-adapted strains of natural origin ( 263K , Hyper ( HY ) , Drowsy ) , the reactions conducted in dsNBH detected approximately 10-fold higher seed dilutions than the reactions conducted in NBH ( Figure 2A ) . Surprisingly , for the synthetic strain SSLOW , the reaction conducted in dsNBH detected 104-fold higher seed dilutions than the reactions conducted in NBH ( Figure 2A ) . In fact , in dsNBH 108-fold diluted SSLOW brain material was persistently detected in a single PMCAb round . For two other strains of synthetic origin LOTSS and S05 [35] , [36] , the amplification rate also increase by at least four orders of magnitude in dsNBH compared to that in NBH ( data not shown ) . In a second format , a set of serial PMCAb reactions were conducted for 263K or SSLOW in NBH or dsNBH with the dilution folds between serial rounds ranging from 1∶30 to 2∶108 . The amplification rate is defined operationally as the highest dilution between PMCAb rounds at which amplification was still capable of compensating for the effect of dilution [34] . For 263K , the amplification rate increased approximately 50 fold , from 100-fold in NBH to 5000-fold in dsNBH ( Figure 2B ) . For SSLOW , the amplification rate increased more than 5×105 fold , from approximately 100-fold in NBH to at least 5×107-fold in dsNBH ( Figure 2B ) . Consistent with previous studies [32] , [34] , SSLOW showed an increase in signal intensity in serial PMCAb suggesting that it undergoes fast ‘adaptation’ to the PMCAb environment . Both experimental formats showed that desialylation of PrPC increases the rate of prion amplification in PMCAb , while the magnitude of an increase was strain-dependent . This effect was considerably higher for synthetic strains than strains of natural origin . In previous studies prion amplification in PMCA was shown to mimic key features of prion transmission including the species barrier [37]–[39] . A drop in amplification efficiency in PMCAb seeded with heterologous PrPSc will be referred to as the cross-seeding barrier . Considering that desialylation of PrPC increases the rate of PrPSc amplification , we were interested to test whether desialylation of PrPC eliminates the cross-seeding barrier in PMCAb . When serial PMCAb reactions in mouse NBH were seeded with hamster strains 263K or HY , stable replication was observed only after four or six serial rounds , respectively , a sign of a significant cross-seeding barrier ( Figure 3A ) . However , when mouse dsNBH was used as a substrate for 263K and HY , stable amplification was observed starting from the first round for both strains with no signs of cross-seeding barrier ( Figure 3A ) . In a reverse transmission experiment , two mouse strains 22L and ME7 were subjected to amplification in hamster NBH . No signal was observed for ten serial rounds suggesting that these two strains could not cross the barrier under the current experimental conditions ( Figure 3B ) . Surprisingly , when 22L and ME7 were subjected to amplification in hamster dsNBH , stable amplification was observed starting from the third or fourth serial round , respectively ( Figure 3B ) . Both experiments show that reducing sialylation levels of PrPC of a host species helps to eliminate or significantly reduce the barrier that prevents cross-seeding . Careful comparison of mouse-adapted 263K or HY revealed that the relative ratio of diglycosylated vs . monoglycosylated glycoforms was higher in dsNBH-amplified products than in NBH-amplified products ( Figure 3A , C ) . These changes suggest that the ratio of di- , mono- and unglycosylated glycoforms is not only a function of prion strain or host , but also depends on PrPC sialylation status . Due to glycan sialylation , the surface of PrPSc particles has a dense negative charge that creates electrostatic repulsion and , presumably , limits the percentage of diglycosylated PrPC able to be recruited by PrPSc . We propose that desialylation of PrPC eliminates electrostatic repulsion permitting a higher percentage of the diglycosylated glycoform to be accommodated . To test the hypothesis that PrPC sialylation controls the ratio of glycoforms within PrPSc , we examined the glycosylation profile of two mouse strains 22L and ME7 after their amplification in mouse NBH or dsNBH . Mouse strains were chosen because , in contrast to hamster strains , they have equal or even slightly higher percentage of monoglycosylated form relative to that of diglycosylated form . Upon amplification in dsNBH , the glycosylation profile of both 22L and ME7 changed immediately from predominantly monoglycosylated to predominantly diglycosylated ( Figure 4A ) . For comparison , the glycosylation profile of NBH-amplified 22L PMCAb-derived material remained very similar to that of brain-derived 22L ( Figure 4B ) . Due to very low amplification efficiency of ME7 in NBH ( Figure 4A ) , it was difficult to compare the glycosylation profile of ME7 NBH- versus dsNBH-amplified material directly . Nevertheless , after adjusting the amount of material loaded on a gel , ME7 amplified in dsNBH showed a considerably higher ratio of di- versus mono- or unglycosylated glycoforms than those observed in ME7 NBH-amplified or brain-derived material ( Figure 4B ) . Similar trend was observed for both hamster strains tested ( 263K and SSLOW ) . Monoglycosylated glycoform was well represented in brain-derived and in lower proportion in PMCAb-derived material , but absent in PMCAb-derived material produced in dsNBH ( Figure 4C ) . Taken together these results indicate that ( i ) the glycoform ratio within PrPSc is not only controlled by the strain or host but also by the sialylation status of PrPC; ( ii ) a decrease in PrPC sialylation levels results in a higher percentage of diglycosylated glycoforms in PrPSc; ( iii ) a shift toward diglycosylated glycoforms appears regardless of the host species , however the extent of the shift is likely determined by the strain-specific conformation . The lymphoreticular system plays an important role in prion pathogenesis , because: 1 ) prion replication in lymphoid tissues precedes neuroinvasion [40] , [41] , 2 ) lymphoid organs are targeted upon cross-species transmission and appear to be more permissive than central nervous system [42] , and 3 ) inflammation facilitates prion invasion [43] , [44] . Considering that desialylation of PrPC facilitates PrPSc replication , we were interested in comparing the sialylation pattern of brain-derived and spleen-derived PrPC . The 2D analysis of spleen tissues was very challenging because the level of PrPC expression in the spleen is 20 to 50-fold lower than that in the brain . Moreover , most of spleen-derived PrPC was proteolytically processed and formed a C2 fragment ( residues ∼100–231 ) that was immunoreactive with 3F4 ( epitope 109–112 ) and SAF-84 ( epitope 160–170 ) antibodies , but not Ab3531 ( epitope 90–102 ) antibody ( Figure S1 ) . This finding was in agreement with a previous report on N-terminal trimming of spleen-derived prion protein [45] . In addition , we found that the spleen-derived C2 fragment is highly prone to aggregation , as a significant proportion of it appeared as a dimer on SDS-gels ( Figure S1 ) . For the above reasons , 2D analysis of brain- and spleen-derived PrPC was performed using the SAF-84 antibody that reacts with full-length PrPC as well as C1 ( residues 111–231 ) and C2 fragments ( Figure 5 ) . Consistent with 1D SDS gel analysis , a significant percentage of full-length PrPC and the C2 fragment were persistently seen as dimers on 2D gels of spleen tissues ( Figure 5A ) . Pretreatment of spleen homogenates with sarcosyl did not help to reduce the amount of dimers . Nevertheless , a notable difference could be observed with respect to the charge distribution between spleen- and brain-derived monomeric full-length PrPC ( Figure 5A , B ) . In spleen material , the distributions of charge isoforms were shifted toward basic pH for both full-length PrPC and the C2 fragment . While the details of sialylation of spleen-derived PrPC remain to be investigated , this result suggests that the metabolism of PrPC sialylation in spleen might be different from that in a brain . To test whether PMCAb-associated changes in sialylation status affect prion infectivity , Syrian hamsters were inoculated with 263K brain-derived and PMCA-derived materials produced using NBH or dsNBH . We also attempted to produce desialylated PrPSc directly by treating brain-derived 263K with sialidase . However , this approach was not successful . For PMCAb conducted in NBH , 103-fold diluted 263K brain material was subjected to 24 serial rounds with 1∶10 dilution between rounds that resulted in a final dilution of brain material of 10−27 . For PMCAb conducted in dsNBH , 105-fold diluted 263K brain material was subjected to 7 serial rounds with 1∶1000 dilutions between rounds that resulted in a final dilution of brain material of 10−26 . On 2D gels , the charge distribution of products generated in PMCAb with dsNBH showed substantially more significant shift toward pI 10 and very little charge heterogeneity in comparison to the brain-derived or PMCAb-derived materials ( Figure 6A ) . The incubation time to the clinical signs of disease was 83±2 and 106±12 days in groups inoculated with brain-derived 263K and PMCAb-derived material , respectively . Two animals inoculated with PMCAb-derived material produced in dsNBH were sacrificed at 223 days p . i . , and no PK-resistant material was found in their brains ( Figure S2A ) . The remaining four animals from this group did not show any clinical signs and were sacrificed at 341 days p . i . No PK-resistant products were detected in their brain using 3F4 or SAF-84 antibodies ( Figure S2B ) . This result suggests that in contrast to the material produced in PMCAb with NBH , the material produced in PMCAb with dsNBH had no detectible infectivity despite being subjected to fewer PMCAb rounds . Lack of infectivity of PMCAb material produced in dsNBH could be due to changes in physical properties , particularly the high sensitivity to proteolytic clearance . To examine proteolytic resistance , PMCAb products generated in NBH or dsNBH were treated with increasing concentrations of PK ( Figure 6B ) . Indeed , for both strains 263K and SSLOW , the products formed using desialylated substrate were substantially less resistant than PMCAb products produced in NBH ( Figure 6B ) . Increased proteolytic sensitivity could lead to faster clearance of such inocula . In previous studies , we showed that 263K gave rise to a novel PrPSc conformation referred to as 263KR+ after 263K brain material was subjected to 12 serial PMCAb rounds in RNA-depleted NBH and then to additional 14 PMCAb rounds in NBH [46] . While 263KR+ amplified very fast in vitro , no clinical disease was observed upon inoculation of 263KR+ in first and second serial passages . We were interested in testing whether lack of infectivity could be due to changes in sialylation status of 263KR+ . 2D analysis revealed that the distribution of 263KR+ charge isoforms was considerably shifted to the basic pI in comparison to that of PMCAb-derived 263K ( Figure 6A ) . Therefore , for both 263KR+ and PMCAb-derived material produced in dsNBH the lack of infectivity correlates well with their low sialylation status . The catabolism of sialoglycoconjugates is regulated by four sialidases ( also referred to as neuraminidases Neu1 , Neu2 , Neu3 and Neu4 ) , all of which catalyze the removal of terminal sialic acid residues from carbohydrates of glycoprotein or glycolipids [47] . While all four sialidases are expressed in neuronal tissues , their levels and subcellular localization differ greatly . Because Neu1 is the most abundant , is expressed in the brain and is localized in lysosomes and on the cell surface , we decided to examine its role in regulating PrPC sialylation status . Brain materials from Neu1−/− , Neu1+/− and wild type mice of two genetic backgrounds ( FVB and BL6 ) were compared using 2D analysis . No notable differences with respect to PrPC charge distribution were observed between wild type , Neu1−/− or Neu1+/− mice of the two groups ( Figure S3 ) . The lack of effect could be because ( i ) Neu1 is not involved in PrPC desialylation , ( ii ) Neu1 deficiency is compensated by other neuraminidases , or ( iii ) PrPC molecules are degraded very fast in lysosomes , so the relative contribution of desialylated PrPC in the total pool of PrPC is very small . To probe the possible role of Neu1 in PrP catabolism further , we also examined the sialylation profile of PrP proteolytic fragment C1 ( residues ∼111–231 ) using SAF-84 antibody . C1 is present in large amounts in mouse brain and similar to PrPC , C1 can be found in di- , mono- and unglycosylated forms; however , the dynamics of the cellular clearance of C1 and its cellular localization could be different from those of full-length PrPC . No differences in sialylation status of C1 fragments were found between Neu1−/− , Neu1+/− and wild type mice ( Figure S3 ) . Because the life-span of Neu1−/− mice is limited to ∼150 days , it was not possible to test whether knocking out Neu1 affected the incubation time to prion disease .
Sialic acids are the most abundant terminal monosaccharides in cell membrane glycans [21] , [22] . Sialylation plays an essential role in key cellular functions including cell signaling , adhesion , differentiation , neuronal plasticity , cell-cell and cell-pathogen recognition , and the activation and trafficking of B and T lymphocytes , among other things [21] , [22] . The sialic acid content is the highest in embryonic and perinatal phases , but drops gradually during adulthood [48] , [49] . Brain and immune tissues including spleen have considerably higher amounts of sialic acid in their membrane fraction than other organs such as heart or kidney [48] . In the current study we showed that de-sialylation of PrPC increases PrPSc amplification rates in PMCAb . Faster amplification of desialylated substrates was likely due to removal of electrostatic repulsion between glycan moieties , which can carry up to 4 negatively charged sialic acid residues each ( Fig . 7A , B ) . This work suggests that a dense negative charge on the surface of PrPSc particles due to sialylated glycans prevents efficient PrPSc replication . Previous studies revealed that partial removal of N-linked glycans from PrPC using treatment with PNGase F or replacing a mixture of di- , mono- and unglicosylated PrPC with only unglicosylated form have a negative effect on PrPSc amplification in PMCA [50] , [51] . Taken together , these results indicate that while glycans are important for efficient amplification of PrPSc , the terminal sialic acid residues have a negative impact . Notably , the positive effect of PrPC desialylation on the replication rate , while present in all prion strains probed in this study , differed considerably in magnitude . Approximately 10–50 fold increases in the rate for strains of natural origin ( 263K , HY , Drowsy ) was strikingly different from the 10 , 000-fold increase for the strains of synthetic origin . Such drastic difference between strains of the two classes indicate that steric clashes between sialic acid residues in neighboring glycans are much more substantial in synthetic strains than in strains of natural origin ( Figure 7A , B ) . The elimination of the negative charges from PrPSc surface led to a much more significant drop in polymerization energy costs for synthetic strains than natural strains . This result highlights structural differences between the two classes of prion strains . The hypothesis that electrostatic repulsion between sialic residues controls PrPSc amplification rate explains why undersialylated PrPC molecules are preferentially recruited during in vitro amplification at the expense of oversialylated PrPC . As a result , the sialylation status of PrPSc changes during PMCAb becoming less sialylated in comparison to brain-derived PrPSc . In previous studies , prion specific infectivity ( the ratio of the infectivity titer to the amount of PrPSc ) was found to decrease gradually during amplification in serial PMCA [52] . A decrease in specific infectivity correlates well with a drop in sialylation status of PMCAb-derive material observed here . In support of this correlation , the current study observed longer incubation time to disease for PMCAb-derived material relative to that of brain-derived PrPSc , and a lack of clinical signs for PMCAb-derived material produced using desialylated substrate . Recent studies reported a gradual change in strain-specific secondary structure during serial amplification in PMCAb [33] . The relationship between changes in PrPSc conformation and sialylation status during serial PMCAb is not clear . Nevertheless , the fact that PMCAb material produced in NBH caused prion disease after 24 PMCAb rounds , whereas PMCAb material amplified in dsNBH did not cause the disease after only 7 amplification rounds , suggests that it is the loss of sialylation rather than a number of PMCAb rounds that had deleterious effect on infectivity . Progression of prion diseases is determined by a number of factors including PrPC-to-PrPSc conversion rate , PrPSc clearance , PrPSc deposition sites , and relative toxicity and size of PrPSc aggregates . Lack of clinical disease in animals inoculated with desialylated PMCAb products can be attributed in part to their high clearance rate . Consistent with this hypothesis , PMCAb materials produced with de-sialylated substrate were more sensitive to proteolytic digestion than standard PMCAb-derived material . Alternative mechanisms that involve interaction with microglia and cells of the immune system might also contribute to the lack of infectivity of PMCAb material produced in dsNBH . The mammalian immune system uses terminal sialylation of cell surface glycoproteins to identify pathogenic microorganisms to set them apart from their own cells , as microorganisms generally lack enzymes essential for sialic acid synthesis [21] , [22] . In the absence of terminal sialic residues , galactose is exposed as the terminal residue of glycans of microbial glycoproteins and serves as a signal for activating the immune response and phagocytotic clearance by macrophages [21] , [22] . If clearance of PrPSc involves mechanisms that are involved in clearance of microorganisms , desialylated PrPSc should be cleared much faster than sialylated PrPSc . Notably , some microbial pathogens recruit sialic acid from the host and sialylate their own glycoproteins in order to become “invisible” to the host's immune systems [23] . In addition to the clearance by macrophages , recent studies revealed that glycoclusters with terminal sialic acid were stable upon injection into mice and accumulated in the spleen , while the same clusters without sialic acid residues were rapidly excreted via the urinary tract [53] . It remains to be determined whether any of the above mechanisms account for lack of infectivity of desialylated PMCAb material . Nevertheless , fast clearance of PrPSc with low level of sialylation in a brain and luck of such clearance in PMCAb could explain why the sialylation levels of brain-derived and PMCAb-derived PrPSc are different . In previous studies , prions with high infectivity titers that lacked sialylation were generated in vitro using recombinant PrP [54] , [55] . Because entire carbohydrate groups were missing in PrPSc produced from recombinant PrP , it is unlikely that the immune system and microglia can identify these synthetic prions as potential pathogens in the same manner as it deals with desialylated PMCAb products . Consistent with this hypothesis , scrapie brain material from transgenic mice deficient in PrP glycosylation at both sites was found to be capable of infecting wild type mice [56] . Notably , transgenic mice that lacked glycosyls at both sites displayed a dramatic increase in the incubation time , incomplete attack rate or lack of infection [56] . These results suggest that in the absence of glycosylation/sialylation , PrPC of the host does not support well the infection or the newly formed PrPSc is not toxic . That PrPC sialylation controls PrPC-to-PrPSc conversion rate has far reaching implications . A decrease in PrPC sialylation could lead to a dramatic plunge of PrPC-to-PrPSc barriers in vivo and provide favorable conditions for ( i ) lowering the energy barrier of the spontaneous PrPC-to-PrPSc conversion in sporadic prion diseases; ( ii ) successful infection of a host or tissues with abnormally low sialylation status by low prion doses; and ( iii ) crossing the species barrier . In support of the last hypothesis , the current study revealed that hamster-to-mouse or mouse-to-hamster cross-seeding barriers can be reduced or abolished entirely in PMCAb , if de-sialylated PrPC is used as a substrate . While PMCAb can not predict the outcomes of transmission species barrier effects in whole organisms , our work opens an intriguing possibility that a species barrier is not only controlled by PrP amino acid sequence and PrPSc strain-specific structure but also by PrPC sialylation status . Noteworthy , because of an irreversible mutation in the gene encoding human N-acetylneuraminic acid hydroxylase , humans and the rest of mammalian species use different sialyc acid residues: humans produce only N-acetyl neuraminic acid ( Neu5Ac ) , while other mammals produce Neu5Ac and N-glycolylneuraminic acid ( Neu5Gc ) [57] . The difference in sialic acid structure affects interaction of pathogenic microbes with the immune systems of humans and other mammalian species . This difference might also contribute to a previously unappreciated mechanism that controls the prion transmission barrier between mammals and humans . Lymphoid organs are targeted upon cross-species transmission and appear to be more permissive than central nervous system [42] . 2D analysis of PrPC charge distribution revealed that spleen-derived PrPC was different from that of brain-derived PrPC with respect to their sialylation pattern , although precise comparison of the two tissues was complicated because of the low expression level , high tendency for aggregation and formation of the C2 proteolytic fragment by spleen-derived PrPC . Further research is needed more sensitive and accurate tools to confirm whether hyposialylation of PrPC in spleen makes the spleen more susceptible to prion infection than brain . Noteworthy , endogenous sialidase activity was found to increase in cells of the immune system , including lymphocytes and monocytes during cell activation and differentiation leading to undersialylation of cell surface glycoproteins [58] , [59] . Because inflammatory conditions support prion replication [43] , [44] , it would be interesting to examine in future studies whether inflammation-induced activation of sialidase gives rise to undersialylated PrPC and facilitates prion infection . Sialylation status of glycoproteins is controlled by sialyltransferases and sialidases ( also called neuraminidases ) , two classes of enzymes that transfer or cleave terminal sialic acids to/from glycoproteins , respectively [60] . In mammals , there are four sialidases ( Neu1 , Neu2 , Neu3 and Neu4 ) that are expressed in a tissue-dependent manner and differ with respect to their cellular localization and enzymatic properties [60] . Among the four sialidases , Neu1 is the most abundant and ubiquitously distributed . It is a part of a multi-enzyme , 1200 kDa hydrolase complex , which is localized predominantly in lysosomes and to lesser extent on the cell surfaces of many tissues and organs including brain [61] . To test whether Neu1 is responsible for desialylation of PrPC , brain materials from Neu1-/- knockout and Neu1+/− mice generated in two genetic backgrounds ( FVB and BL6 ) were analyzed using 2D gels . No significant changes in PrPC sialylation patterns were observed in Neu1−/− or Neu1+/− in comparison to those of corresponding wild type mice ( Figure S1 ) . These results indicate that either Neu1 is not involved in PrPC desialylation or Neu1 deficiency is compensated by other neuraminidases . Alternatively , PrPC molecules might be degraded so fast following desialylation , that the relative contribution of desialylated PrPC in the total pool of PrPC is too small to be detected by the current approach . In either case , Neu1 might not be the right target if one wants to alter the sialylation status of PrPC in vivo for therapeutic intervention . The ratio of di- , mono- and unglycosylated glycoforms within PrPSc is believed to be an intrinsic property of a prion strain or PrPSc subtype ( in sporadic prion diseases ) [62]–[65] . As such , the PrPSc glycoform ratio is used widely for strain typing and classification of PrPSc subtypes in sporadic CJD [64] , [65] , and changes in the glycoform ratio are thought to be indicative of a strain mutation or strain adaptation to a new host or environment [66] , [67] . Surprisingly , the current study revealed that the PrPSc glycoform ratio is not only controlled by prion strain or the host , but also by the sialylation status of PrPC . A decrease in PrPC sialylation levels resulted in a shift of the glycoform ratio toward diglycosylated forms at the expenses of mono- and unglycosylated glycoforms for both mouse and hamster strains . Such a relationship is explained well by the model that postulates that electrostatic repulsion created by sialic acid residues on the surface of PrPSc particles limits the percentage of diglycosylated molecules that can be accommodated within PrPSc ( Figure 7C ) . A decrease in sialylation levels reduces electrostatic repulsion leading to an increase in the percentage of diglycosylated molecules . In previous studies , treatment of prion-infected cultured cells with swainsonine , a compound that blocks synthesis of complex N-linked glycans , was shown to select minor strain variants or “mutants” resistant to swainsonine [68] . This process was accompanied by a change in the PrPSc glycoform ratio in favor of diglicosylated forms at the expense of monoglycosylated PrPSc glycoforms [68] , [69] . The extent to which swainsonine-related selection of minor variants and changes in glycoform ratio were due to lack of sialic acid residues is unclear . Recent studies established a possible link between protein sialylation and Alzheimer's diseases [70] . Deficiency of the lysosomal sialidase Neu1 was found to lead to the spontaneous occurrence of an Alzheimer's disease-like amyloidogenic process in mice . Loss of Neu1 resulted in accumulation of an over-sialylated amyloid precursor protein in lysosomes and excessive release of Aβ peptides by lysosomal exocytosis [70] . The current study opens a new avenue in prion research that might shed new light on the mechanism of prion replication and contribute to development of new therapeutic approaches . A number of sialic acid metabolic precursors or sialidase inhibitors are currently available and approved by FDA . Nevertheless , the impact and effectiveness of pharmacological intervention that target PrPC sialylation on progression of prion diseases are difficult to predict . The sialylation status of PrPC does not only control the rate of prion replication and magnitude of the species barrier , but is also likely to affect prion uptake and transport by macrophages , prion clearance rate , toxicity of PrPSc particles , and interaction of PrPSc with cells of the immune system and microglia . These topics have to be investigated in future studies . In summary , the current work demonstrated that hyposialylated PrPC molecules are a preferable substrate for prion amplification in PMCAb . PMCAb-derived PrPSc is less sialylated than brain-derived PrPSc . De-sialylation of PrPC significantly speeds up PrPSc amplification in a strain-dependent manner and significantly reduces or eliminates the species barrier . A decrease in PrPSc sialylation correlates with a drop in infectivity of PMCAb-derived material . The sialylation status of brain-derived PrPC appears to differ from that of spleen-derived PrPC . The sialylation status of PrPC controls the PrPSc glycoform ratio with a decrease in PrPC sialylation levels resulting in a higher percentage of the diglycosylated glycoform in PrPSc . Knocking out lysosomal sialidase Neu1 does not change the sialylation status of PrPC . The current work highlights the previously unappreciated role of PrPC sialylation in prion diseases and opens new directions in prion research , including development of new therapeutic approaches .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee of the University of Maryland , Baltimore ( Assurance Number A32000-01; Permit Number: 0309001 ) . Hyper and Drowsy scrapie brain materials were kindly provided by Richard Bessen ( Colorado State University , Fort Collins , CO ) ; 263K , 22L and ME7 scrapie brain materials were kindly provided by Robert Rohwer ( Veterans Affair Maryland Health Care System , Baltimore , MD ) ; SSLOW scrapie brain homogenate was prepared using animals from the 4th passage of SSLOW [71] . Neu1−/− , Neu1+/− and wild type mouse brains were collected from four month old FVB mice and five month old BL6 mice [72] . The mice were perfused with 20 ml PBS/5mM EDTA ( pH 7 . 4 ) , and then brains were collected and frozen in liquid nitrogen . Weanling Golden Syrian hamsters ( all males ) were inoculated intracerebrally under 2% O2/4 MAC isoflurane anesthesia . Each animal received 50 µl of brain homogenate or PMCAb products . After inoculation , hamsters were observed daily for disease using a ‘blind’ scoring protocol . Hamsters were euthanized as they approach the terminal stage of the disease . 10% normal brain homogenate ( NBH ) from healthy hamsters was prepared as described previously [35] and used as a substrate for PMCAb [39] . The sonication program consisted of 20 sec sonication pulses delivered at 170W energy output applied every 20 min during a 24 hour period . For each subsequent round , 10 or 20 µl of the reaction from the previous round were added to 90 or 80 µl of fresh substrate , respectively . Each PMCAb reaction was carried out in the presence of two 2/32” Teflon beads ( AmazonSupply . com ) . To analyze production of PK-resistant PrP material in PMCAb , 10 µl of sample were supplemented with 5 µl SDS and 5 µl PK , to a final concentration of 0 . 25% SDS and 50 µg/ml PK , followed by incubation at 37°C for 1 hour . The digestion was terminated by addition of SDS sample buffer and heating the samples for 10 min in a boiling water bath . Samples were loaded onto NuPAGE 12% BisTris gels , transferred to PVDF membrane , and probed with SAF-84 , Ab3531 or 3F4 antibodies . To produce de-sialylated substrate , 10% NBH from healthy hamsters prepared for PMCAb was treated with Arthrobacter ureafaciens sialidase ( cat # N3786 , Sigma-Aldrich , St . Louis , MO ) as follows . The lyophilized enzyme was dissolved in MilliQ water to the final concentration of 500mIU/ml . After preclearance of NBH at 500 × g for 2 min and addition of the buffer supplied by manufacturer , 7mIU/ml sialidase were added to the supernatant , and the reaction was incubated on a rotator at 37 °C for 5 h . The resulting substrate was used in dsPMCAb using the sonication protocol described for PMCAb . To prepare mock sialidase treated PMCAb substrate , the procedures were the same with adding MilliQ water instead of sialidase solution . Samples of 10 µl volume prepared in 1xSDS sample loading buffer as described above were solubilized for 1h at room temperature in 80 µl solubilization buffer ( 8M Urea , 2% CHAPS , 5mM TBP , 20mM Tris pH 8 . 0 ) , alkylated by addition of 135 µl 0 . 5M iodoacetamide and incubated for 1h at room temperature . Then , 1150 µl of ice-cold methanol was added , and samples were incubated for 2h at −20°C . After centrifugation at 13 , 000 rpm and 4°C , supernatant was discarded and the pellet was re-solubilized in 160 µl rehydration buffer ( 7M urea , 2 M thiourea , 1%DTT , 1% CHAPS , 1% Triton X-100 , 1% ampholyte , trace amount of Bromophenol Blue ) . Fixed immobilized pre-cast IPG strips ( cat . # ZM0011 , Life Technologies , Carlsbad , CA ) with a non-linear pH gradient 3–10 were rehydrated in 155 µl of resulting mixture overnight at room temperature inside IPGRunner cassettes ( cat . # ZM0008 , Life Technologies , Carlsbad , CA ) . Isoelectrofocusing ( first dimension separation ) was performed at room temperature with rising voltage ( 175V for 15 minutes; 175–2 , 000V linear gradient for 45 minutes; 2 , 000V for 30 minutes ) on Life Technologies Zoom Dual Power Supply using the XCell SureLock Mini-Cell Electrophoresis System ( cat . # EI0001 , Life Technologies ) . The IPG strips were then equilibrated for 15 minutes consecutively in ( i ) 6 M Urea , 20% glycerol , 2% SDS , 375mM Tris-HCl pH 8 . 8 , 130mM DTT , and ( ii ) 6 M Urea , 20% glycerol , 2% SDS , 375mM Tris-HCl pH 8 . 8 , 135mM iodoacetamide , and loaded on 4–12% Bis-Tris ZOOM SDS-PAGE pre-cast gels ( cat . # NP0330BOX , Life Technologies ) . For the second dimension , SDS-PAGE was performed for 1h at 170V . Immunoblotting was performed as described above .
|
The central event underlying prion diseases involves conformational change of the cellular form of the prion protein ( PrPC ) into disease-associated , transmissible form ( PrPSc ) . The amino acid sequence of PrPC and strain-specific structure of PrPSc are among the key parameters that control prion replication and transmission . The current study showed that PrPC posttranslational modification , specifically sialylation of N-linked glycans , plays a key role in regulating prion replication rate , infectivity , cross-species barrier and PrPSc glycoform ratio . A decrease in PrPC sialylation level increased the rate of prion replication in a strain-specific manner and reduced or eliminated a species barrier when prion replication was seeded by heterologous seeds . At the same time , a decrease in sialylation correlated with a drop in infectivity of PrPSc material produced in vitro . The current study also demonstrated that the PrPSc glycoform ratio , which is an important feature used for strain typing , is not only controlled by prion strain or host but also the sialylation status of PrPC . This study opens multiple new directions in prion research , including development of new therapeutic approaches .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases",
"veterinary",
"diseases",
"zoonoses",
"medicine",
"and",
"health",
"sciences",
"neurology",
"neurodegenerative",
"diseases",
"biology",
"and",
"life",
"sciences",
"veterinary",
"science",
"prion",
"diseases"
] |
2014
|
Sialylation of Prion Protein Controls the Rate of Prion Amplification, the Cross-Species Barrier, the Ratio of PrPSc Glycoform and Prion Infectivity
|
The cohesion of sister chromatids in the interval between chromosome replication and anaphase is important for preventing the precocious separation , and hence nondisjunction , of chromatids . Cohesion is accomplished by a ring-shaped protein complex , cohesin; and its release at anaphase occurs when separase cleaves the complex's α-kleisin subunit . Cohesin has additional roles in facilitating DNA damage repair from the sister chromatid and in regulating gene expression . We tested the universality of the present model of cohesion by studying cohesin in the evolutionarily distant protist Tetrahymena thermophila . Localization of tagged cohesin components Smc1p and Rec8p ( the α-kleisin ) showed that cohesin is abundant in mitotic and meiotic nuclei . RNAi knockdown experiments demonstrated that cohesin is crucial for normal chromosome segregation and meiotic DSB repair . Unexpectedly , cohesin does not detach from chromosome arms in anaphase , yet chromosome segregation depends on the activity of separase ( Esp1p ) . When Esp1p is depleted by RNAi , chromosomes become polytenic as they undergo multiple rounds of replication , but fail to separate . The cohesion of such bundles of numerous chromatids suggests that chromatids may be connected by factors in addition to topological linkage by cohesin rings . Although cohesin is not detected in transcriptionally active somatic nuclei , its loss causes a slight defect in their amitotic division . Notably , Tetrahymena uses a single version of α-kleisin for both mitosis and meiosis . Therefore , we propose that the differentiation of mitotic and meiotic cohesins found in most other model systems is not due to the need of a specialized meiotic cohesin , but due to additional roles of mitotic cohesin .
Cohesin is a ring-shaped protein complex which holds sister chromatids together to prevent their untimely separation prior to anaphase ( see [1] ) . It consists of four core components , Smc1 and Smc3 , Scc3 , and a member of the conserved α-kleisin family of proteins , Mdc1/Rad21/Scc1 in mitotic cells , or Rec8 in meiotic cells [1] , [2] . In mitotic cells , newly synthesized sister chromatids are linked by cohesin . In some organisms , chromatids separate along their arms during prophase and at the centromeres during anaphase , while in other organisms separation occurs in a single step during anaphase . The resolution of cohesion during anaphase is initiated by the cleavage of the α-kleisin component of the cohesin , the opening of the ring structure , and the disappearance of the cohesin from the chromosomes . This allows the mitotic separation of the sister chromatids . The fact that ring opening releases the chromatids , together with other evidence ( see [1] ) , led to the popular model that cohesin works by the enclosure of the two sisters by a single ring . During meiosis , homologous chromosomes first pair and become connected by a protein structure , the synaptonemal complex ( SC ) ( see [3] ) . At the same time , deliberate DNA double-strand breaks ( DSBs ) are formed and resolved in a way that leads to recombination [4] . Subsequent segregation of homologous chromosomes and sister chromatids is coordinated by the sequential loss of a specialized meiotic cohesin from the arms prior to the first meiotic division , and then from centromere regions prior to the second division . First , release of cohesin from regions distal to chiasmata allows the separation of homologous chromosomes , then release of centromeric or proximal regions separates sisters . Meiosis-specific regulation of cohesion works only when a specialized meiotic cohesin ( containing the meiosis specific α-kleisin Rec8 , and in some organisms also meiotic versions of other cohesin components ) is loaded onto chromosomes during the premeiotic S-phase ( for review see [5] ) . Kleisin cleavage in mitosis and meiosis is performed by a protein called separase , whose checkpoint-dependent activation relies upon the orderly association of chromosomes or bivalents with the division spindle , thus assuring their faithful disjunction ( see [1] , [6] ) . During meiotic prophase , cohesin is associated with the axial elements of SCs [7]–[11] and is involved in the repair and , in some organisms , the formation of meiotic DNA DSBs [12] . In addition to its well-established function in controlling chromatid and chromosome segregation , cohesin was found to play some additional roles , such as in DNA damage repair , where it promotes recombinational repair via sister DNA molecules [13] , [14] , in maintaining epigenetic inheritance states [15] , in gene regulation [16] ( and lit . cit . therein ) and in chromosome conformation [17] . Tetrahymena thermophila is a ciliated protist . Like the other ciliates , Tetrahymena possesses its soma and its germline organized as two nuclei within one cell ( see [18] ) . The germline micronucleus ( MIC ) carries a diploid number of 10 chromosomes that undergo closed mitosis and meiosis ( Figure 1 ) . The MIC genome is not expressed , and so the MIC is largely dispensable for vegetative proliferation of the cell; its only function is the propagation of heritable information during sexual reproduction . Transcription takes place only in the somatic macronucleus ( MAC ) . The MAC contains ∼45-fold amplified copies of the genome which are distributed in ∼180 minichromosomes that lack centromeres and remain uncondensed . It does not divide by a mitotic process , but splits into roughly equal parts prior to cell division . Tetrahymena cells of complementing mating types join ( “conjugate” ) under starvation conditions and initiate synchronous meioses in their MICs ( Figure 1 ) . The MAC does not undergo meiosis but it degenerates , and a new MAC regenerates from the MIC in new sexual progeny . Meiosis is characterized by several unusual features: First , there is no dedicated premeiotic S-phase; cells at ( micronuclear ) G2 conjugate and enter the meiotic program . Moreover , Tetrahymena lacks an SC [19] and , most notably , the MIC undergoes an immense elongation , to up to 50 times its original diameter , during meiotic prophase [20]–[22] . Within the elongated MIC , chromosomes are arranged in parallel , which facilitates homologous pairing and recombination [23] . Here , we take advantage of Tetrahymena's evolutionary distance from other commonly studied model systems to learn more about what features of cohesin are conserved , and what have been adapted to the needs of the particular organism . We also exploit the nuclear duality of Tetrahymena to separate functions of cohesin proteins that are important for chromosome division from functions related to gene expression and regulation . We demonstrate that Tetrahymena has evolved notable differences to the standard eukaryotic cohesion machinery . Investigation of these remarkable adaptations will lead to new insights into the flexibility of the chromosome cohesion and division processes .
We performed a bioinformatic search of the T . thermophila proteome [24] for the presence of cohesin components ( see Text S1 ) . The protein encoded by TTHERM_00245660 was the top and significant hit identified in searches for homologs to the Scc1 ( Mcd1 ) /Rad21/Rec8 family of α-kleisin proteins [2] . It produced a significant match in region aa 579–607 of the 619 amino acid-protein to the conserved C-terminal winged helix domain [25] . An additional weaker similarity was found for the N-terminal aa 30–130 ( Figure S1 ) . Based on the alignment of the conserved N- and C-terminal regions , we generated a phylogenetic tree that shows a relatively close sequence relationship of the group of mitotic α-kleisins , whereas meiotic members and TTHERM_00245660p show higher sequence divergence ( Figure 2 ) . Another protein , TTHERM_00219160p , is characterized by a C-terminal winged helix domain with more distant similarity to kleisins . It was found as the second best and significant hit in profile searches with the winged helix domain that was used to identify TTHERM_00245660p . This similarity was also confirmed in reciprocal searches ( see Text S1 , Figure S1 ) . However , sequence similarity beyond the C-terminal region could not be detected . TTHERM_00219160 mRNA has very low expression compared to the other cohesin gene homologs , and does not show significant up-regulation during conjugation , as the others do [26] . GFP tagging showed no detectable expression or localization of TTHERM_00219160p , even at high levels of replacement of the wildtype gene with the tagged gene ( data not shown ) . All of this evidence suggests that this is not a true kleisin protein with cohesin function , leaving TTHERM_00245660p as the only likely candidate . It is notable that Tetrahymena would have only a single α-kleisin homolog , because all eukaryotes studied to date in this respect have mitotic ( Scc1/Rad21/Mcd1 ) and meiotic ( Rec8 ) versions of this protein . Despite the lack of any closer relationship to the meiotic subgroup of α-kleisins ( Figure 2 ) , we will designate TTHERM_00245660p as Rec8p ( and the gene as REC8 ) in the following , because of its function in mitosis and meiosis ( see below ) . It shares this feature with the meiotic budding yeast Rec8 which , in principle , can take over the mitotic cohesion function [27]–[29] . Other core components of a putative cohesin complex show more conservation than Rec8 . ORFs TTHERM_01048090 and TTHERM_00294810 encode clear Smc1 and Smc3 homologs , respectively [24] , [30] . It is likely that the SMC3 ORF prediction is incorrect [24] , because cDNA coverage for SMC3 in the Tetrahymena Functional Genomics Database ( http://tfgd . ihb . ac . cn ) suggests a protein of only 1187 amino acids . To confirm that putative cohesin proteins form a complex in Tetrahymena , an immunoprecipitation ( IP ) was performed using cells expressing Rec8-GFP . Wild type cells were used for a control IP . Mass spectrometry analysis of the precipitating proteins showed that Smc1 and Smc3 were enriched 13× and 9× , respectively , in the Rec8-GFP pulldown sample over the untagged Rec8 control sample . Of 1837 identified proteins , Smc1 and Smc3 ranked 1st and 4th with respect to the number of unique peptides identified in the pulldown sample but only at positions 521 and 703 in the control ( Table S1 ) . Sequence coverage was 56% vs . 3% for Smc1 and 47% vs . 8% for Smc3 in the pulldown and control samples ( Figure S2 ) . This is strong evidence for the involvement of Rec8 , Smc1 and Smc3 in a protein complex and hence a confirmation of their correct identification as cohesin proteins . An Scc3/Rec11 homolog candidate , encoded by TTHERM_00225630 , was identified in profile searches versus the T . thermophila proteome with the region of best conservation from known Scc3/Rec11 orthologs , and was also confirmed in reciprocal searches ( see Text S1 ) . Although we did not find this homolog precipitating with Rec8-GFP , we will designate it as Scc3p . In addition to the core subunits of the cohesin complex , we were also able to identify the separase protein , another component of the cohesion/segregation machinery . TTHERM_00297160 is annotated in the Tetrahymena Genome database as the separase gene ESP1 [24] . The Esp1 protein sequence produced best reciprocal BLASTp hits with Saccharomyces cerevisiae Esp1 , Schizosaccharomyces pombe Cut1 , Caenorhabditis elegans SEP-1 , Arabidopsis thaliana AESP and human ESPL1 . We did not yet find obvious candidates for accessory cohesion proteins such as Eco1 , Pds5 , Wapl , sororin or securin . In order to observe the localization of cohesin in Tetrahymena , we used strains expressing tagged fusion proteins Rec8-GFP and Smc1-HA . Immunostaining against GFP or HA revealed a micronuclear localization of both proteins , in both vegetatively growing and sexually reproducing cells . This localization is maintained throughout all stages of mitosis and meiosis ( Figure 3A–3D , Figure S3A , Figure S4 ) . In addition , we tagged the putative Scc3p with mCherry and found its localization to be the same ( Figure 3E ) . In elongated meiotic nuclei , which correspond to the prophase stage of meiosis , Rec8p , Smc1p and Scc3p can be seen along the entire length of the chromosomes ( Figure 3C , 3D , 3E , Figure S4 ) . When only one of two conjugating meiotic cells carried the tagged proteins ( as shown for Smc1-HA and Scc3-mCherry in Figure 3D , 3E and Figure S4 ) , the protein was found retained in this cell . This is in contrast to the previously observed exchange of proteins between conjugating partners [31] and indicates a low turnover of cohesin proteins . Notably , in mitotic anaphase as well as in anaphase I and II of meiosis , the proteins could be seen along the full length of the stretched chromosome arms ( Figure 3A–3E ) . This is in contrast to the expected behavior shown by other organisms [7] , [32]–[36] , where cohesin disappears from chromosome arms during anaphase , but is similar to fission yeast mitosis [37] , where the separation of sister chromatids is not accompanied by the removal of cohesin . It is possible that the anaphase signal could result from cleaved cohesin rings that remain in the nucleoplasm or loosely associated with chromosomes during closed divisions . In order to address this possibility , detergent-spread fixations were performed in which free proteins are washed out of the MIC [38] . Immunofluorescent stainings of these fixations still showed the continuous localization of Smc1-HA and Rec8-GFP on chromosome arms throughout anaphase ( Figure 3F ) . This indicates that cohesin remains attached to chromatin during divisions . It was shown by Feulgen microspectrophotometry that in the Tetrahymena vegetative cell cycle , MIC DNA replication starts during division , as early as late anaphase or telophase [39] . We confirmed this staging by BrdU incorporation ( Figure S3B ) . This early replication may provide an explanation for the presence of cohesin along mitotic chromatids . Cohesin may be present in a non-cohesive form in order to establish cohesion as soon as replication starts . Tetrahymena allows separate observation of different cohesin functions because mitosis and transcription are carried out in separate nuclei within one and the same cell . We did not observe Rec8p nor Smc1p in the MAC of vegetatively growing or meiotic cells ( Figure 3A–3D , 3F ) . In principle , there should be no strict need for cohesin in the MAC , because the polyploid macronuclear chromosomes separate randomly by an amitotic process , and therefore it is not necessary to hold sister chromatids together after replication . However , an increasing number of non-cohesive functions of cohesin have been reported in studies of other model organisms ( for reviews see [1] , [6] , [40] ) . If cohesin performs these functions in Tetrahymena , it could conceivably be needed in the MAC as well as the MIC . Because cohesin has previously been shown to localize to sites of broken DNA [41] , [42] , we hypothesized that macronuclear cohesin might be induced by DNA damage . It has been shown in Tetrahymena that DNA damage induces expression of Rad51p in the MAC [43] . This suggests that a recombinational repair pathway is triggered that might rely on the physical linkage of sister DNA molecules by cohesin . Cells expressing Rec8-GFP were treated with UV-C , γ-radiation , MMS , cisplatin , or bleomycin ( see Materials and Methods ) , all of which are known to induce DNA damage in Tetrahymena [23] , [44] . Cells were fixed 0 . 5 or 1 . 5 h after treatment . Immunostaining against GFP showed no macronuclear localization while the DNA repair protein Rad51 was abundant in the MAC in response to the DNA damage ( Figure 3G ) . 200 cells with Rad51-positive MACs were evaluated per experiment . In all cases , Rec8p was detected in the MIC but never in the MAC . A similar experiment did not detect damage-induced Smc1-HA in the MAC , either ( data not shown ) . These results seem to indicate that cohesin does not localize to the MAC . Therefore , if cohesin functions in the MAC , it must be at concentrations too low to observe cytologically . To learn more about the function of cohesin in Tetrahymena , depletion of Rec8p or Smc1p was performed by RNA interference ( RNAi ) . For this , cells were transformed with constructs to express hairpin ( hp ) RNA molecules from a cadmium-inducible promoter , and RNAi was elicited in SMC1hp and REC8hp cells by growth in medium with CdCl2 . Induction of rec8RNAi ( rec8i ) in growing cells resulted in a very slow depletion of Rec8p , as can be seen by Western blotting in cells expressing both the Rec8-GFP and the REC8hp construct ( Figure 4A ) . After 4 h of RNAi induction ( approximately 1–2 cell divisions ) , large amounts of Rec8p remained in the cells , whereas REC8 mRNA was substantially reduced even after 2 h , as seen by RT-PCR ( Figure 4B ) . Only after 24 hours of RNAi , was the protein almost completely depleted ( Figure 4A ) . This data also confirms that Rec8p has a very low turnover in the cell ( see above ) , in contrast to RNAi targets studied previously , which needed only short periods of RNAi induction to produce phenotypes [38] , [45] . Depletion of Rec8p and/or Smc1p impaired the growth of cells . Viability was tested by isolating single cells of wild type ( WT ) , REC8hp , SMC1hp or REC8hp/SMC1hp strains ( n = 47 each ) in drops of medium containing 0 . 1 µg/ml CdCl2 to induce RNAi . After 48 h of growth , 100% of WT clones had more than 100 cells , whereas 0% of rec8i , 4 . 2% of smc1i and 6 . 4% of rec8i/smc1i clones had grown to the same density . Cells depleted of cohesins for at least 24 hours often showed lagging chromosomes in mitotic anaphase and , unlike in wild-type controls ( Figure 1A ) , the MACs began to elongate before mitosis was complete ( Figure 4C ) . This suggests that a lack of cohesion causes abnormal spindle attachments that result in a delay in chromatid segregation . rec8i/smc1i cells displayed 45% ( n = 100 ) abnormal anaphases whereas the phenotype was less prevalent in the smc1i line ( 13% ( n = 100 ) of anaphases were defective ) and absent in the rec8i line . Thus , the additive effect of the double RNAi is necessary to strongly affect segregation . Notably , telophase MICs appeared normal , presumably because lagging or misoriented chromatids eventually manage to migrate to the poles . MACs also had problems in splitting ( Figure 4C , Figure S5A ) , presumably as a consequence of delayed MIC mitosis . Cells often contained large amounts of DNA left between newly divided MACs , which is a less prevalent phenomenon in the wild type [46] . We next tested if depletion of Rec8p influences Smc1p localization . We constructed a strain that carries both the REC8hp and Smc1-HA . This strain showed the normal micronuclear localization of HA-tagged Smc1p ( in 200 of 200 evaluted nuclei ) when REC8 RNAi was not induced ( Figure 4D ) . However , when RNAi was induced ( rec8i ) , Smc1p staining was lost ( Figure 4D ) . In quantitative terms , in 96% of nuclei , the Smc1 signal was completely lost , and in 4% it was strongly reduced ( n = 200 nuclei ) . When this strain was mated to wild type , Smc1-HA localized to meiotic MICs ( in 200 of 200 evaluted nuclei ) , but mostly failed to do so upon RNAi depletion of Rec8 ( Figure 4D ) . The Smc1 signal was reduced or completely lost , in 8% and 85% of nuclei , respectively ( n = 200 nuclei ) . The dependency of Smc1 localization on Rec8 confirms a strong interdependence of Rec8 and Smc1 protein expression or stability , as would be expected for members of a protein complex such as cohesin . Moreover , it indicates that Rec8p is the predominant , if not only , kleisin partner of Smc1p , otherwise more Smc1p , which was forming a complex with another kleisin , would persist . To evaluate the meiotic phenotype of rec8i and smc1i cells , RNAi was performed on growing cells for 24 hours to deplete cohesin prior to inducing meiosis . The cohesin-depleted cells were then mated with a WT strain expressing Rec8-GFP . Due to the low turnover rate of the cohesion proteins ( see above ) , RNAi affected only the hairpin-carrying partner of mating pairs and allowed direct comparison of defective stages with the corresponding meiotic stages in the WT partner . DAPI staining of these mating cells showed that rec8i and smc1i cells begin meiosis normally and elongate the MIC , as in WT . However , whereas in WT cells , the MIC contracts and chromosomes condense to form bivalents in metaphase , in rec8i or smc1i cells , the chromosomes never fully condense ( Figure 4E , Figure S5 ) . In these cells , meiosis arrests at an abnormal metaphase-anaphase-like state , with clumps of fragmented and stretched chromatin , instead of distinct chromosomes , while the WT partner progresses normally through the meiotic divisions ( Figure 4E ) . Quantitation of the rec8i meiotic arrest showed that in pairs where the WT partner had progressed to metaphase or beyond , 70% of the rec8i partners were arrested in the fragmented metaphase state , 20% attempted to undergo an abnormal anaphase , and 10% showed a normal looking division ( Figure 4F ) . To determine if the rec8i arrest phenotype was dependent on meiotic DSBs , we depleted the Spo11 DSB nuclease by RNAi . spo11 RNAi of one cell inhibits or reduces DSB formation in the partner cell as well ( data not shown ) . When we mated rec8i cells with spo11i partners , the rec8i arrest phenotype was partially rescued . 60% of cells completed anaphase I , 35% showed abnormal divisions , and only 5% arrested with fragmented metaphase chromosomes ( Figure 4F ) . This suggests that rec8i cells have a defect in DSB repair . To further test if meiotic DSB repair was affected , matings of rec8i and WT cells were stained with antibodies against γ-H2A . X or the recombination protein Dmc1 . These DSB markers are normally seen primarily in the elongated pachytene MIC , and they disappear in later stages after DSBs are repaired [38] , as can be seen in the WT partner ( Figure 4G ) . However , γ-H2A . X and Dmc1 staining can still be seen in the rec8i cells , after their WT partners have completed DSB repair . This suggests that DSBs are not completely repaired in the absence of cohesin . The persistence of DSBs was also confirmed by electrophoretic detection of DSB-dependent chromosome fragments in mating rec8i cells ( Figure 4H ) . Whole cell DNA preparations of mating cells were separated by pulsed field gel electrophoresis , blotted , and probed with a repetitive sequence found only in the MIC [44] . Under the conditions used , unbroken MIC chromosomes do not enter the gel , but chromosomes fragmented by DSBs migrate as a mass , creating a distinct band . In WT cells , DSBs appeared at 3 hours after induction of meiosis , and disappeared at 5–6 h as breaks were repaired and meiosis was completed ( Figure 4H ) . In rec8i matings , however , the band representing meiotic DSBs appeared normally at 3 h , but seemed to accumulate to a higher level at 4 and 5 h , and did not disappear by 6 h after induction of meiosis . Therefore , we can conclude that cohesin is required for the repair of meiotic DSBs . Because defective DSB repair causes reduced meiotic pairing [38] , we wondered if pairing was affected by the absence of Rec8 . In meiosis of Tetrahymena , close pairing of homologous chromosomes is established in elongated MICs . Fluorescence in situ hybridization ( FISH ) can be used to monitor meiotic pairing of a chromosomal locus , which is indicated by the presence of a single FISH signal instead of two ( Figure 4I ) . While a single signal could also result from mitotic nondisjunction ( see above ) , rec8i did not show a mitotic phenotype . Therefore we were able to use rec8i cells to score pairing , using a probe to an intercalary chromosomal locus . In WT cells , 34 . 7% of meiotic prophase nuclei showed two FISH signals , corresponding to the unpaired homologous loci , and 61 . 3% showed only one signal , indicating pairing of that locus had occurred . In rec8i cells , pairing was reduced to 42 . 3% . ( Figure 4I ) . At the same time , FISH was also used to evaluate cohesion . 10 . 0% of rec8i elongated nuclei showed 3 or 4 signals , as compared to 4 . 0% in WT ( Figure 4I ) . Altogether , the depletion of cohesin results in a reduction of meiotic sister chromatid cohesion and homologous pairing . The continuous association of cohesin with mitotic and meiotic chromosomes during and after anaphase led us to ask if the separase cleavage mechanism used in other organisms was conserved in Tetrahymena . We therefore designed an RNAi construct to target Tetrahymena's separase homolog , ESP1 . After induction of the esp1i construct in vegetative growth , cells immediately showed problems dividing the MICs . Normally the MIC divides before the MAC ( Figure 1A and Figure S3 ) , whereas in esp1i cells , MICs arrested in an anaphase-like state by the time the MAC divided ( Figure 5A ) . As a result , the cleavage furrow formed while the MIC was still attempting division , leaving one daughter cell containing the undivided MIC , and one cell without a MIC ( Figure 5A ) . Staining for phosphorylated ( Ser10 ) histone H3 , which is a marker for chromosome condensation [47] , showed that the MICs stay in a mitosis-like condensation state ( Figure 5A ) . After 24 h of esp1i induction , only 10 . 5% of cells contained MICs , as opposed to 98 . 5% of uninduced cells . This decreased to 7% after 48 h , and 3% after 72 h of esp1i induction ( n = 200 cells per genotype and timepoint ) . After 48 h of esp1i induction , the remaining MICs were very large , presumably as a result of polyploidization by numerous rounds of replication and failed separation . Many of these large MICs had a ropy appearance , suggesting a bundled , polytenic state of chromosomes ( Figure 5B ) . FISH staining of a unique chromosomal locus showed signal clusters whose strength indicated that they contained multiple copies , again suggesting a polytenic organization of unseparated chromatids ( Figure 5C ) . To obtain a rough estimate of the degree of polyploidization , FACScan was performed of cultures 48 h after esp1i induction . While nuclei from WT cells sorted into two clear peaks corresponding to MICs and MACs , esp1i cells showed only one peak for both MICs and MACs ( Figure 5D ) . This indicates that aberrant MICs may reach roughly the same DNA content as MACs and suggests that they may become 16–32-ploid . When two ESP1hp cells were mated and RNAi was induced during meiosis , MICs elongated normally and chromosomes condensed at metaphase I . However , most cells arrested in anaphase I , and chromosomes coalesced back to enlarged undivided MICs ( Figure 5E , 5F ) . Together , these observations suggest that cohesin cleavage does not occur in the absence of Esp1p , which prevents sister chromatid separation in mitosis and the separation of homologs in meiosis I .
Efficient depletion of cohesin proteins caused problems with MIC chromosome segregation in mitosis , which is suggestive of loss of cohesion . Loss of Rec8p also led to a reduction of cohesion in meiosis . It has been shown in other organisms that cohesion is relatively insensitive to reduced amounts of cohesin [48] , [49] whereas even small decreases in the levels of cohesin drastically affect meiotic DNA repair [7] , [48] , [50] , [51] . This explains why meiosis is more affected than mitosis in the rec8i and smc1i single knockdown strains . Also , cohesin-independent mechanisms , such as relational coiling or catenation of the sister chromatids , protein-protein interactions , or locally delayed replication could provide additional forces to hold sister chromatids together and contribute to proper segregation [52]–[54] . After long periods of cohesin depletion , mitotic cultures deteriorated . However , this was not due to MIC aneuploidy caused by segregation problems , because Tetrahymena can propagate with a defective or severely aneuploid MIC for numerous generations [55] . In contrast to MICs , MACs did not display cytologically detectable cohesin . Not even upon artificial DNA damage could we induce notable cohesin loading , although the Rad51 DNA damage response was robust . Nevertheless , depletion of cohesin results in a defect in MAC division . Dividing MACs in cohesin-depleted cells leave behind substantial amounts of DNA in the cleavage furrow , and this DNA remains in the cells as extra-nuclear bodies . This defect , along with the growth defect observed in cohesin depleted cells , may indicate a role of cohesin in normal cell division , either through a direct function in the MAC , or through cross-talk between the MIC and the MAC [56] . We have shown that meiotic cells depleted of either Rec8p or Smc1p arrest prior to anaphase I . Similar to the situation in budding yeast [7] , [57] , [58] and C . elegans [8] , meiotic DSBs are formed normally in the absence of Rec8p . However , these breaks are not repaired , which likely causes the prophase arrest . Cohesion has been previously found to be crucial for meiotic DSB repair , to the extent that even a partial loss of cohesin function can result in unrepaired breaks [7] , [48] , [50] , [51] . Exactly how cohesin participates in DNA repair is not known . It is thought that cohesin facilitates repair directed by the sister chromatid by keeping sisters tightly bound together ( see [59] ) . However , one could speculate that close access to the sister chromatid would be less crucial for meiotic DSB repair , which can , and often prefers , to use the homologous chromosome as a template . Therefore , it seems likely that cohesin plays a more direct role in DNA repair than simply providing cohesion , because the repair function is much more sensitive to slight perturbations in cohesin dosage [48]; see [14] . We have also shown that depletion of cohesin in Tetrahymena results in a moderate reduction in meiotic pairing . In other organisms , cohesin forms the backbone of the SC , a structure that is central to meiotic recombination ( see Introduction ) . Although Tetrahymena does not form an SC , it is quite likely that cohesin plays a role in defining chromosome axes and providing the necessary architecture to support homologous recombination . Precise pairing of homologs in meiosis of Tetrahymena is dependent on homologous recombination [23] , and so the reduction of pairing in the absence of cohesin may be a result of the defect in DSB repair . The persistence of cohesin on anaphase chromosome arms might suggest that dissolution of cohesion works by a mechanism different from kleisin cleavage . However , we have shown that depletion of Esp1p prevents the segregation of chromosomes in both mitosis and meiosis . This phenotype is consistent with the canonical separase-dependent loss of cohesion . There are several possible explanations for the maintenance of Rec8p and Smc1p on chromosome arms: First , it may be related to the absence of an apparent G1 interval in Tetrahymena and the early onset of MIC DNA replication during late anaphase or telophase [39] ( see also [60] ) . Thus , the permanent presence of cohesins could be due to the temporal overlap of their removal from separating sisters in anaphase and loading to chromatin in preparation for cohesion establishment during replication , shortly after . This explanation may seem less applicable to the association of cohesins with anaphase I chromosome arms , since this division is not followed by DNA replication . However , the two meiotic divisions occur in rapid succession and are followed quickly by post-meiotic replication [60] . Therefore , carrying non-cohesive cohesin along on the chromosomes may allow a more rapid establishment of cohesion . To maintain this “dormant” population of cohesin on chromosomes , there could be a mechanism in which all cohesin rings are opened by separase cleavage of Rec8p and yet continue to bind to chromatin . Because this would require replacement of the cleaved Rec8 , it seems more likely that only a small subset of cohesin complexes connect sisters , and only the intersister subpopulation is removed by cleavage . This interpretation was preferred by Tomonaga et al . ( 2000 ) [37] for their similar observation in fission yeast mitosis . Prolonged depletion of Esp1p in vegetative cells results in many cells without MICs and a few cells with grossly enlarged MICs , presumably as a result of multiple rounds of replication without division . Polyploidization was also found after inactivation of separase in several mouse tissues , in D . melanogaster and in non-yeast fungi , whereas budding yeast chromosomes break during attempted separation and daughter cells die as a result [61] ( and lit . cit . therein ) . In Tetrahymena , the chromosomes of polyploid MICs appear to be polytenic , because FISH against one chromosomal arm locus shows clustered or banded signals . The embrace model of cohesion limits the number of 10-nm chromatin fibers that can be encircled by a cohesin ring of an estimated diameter of ca . 40–45 nm to about four to six . This is hardly consistent with our observation of ∼4 chromatid bundles and the estimation of 16–32-ploid chromatid content . Thus , chromatid bundles may be formed by cohesin rings randomly encircling sisters as they are replicated within the bundle , so that a large , networked , multi-sister bunch is created , instead of individual pairs of sisters . Alternatively , some form of non-embrace-type of cohesion could be employed ( see [1] ) . We have found that Tetrahymena possesses components of the canonical cohesion-segregation machinery characterized in other model systems such as Saccharomyces cerevisiae , Schizosaccharomyces pombe , Arabidopsis and mice . Interestingly , only one α-kleisin homolog was found in Tetrahymena , which functions in both mitosis and meiosis . In contrast , most other organisms ( with the possible exception of Drosophila [62] ) have two paralogs , e . g . the mitotic Scc1/Mcd1 and meiotic Rec8 of yeast . Mammalian systems even have three homologs: Rad21 , Rec8 , and Rad21L , with the latter two being meiosis specific [63]–[65] . One reason for having only one multipurpose α-kleisin in Tetrahymena is its lack of a dedicated pre-meiotic S phase [66] , [67] . Cells go from G2 alternatively into mitosis or meiosis , and therefore there is no opportunity to load a specific cohesin . Special features in the meiotic kleisin have been considered important for maintaining cohesion of sister centromeres in meiosis I so that they segregate together . This is in contrast to the mitotic cohesin , which ensures sisters segregate to opposite poles ( for review see [5] ) . It will be interesting to learn how Tetrahymena Rec8p performs both functions . On the other hand , meiotic Rec8 can take over most mitotic cohesin function in budding yeast [27]–[29] . Thus , the primary question may not be how Tetrahymena can make with a single α-kleisin , but rather why budding yeast and the others require a mitotic and a meiotic version . It was found that Scc1 , but not Rec8 , can induce DSB-dependent cohesion in mitosis [68] , and it is conceivable that Scc1 also performs better in non-canonical cohesin roles such as gene regulation . Thus , a specialized kleisin may have diverged during evolution to be optimized for functions that may be of subordinate importance in Tetrahymena , due to its allocation of gene expression and propagation to different nuclei . Clearly , the function of cohesin in Tetrahymena has many parallels with previously studied organisms . However , future studies are needed to explore the numerous differences that we have found , including Tetrahymena's use of a single α-kleisin and the continuous association of cohesin with chromatin throughout anaphase .
Tetrahymena thermophila strains B2086 and Cu428 served as wild types and as the source strains for transformation with RNAi constructs . Cells were propagated vegetatively at 30°C according to standard methods ( see [69] ) . For meiosis experiments , cells were grown to a density of ∼2×105 cells/ml and made competent for conjugation by starvation in 10 mM Tris–Cl ( pH 7 . 4 ) for at least 16 h . Conjugation and meiosis were induced by mixing starved cultures of different mating types . The ESP1hp construct was created by amplifying a ∼500 bp region of the ESP1 ORF and cloning it into the rDNA based RNAi vector to create a hairpin expression cassette [45] ( see Table S2 for primer sequences ) . Transformation and selection was performed as previously described [38] . For REC8 RNAi , the transformation strategy was changed slightly . A new vector was created using pBS-CHX , a vector targeting knock-ins to the RPL29 gene which also confers cycloheximide resistance ( gift of Chad Pearson ) . The hairpin expression cassette was released from the rDNA vector using NotI digestion and cloned into the NotI site of the pBS-CHX multiple cloning site ( MCS ) . Then the XmaI site in the MCS was destroyed by blunting and religation in order to allow direct cutting and pasting of hairpin fragments into the PmeI/XmaI and ApaI/XhoI halves of the expression cassette . An ∼500 bp fragment of the REC8 gene was then amplified from genomic DNA using primers that added the appropriate restriction sites , and these fragments were cloned into the two sides of the hairpin cassette . For a map of vector REC8hpCYH see Figure S6 . This construct was digested with BlpI and introduced into vegetatively growing Tetrahymena by biolistic transformation . Cells were grown overnight , then selected in 7 . 5 µg/ml cycloheximide . Transformants were successively grown in higher concentrations of cycloheximide up to 20 µg/ml , to increase the macronuclear copy number of the hairpin containing chromosome . In order to create a strain for double RNAi of both REC8 and SMC1 , it was necessary to use a different selection marker for the SMC1hp , therefore another RNAi vector was created using pMNMM3 ( gift of Kazufumi Mochizuki ) , which carries the NEO5 ( paromomycin resistance ) gene and targets knock-ins to the MTT1 locus , utilizing the MTT1 cadmium-inducible metallothionein promoter for expression . To facilitate cloning into this vector , a new hairpin linker was amplified from genomic DNA using primers that introduced BamHI and PmeI restriction sites on one end , and XmaI and PstI sites on the other end . This was cloned into the MCS of pMNMM3 , and then the ∼500 base pair fragment of SMC1 was amplified and cloned into either end of the linker as in previous constructs . For a map of vector SMC1hpNEO see Figure S6 . The construct was digested with NotI and XhoI and introduced into vegetatively growing Tetrahymena ( either WT strains or REC8hp strains ) by biolistic transformation [70] . Transformants were selected in 120 µg/ml paromomycin , then grown in successively higher concentrations of drug , up to 2 mg/ml . The SPO11hp construct was prepared in the same way as the REC8hp , using the primers listed ( Table S2 ) . Transformation and selection was also performed as for rec8i . In all cases , RNAi was induced by expression of dsRNA from the MTT1 promoter by the addition of 0 . 2 µg/ml of CdCl2 ( final concentration: ) to cells carrying the hairpin construct . Rec8-GFP , Smc1-HA and Scc3-mCherry expressing cells were created using a knock-in approach to fuse the tag to the C-terminus of each gene at its native genomic locus . Tagging constructs were created by amplifying the last ∼500 bp of the ORF as well as a ∼500 bp region downstream of the gene , and fusing these two products with the tagging cassette using overlapping PCR . ( See Table S2 for primer sequences ) . The tagging cassettes were amplified from pHA-Neo4 or pEGFP-Neo4 ( gifts of Kazufumi Mochizuki ) . Tagging constructs were introduced into the MAC of vegetatively growing B2086 and Cu428 cells by biolistic transformation . Transformants were selected in media containing paromomycin in increasing concentrations up to 50 mg/ml . For detecting Rec8-GFP by Western blotting , protein extracts were prepared by trichloroacetic acid precipitation , run on 10% SDS-PAGE gels and blotted . Membranes were incubated with anti-GFP antiserum and with appropriate HRP-conjugated secondary antibody , and the protein bands were detected by chemiluminescence . Immunoprecipitation against GFP tag was performed using magnetic GFP-trap beads ( ChromoTec , Martinsried , Ger ) , according to the manufacturer's protocol . Lysates were created by sonicating 5×107 mating cells ( 5 h after mixing ) in 2 ml of lysis buffer +1 mM PMSF and protease inhibitors ( 1× cOmplete mini , Roche , Indianapolis , IN ) , 4×25 sec at 37% power , duty cycle 5 . After 2 h incubation with lysate , the beads were washed 7×5 min with wash buffer ( 10 mM Tris/Cl pH7 . 5 , 150 mM NaCl , 0 . 5 mM EDTA ) , and 2×5 min with 150 mM NaCl . The bead bound proteins were trypsinized , and peptides were loaded on a Dionex UltiMate 3000 HPLC system ( Thermo Scientific , San Jose , CA ) . Peptides separated in a 0 . 1% formic acid/0% acetonitrile – 0 . 08% formic acid/80% acetonitrile gradient in water were injected into the mass spectrometer via an electrospray-interface . MS/MS analysis was carried out with a Q Exactive mass spectrometer ( Thermo Scientific ) , and peptide spectra were recorded over a mass range of 350–2000 m/z ( for details see [71] ) . For peptide identification , the . RAW-files were loaded into Proteome Discoverer 1 . 4 . 0 . 282 ( Thermo Scientific ) . MS/MS spectra created were searched using Mascot 2 . 2 . 07 ( Matrix Science , London , UK ) against the NCBI non-redundant protein sequence database , using the taxonomy group Alveolata . For induction of DNA damage by UV irradiation , 5 ml aliquots of cells were placed in a 90 mm plastic Petri dish . Open dishes were placed in a Stratalinker crosslinker and treated with 254 nm UV ( UV-C ) at a dosage of 150 Joules/m2 . Treatment with ionizing radiation was performed by exposure to 5000 rads of γ-radiation from a 137Cs source . For chemical induction of DNA damage , cells were treated with 100 µg/ml cisplatin ( from a 2 mg/ml stock solution in 10 mM Tris-HCl ) , 50 µg/ml bleomycin ( from a 10 mg/ml stock solution in 10 mM Tris-HCl ) or 4 mM methyl methane sulfonate ( MMS , from a 100 mM stock solution in 10 mM Tris-HCl ) . Cells were fixed and prepared for immunostaining ( see below ) . Immunostaining of DNA repair protein Rad51 allowed cytological detection of DNA damage in the MAC . Cell suspensions were centrifuged and Carnoy's fixative ( methanol , chloroform , acetic acid 6∶3∶2 ) was quickly added to the pellet . This disrupts the cells and separates MICs and MACs . Carnoy's fixative was replaced by 70% ethanol after 1 h at room temperature , and the fixations were stored in the freezer . Shortly before measuring , the nuclei were pelleted by centrifugation , resuspended in 1× PBS and stained by the addition of DAPI ( 4′ , 6-diamidino-2-phenylindole; final concentration: 0 . 2 µg/ml ) . Different preparation methods were applied for subsequent immunostaining and FISH . A combined formaldehyde fixation and detergent permeabilization treatment [72] was used for subsequent chromosome staining with DAPI or for immunostaining of Dmc1p/Rad51p , PhosH3 , and GFP and HA tags . An enforced detergent spreading method for the removal of free nuclear proteins [44] was applied for probing chromatin associated proteins . For immunostaining of γ-H2A . X , cells were fixed with Schaudinn's fixative , washed and resuspended in methanol [73] . For subsequent FISH , cells were fixed with Carnoy's fixative ( methanol , chloroform , acetic acid 6∶3∶2 ) . Drops of fixed cell suspensions were dried down on slides . For immunostaining , slides were washed with 1×PBS and 1×PBS+0 . 05% Triton , incubated with primary antibodies over night at 4°C , washed again , incubated with appropriate FITC- or Cy3-labeled secondary antibodies for 2 h , and washed and mounted with anti-fading buffer supplemented with 0 . 5 ìg/ml DAPI . The following primary antibodies were applied: Mouse monoclonal antibody against the related DNA repair proteins Dmc1 and Rad51 ( 1∶50 , Clone 51RAD01 , NeoMarkers , Fremont , CA ) , mouse anti-γ-H2A . X antibody ( 1∶200 , BioLegend , San Diego , CA ) , rabbit anti-phosphorylated H3Ser10 ( 1∶500 , Upstate Biotechnology , Charlottesville , VA ) , rabbit anti-GFP ( 1∶100 , Molecular Probes , Eugene , OR ) , monoclonal mouse anti-HA ( 1∶50 , Roche Diagnostics GmbH , Mannheim , GER ) and rabbit anti-HA ( 1∶100 , Sigma St . Louis , MO ) . For FISH , a probe against an intercalary chromosomal region , scaffold scf_8254686 ( http://ciliate . org ) , the same as described in ref . [23] , was used . DNA on slides and the Cy3-labeled hybridization probe were denatured and hybridized for 36–48 h at 37°C [72] . In all cases , chromatin was counterstained with DAPI , and z stacks of pictures were taken under a fluorescence microscope equipped with the appropriate filters . Picture stacks were deconvolved , projected , assigned false colors , and multicolor images were merged . For details of the detection of DSB-generated fragments by PFGE , see [44] . In short , chromosome-sized DNA was prepared in agarose plugs . The run was performed in 1% agarose with 0 . 5× TBE buffer at 200 V , 6°C for 14 h with 60-s pulses , 10 h with 90-s pulses , and 1 h with 120-s pulses in a Bio Rad Chef-DR III system . Under these conditions , intact MIC chromosomes do not enter the gel , whereas fragments migrate as a single band . Since numerous small MAC chromosomes are distributed along the entire gel and cover the DSB-generated signal , MIC-borne DNA fragments were highlighted by Southern detection of a MIC-specific DNA [44] . The membrane was stripped and re-hybridized with a probe against a 121-kb MAC chromosome as a marker to test equal DNA loading and Southern transfer for different time points .
|
During cell division , identical DNA molecules , packaged in the sister chromatids of a chromosome , must be distributed to daughter cells . The cohesion of sister chromatids in the interval between DNA replication and mitotic anaphase is important for preventing the precocious separation , and hence nondisjunction , of chromatids . Cohesion is accomplished by a ring-shaped protein complex , cohesin; and a popular model of cohesion holds that sister chromatids are encircled by cohesin rings and separate upon opening of the rings . During meiosis , cohesin , together with chiasmata , has the additional function of holding bivalents together . Cohesin also has functions in gene regulation and DNA damage repair , and has recently garnered attention as a factor involved in human congenital birth defects . We have studied cohesin in the protist Tetrahymena , which has mitosis/meiosis and transcription performed by different nuclei within the same cell . We exploited this unique feature to experimentally separate the functions of cohesin in chromosome segregation and gene regulation . While the cohesin machinery is generally conserved between eukaryotes , Tetrahymena's phylogenetic distance from standard model organisms allowed us to discover some notable adaptations during the course of evolution .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"meiosis",
"chromosome",
"structure",
"and",
"function",
"microbiology",
"mitosis",
"model",
"organisms",
"cell",
"division",
"tetrahymena",
"thermophila",
"chromosome",
"biology",
"protozoan",
"models",
"biology",
"cell",
"biology",
"cytogenetic",
"techniques",
"genetics",
"molecular",
"cell",
"biology",
"cytogenetics",
"genetics",
"and",
"genomics"
] |
2013
|
A Single Cohesin Complex Performs Mitotic and Meiotic Functions in the Protist Tetrahymena
|
Measures of allele and haplotype diversity , which are fundamental properties in population genetics , often follow heavy tailed distributions . These measures are of particular interest in the field of hematopoietic stem cell transplant ( HSCT ) . Donor/Recipient suitability for HSCT is determined by Human Leukocyte Antigen ( HLA ) similarity . Match predictions rely upon a precise description of HLA diversity , yet classical estimates are inaccurate given the heavy-tailed nature of the distribution . This directly affects HSCT matching and diversity measures in broader fields such as species richness . We , therefore , have developed a power-law based estimator to measure allele and haplotype diversity that accommodates heavy tails using the concepts of regular variation and occupancy distributions . Application of our estimator to 6 . 59 million donors in the Be The Match Registry revealed that haplotypes follow a heavy tail distribution across all ethnicities: for example , 44 . 65% of the European American haplotypes are represented by only 1 individual . Indeed , our discovery rate of all U . S . European American haplotypes is estimated at 23 . 45% based upon sampling 3 . 97% of the population , leaving a large number of unobserved haplotypes . Population coverage , however , is much higher at 99 . 4% given that 90% of European Americans carry one of the 4 . 5% most frequent haplotypes . Alleles were found to be less diverse suggesting the current registry represents most alleles in the population . Thus , for HSCT registries , haplotype discovery will remain high with continued recruitment to a very deep level of sampling , but population coverage will not . Finally , we compared the convergence of our power-law versus classical diversity estimators such as Capture recapture , Chao , ACE and Jackknife methods . When fit to the haplotype data , our estimator displayed favorable properties in terms of convergence ( with respect to sampling depth ) and accuracy ( with respect to diversity estimates ) . This suggests that power-law based estimators offer a valid alternative to classical diversity estimators and may have broad applicability in the field of population genetics .
Allele and Haplotype diversity are fundamental properties in the domain of population genetics for describing the general characteristics of any population of diploid organisms . In the context of hematopoietic stem cell transplant ( HSCT ) matching , proper estimates of allele and haplotype diversity are essential for describing the population genetics that govern the clinical suitability of a donor/patient match . Matching algorithms scan large unrelated donor registries to identify suitable matches based upon individual Human Leukocyte Antigen ( HLA ) genetic typings . HLA is defined by the super locus of genes contained in the major histocompatibility complex ( MHC ) on the 6th chromosome that encodes proteins governing the adaptive immune system response [1] . Successful transplantation requires careful matching of donor/recipient HLA to avoid adverse immune reactions that result in graft rejection or graft versus host disease [2 , 3] . The ambiguous nature of HLA typing , however , presents challenges for transplant matching , given that donor registries contain a range of typing methods and allele definitions that have evolved since the 1980’s . This “mixed resolution” data contains uncertainties regarding the exact alleles each donor has and their phase on the chromosome . The standard approach for resolving these uncertainties is to impute each donor’s HLA typing to a common allele resolution and set chromosomal phase using expectation maximization ( EM ) algorithms [4] . Given a proper candidate set of haplotypes , EM algorithms work well to estimate the distribution of this defined population , which becomes the reference data for computing accurate donor/patient match predictions . However , we often do not know what constitutes a complete or reasonable candidate set of haplotypes describing the clinically relevant variation among individuals . This problem arises due to the following main reasons: Lack of Adequate Sampling Depth—the MHC region has the highest allele diversity within the human genome , so even a large sample relative to the population is unlikely to observe all clinically relevant haplotypes . For example , in the European American population 44 . 65% of haplotypes in the Be The Match registry are singletons ( i . e . are represented by only one individual ) , yet the most frequent haplotype represents 6 . 11% of the sample so one might falsely conclude that variation is lower than it really is . This motivates the need to develop reliable methods for assessing sampling depth and the benefits of continued sampling; Ambiguous Nature of HLA Typing—true haplotype variation is confounded by ambiguous typing methods , which describe each individual as a set of possible alleles at each loci ( consistent with HLA typing ) . When candidate haplotypes are constructed from cross-combinations of these possible alleles , an intractably large set results due to the “curse of dimensionality”[5] . Currently , this exhaustive set of haplotypes is heavily trimmed prior to estimation with the EM algorithm , but the trimming strategies are not quantitatively informed; Variation in Diversity Among Populations—ancestral migration patterns have resulted in different patterns of allele and haplotype diversity among ethnic groups , implying that sampling depth requirements are likely to vary by population . Clearly , there is a need to develop quantitative estimates of allele and haplotype diversity . More formally , let A denote the number of unique HLA alleles in a population and H the number of unique HLA haplotypes ( see Table 1 for all symbols used ) . Estimating A and H is a difficult problem given that we do not directly observe all alleles and haplotypes in the population . Instead , we must establish and extrapolate a relationship between the number of kinds we observe and the frequency of each kind . Mathematical frameworks for describing this relationship have been developed in the study of species richness and occupancy . Capture-recapture methods have been adapted from ecology to genetics for estimating the total number of different alleles/haplotypes [6] . Such methods could be applied to estimate A and H by tagging resamples of alleles and haplotypes . Although interesting , these capture recapture methods require a very large sampling depth for heavy tailed distributions , as will be shown . Occupancy distributions [7]offer a more natural representation of our data sampling process and use the mathematical concept of regular variation to extrapolate A and H using a power law function that describes the distribution of allele and haplotype frequencies in the population [7 , 8] . The power law relationship was constructed under the assumption that an infinite number of categories ( or kinds ) exist in the population; so for our purposes we apply a boundary constraint to estimate a finite number of categories under this general framework . In the medical field of HSCT , estimates of A are largely catalog based; unique HLA genomic sequences are stored in databases as alleles using locus-specific nomenclature ( the best known being the IMGT/HLA database cataloging 10 , 653 allele definitions at the time of writing [9] ) . Further information on the prevalence of specific HLA alleles in the general population is provided by the Common and Well Documented ( CWD ) allele list [10] . Estimates of H are typically based on EM derived HLA haplotype frequencies , which represent nomenclature-based allele combinations that are likely to be found in the population . These estimates take into account that genetic information is inherited in blocks making the true value of H much smaller than the expansion of all possible allele combinations . For example , estimates of H within the United States population in 2013 identified over 160 , 000 unique 6-locus haplotypes in the Be The Match Registry , where 99% of the time H was comprised of combinations of “common” alleles and roughly 90% of the known IMGT documented alleles ( A ) were absent from the haplotype frequencies [11] . Thus , current methods for estimating of A and H rely upon direct observation of a given allele or haplotype . For estimating the “common” portions of A and H , this is a reliable strategy for sufficient observational data is likely to exist . However , such an approach is unsuitable for estimating the “rare” portion of A and H lacking direct observational data . Evidence for the existence of a large number of “rare” alleles and haplotypes in the overall population is outlined by Klitz [12] , who modeled the effects of gene conversion , point mutations , and recombination to establish the heavy tailed nature of allele and haplotype distributions . To estimate A and H , we , therefore , construct a heavy tail relationship based upon occupancy distributions and the properties of regular variation . This power-law framework is useful for modeling the probability mass ( and number ) of A and H that are unseen in the sample and represented by the “invisible” tail of the distribution . We perform this on 6-locus HLA data ( A~C~B~DRBX~DRB1~DQB1; where DRBX = DRB3/4/5 ) from the Be The Match Registry , which is one of the largest data sets characterizing human HLA population genetics containing the typing information for 6 . 59 million donors [11] . We also assess registry coverage , which is the proportion of the total population occupied by the observed haplotypes , of A and H with respect to the potential base of patients seeking HSCT to inform donor recruitment strategies . Last , we discuss broader applications of the power-law methodology outside HSCT for modeling species richness in the field of ecology , which is characterized by similar heavy-tailed distributions .
In order to estimate the total number of unique haplotypes H having probability pj for haplotype j 1≤j≤H , 0≤pj≤1 , we define a counting function ( vH ( x ) ) , also denoted an exceedance function , representing the number of haplotypes with a relative frequency ( i . e . pj ) above or equal to x . Observations suggest that vH ( x ) can be approximated by a scale free function: vH ( x ) ∝x−β ( 1 ) Note that in reality , vH ( x ) can only take discrete values . However , we here take a continuous approximation of . VH ( x ) This approximation stops being valid at some maximal and minimal haplotype relative frequencies ( Xmin and Xmax ) Obviously , a ( unique ) haplotype cannot have a frequency of less than 1 and , thus , a relative frequency of less than1 / Ntotal ( Xmin ≥ 1 / Ntotal ) , where Ntotal is the total number of haplotypes in the population ( twice the total population size since each sampled person has two sets of haplotypes ) . Ntotal is typically much larger than the number of unique haplotypes , H . Similarly , above a given relative frequency , vH ( x ) becomes smaller than 1 , and the approximation stops being valid . The total number of unique haplotypes in this formalism is H = VH ( Xmin ) . Given the continuous formalism , one can also define a haplotype count density ( which is not a formal density function , since it is not normalized to 1 ) , representing the number of unique haplotypes H in a given relative frequency interval of dx: μH ( x ) dx=vH ( x ) −vH ( x+dx ) ( 2 ) Based on the definition of vH ( x ) , μH ( x ) obeys: ∫xminxmaxμH ( x ) dx=H ( 3 ) Moreover , if vH ( x ) is a scale free function , so is μH ( x ) : μH ( x ) dx∝x−αdx , where α=β+1 ( 4 ) μH ( x ) is not a probability density function ( PDF ) . However it can be normalized , by defining: pH ( x ) =μH ( x ) H ( 5 ) which is a proper density distribution function . Thus , pH ( x ) provides a description of how common ( i . e . the density ) each pj is in the population . Further , a proper cumulative distribution function ( CDF ) for the pj’s ( the relative haplotype frequencies ) can be derived by normalizing VH ( x ) where Pr ( pj≤x ) =1−vH ( x ) /H , since vH ( x ) /H represents the complementary CDF ( i . e . 1-Pr ( pj≤x ) ) . Also , as noted above , the power law exponent of μH ( x ) is 1 unit greater than the power law exponent of VH ( x ) since these two quantities share the properties linking CDFs to PDFs , namely that a PDF is proportional to the first derivative of its CDF . Previous models of haplotype frequency using slowly varying distributions assumed unlimited population size and the existence of an infinite number of unique haplotypes H [7 , 8 , 13 , 14] . In order to estimate the properties of the real population haplotype distribution , a parallel theory is required for finite populations . The finite population affects the upper and lower bounds of the distribution . The lower bound is determined by the absence of relative frequencies smaller than 1 / Ntotal and is common to all types of distributions . The upper bound is specific to heavy tail distributions . In such distributions , the contribution of the right tail to the overall probability mass is considerable . As such , the normalization constant of the distribution is greatly affected by the maximal relative frequency in the population ( Xmax ) . We here expand these methods to finite sized populations ( and not only finite samples of an infinite population ) . In order to incorporate the finite size of the population , minimal and maximal relative haplotype frequencies are defined ( Xmin and Xmax ) . As mentioned , the minimal frequency for a haplotype cannot be less than one person in the whole Ntotal population that carries one copy of this haplotype ( Xmin≥Ntotal-1 ) . However , in many realistic cases , the scale free distribution only starts at a higher value than that . The upper cutoff ( Xmax ) can be determined by the total population Ntotalin the following way: Let us separate the values of x into bins . As mentioned , the minimum difference between relative frequencies is Ntotal-1 . We can thus define Ntotal-1to be the minimal bin size . Per definition , at the largest ‘x’ value , there is typically one unique haplotype . Thus , vH ( xmax ) = 1 , leading to: μH ( Xmax ) dx=1⇒μH ( Xmax ) 1Ntotal≤1⇒pH ( Xmax ) ≤NtotalH ( 6 ) Assuming Xmax is an upper threshold , we have the equal sign in the equation and in this case Xmax can then be numerically extracted from Eq 6 . In many realistic cases , the scale free distribution has an exponential cutoff , starting at much lower values and Xmax can be lower ( see S2 Text ) . The observed portion of H has been growing over the years , yet there is still a significant portion that remains unobserved . It is thus of interest to estimate the total size of H , and the relation between the sampling size and the portion of H that we actually observe . If the haplotype relative frequency distribution pH ( x ) was light tailed , most haplotypes would have the same order of frequency . In such a case , it would have been enough to sample a limited part of the population to detect the vast majority of different haplotypes and estimate H . More specifically , if the sample would be at least an order larger than H , most unique haplotypes would be detected . However , in heavy tailed distributions ( such as power-law distributions ) the vast majority of haplotypes are very rare . A deep sampling is thus required for a precise estimate of H and a very deep sample is required to observe all unique haplotypes . On the other hand , the contribution of these rare haplotypes to the coverage is often negligible since most of the population has common haplotypes . The precise balance between these two phenomena is determined by the power exponent of the distribution ( α ) . Formally , the expected number of unique haplotypes ( U ( R ) ) in a sample size ( R ) can be estimated , assuming a truncated power law formalism as defined above , to be ( see S1–S3 Texts ) : U ( R ) =H−2−αXmax2−α−Xmin2−αRα−1[γ ( 1−α , RXmax ) −γ ( 1−α , RXmin ) ] , ( 7 ) where γ is the partial gamma function . The total number of unique haplotypes can be computed to be: H = 2−α1−α Xmax1−α−Xmin1−αXmax2−α−Xmin2−α ( 8 ) The estimates above require the slope of the distribution ( α ) . Many methods have been proposed to estimate this slope , including a fit of the CDF or the PDF , or a fit of the power of the Zipf plot [15] . However , such methods have been argued to be flawed , as discussed in an excellent review by Clauset et al . [16] . The same authors have proposed estimates for the slope of the scale free distribution for either continuous [17] ( Eq 9 ) or discrete ( Eq 10 ) distributions: α=1+n[∑j=1nlnpjpmin]−1 ( 9 ) α=1+n[∑j=ln1nyjymin−0 . 5]−1 , ( 10 ) where n is the total number of observed haplotypes , pj's are the haplotype relative frequencies , pmin is the relative frequency of the rarest observed haplotype , y's are the parallel values in absolute frequencies for the discrete case and ymin is the minimal absolute frequency . Another estimator based on the asymptotic properties of regular variation was proposed by Ohannessian et al . for α∈[1 , 2] [7 , 18]:α=Kn , 1Kn+1 . This estimator is based on the ratio between Kn , 1—the number of haplotypes which appear only once in the sample , and Kn—the total number of unique haplotypes in the sample . Note that this estimator was derived for the CDF and β = α-1 , which is 1 less than our PDF based estimator . While these methods are precise for a full power law distribution , they do not converge to the proper value when the distribution is truncated either from above or from below ( Fig 1A ) . We also checked the convergence of Eq 10 when using minimal absolute frequency y*which is larger or equal to yminmentioned above:α=1+n[∑j=Jminnlnyjy*−0 . 5]−1 . This truncated power law estimate does converge to the right value , but very slowly ( Fig 1A orange circles ) . Formally , we estimated the fit to the distribution using a Kolmogorov Smirnov test , and used the lower cutoff Jmin producing the best fit . We have thus adopted a different approach based on a numerical fit of the observed value of U ( R ) to Eq 7–8 for different values of R . We compute U ( R ) from the observed data for different R values and find the parameters leading to a best fit between Eq 7–8 and the observed data . The parameters are bounded by Eq 6 for Xmax and Ntotal-1 for Xmin . The numerical minimization is initialized with Xmax from the observed data , Xmin at its boundary value , and α as it is computed from Eq 9 ( for a schematic figure see Fig 2 ) . The value of α computed using this estimate converges to the proper value at limited sampling depth ( Fig 1A ) . To validate the correctness of the haplotype number estimate in Eq 7 and 8 , we did multiple validations as follows: As a first validation , we chose values of Xmax , Xmin and α , and then calculated H using Eq 8 . We generated H haplotypes with relative frequencies ( pj ’ s ) taken from a pure power law distribution with exponent of α . Next , we generated samples of different sizes ( R ) based on the pj ’ s values . We compared the expected value of U ( R ) ( Eq 7 ) to the actual number of different unique haplotypes that we got in the sample , with an excellent fit ( less than 1% deviation—S1 Fig ) . A similar analysis with a truncated power law gave similar results . We then compared the convergence rate of our method with other frameworks for estimating H . We used a simulated population from a truncated power law , as described above and formed sub samples of incremental size . We then generated estimates of H as a function of increasing sub-sample size . We compared our method with a capture-recapture formula [6] . This method was adapted to compute the total number of different alleles instead of the total population as is performed in ecology . In order to compute the capture recapture estimate , we divided each sub-sample into two smaller sub samples: A and B . We then computed the ratio between the product of the number of different alleles/haplotypes in A and B and the number of different alleles/haplotypes in their intersection:|Unique ( A ) |*|unique ( B ) |/|Unique ( A∪B ) | . We further compared our method with a Jackknife based estimator [19] . Capture-recapture and power-law estimates were found to converge to the true value of H , but the capture-recapture method required a very deep sample of the population to attain accuracy whereas the power-law method converged quickly and offered accurate estimates , even with limited sampling ( Fig 1B ) . The Jackknife based estimator did not converge to the proper value . The field of species richness estimation is very wide and a large number of methods to estimate species richness have been proposed ( e . g . [20] and supplementary material therein ) . Many of these methods have been merged into the CatchAll software [21] . To further compare our method , with existing state of the art methods , we have tested these measures on the same samples . All methods proposed by CatchAll showed a very slow convergence , or actually no convergence ( Fig 3 ) . In order to further validate the methodology in a realistic regime , we used 7 . 8 million samples from Be the Match Registry , which identified 88 , 621 haplotypes as estimated in the EM algorithm from the European American population , and used a sub-sample of half a million haplotypes to estimate the parameters of the distribution ( Xmax , Xmin and α ) . We then validated that the resulting values fit the observed distribution all the way to the full sample size ( 7 . 8e6 ) and extrapolated it to the total European American population as defined by the Census [22] . The final value is an estimate of the total number of haplotypes in the European American population in the US ( Fig 4; see S2 Fig for the other populations ) . When we estimated Xmin , Xmax and α from the distribution ( purple full line ) an excellent fit was obtained . To investigate the sensitivity of the model to parameter estimates of Xmin , Xmax and α , we fixed the value of α based on the Eq 9 , and optimized Xmin and xmax or fixed one of these variables at their extreme values and optimized α . The fit diverges from the observed result ( Fig 4 and S2 Fig ) . The strongest deviation is for a fixed α , and the smallest is for a fixed Xmin . A fixed Xmin is expected to have the smallest effect , since its estimate is closest to the reality in the studied distributions . To summarize , the method above has been shown to converge properly in simulated and sampled data . However , for accurate estimates , it is recommended to treat all parameters Xmin , Xmax and α as free parameters to be estimated in the model . The resulting total number of haplotypes ( H ) in the European American population in the US is 286 , 787 . Values for other sub-populations are given in Table 2 . For most populations , the haplotype frequency distribution follows a power law with exponent α of the PDF between 1 . 4 and 1 . 9 . Further , estimates of the power law exponent converged before the full sample size was analyzed ( Table 3 and S3–S8 Figs ) . An important aspect of Eq 7 is that it does not saturate until the sample size is close to the full population size . We therefore expect that haplotypes discovery rate should remain appreciable with continued sampling given that the current sample size is about 3 . 97% of the European American populations . In contrast , if the distribution were light-tailed , we would be expected to detect all of the haplotypes that exist . Since the distribution is heavy tailed , we estimate to have detected only 30% of H that exist in the US population ( Table 2 ) . An important aspect of the low value of α ( α ≤ 2 in most cases ) is that there is a large difference between the fraction of observed haplotypes and the population coverage , defined as the fraction of samples ( two per person ) with a known haplotype . When α is close to 1 , most of the haplotypes have practically no contribution to the coverage . Actually only very few haplotypes in the right hand tail of vH ( x ) ( the larger values of x ) contribute to the population coverage . For example , in the European American population , the most frequent 1% of haplotypes provide 73 . 18% of the coverage . The fraction of the population with haplotypes which have already been identified can be computed to be ( see S4 Text ) : FractionCovered ( R ) =1−2−αXmax2−α−Xmin2−αRα−2[γ ( 2−α , RXmax ) −γ ( 2−α , RXmin ) ] , ( 11 ) where γ is the partial gamma function . As was the case for Eq 7 , we compared the values obtained from this calculation to the values obtained in the simulation mentioned above , and obtained an excellent fit ( S9 Fig ) . One can apply Eq 8 to the European American population , for example , and obtain that only around 88 , 620 haplotypes are identified today , out of over 286 , 787 haplotypes that are estimated to exist in the total population ( Fig 5A and Table 2 ) . However , the coverage of these 88 , 620 haplotypes is about 99 . 4% ( Fig 5C and Table 2 ) . The relationship between the coverage and the number of haplotypes depends on α , and varies among the studied populations . We analyzed 5 combined populations , for which clear census based estimates of the total population size are available ( African-Americans—AFA , Asian Pacific Islanders—API , European Americans—CAU , Hispanic—HIS and Native Americans—NAM ) , and calculated the sample size needed in order to have the same percent coverage of the European American population ( Fig 5D and Table 2 ) . The combined populations have similar distributions ( Fig 5A ) , with the European American population having the flattest distribution ( least diverse , or most contribution from frequent haplotypes ) , and the Native American having the steepest slope ( most diverse , and the least contribution from frequent haplotypes ) . However , since as mentioned above the number of haplotypes ( H ) is affected by the total population size , the Native American population is actually the least diverse , while the European American population is the most diverse , when H is taken as a measure of the diversity ( Fig 5B ) . While most haplotypes are not sampled , most individuals in the population carry known haplotypes . As mentioned earlier , 99 . 4% of the European American population is covered by observed haplotypes ( Fig 5C ) . Coverage is also more than 98% for the Hispanic , African-American and Asian/Pacific Islander populations . Coverage is less for the Native American population at 93 . 4% percent . In order to reach the percent coverage that currently exists for the European American population , 1 to 6 million additional donors would need to be recruited for each of the other populations ( Fig 5D ) . This represents a factor between the required and existing sample size of around 3 for the Asian Pacific Islanders , but up to 15 for Native Americans , as summarized in Table 2 . The contrast between the small fraction of haplotypes known and the large coverage implies that the frequency of each undiscovered haplotype is expected to be very small . The probability that a haplotype from a newly sampled host is not present in the current sample of size R is ( see S4 Text ) : Z ( R ) =2−αXmax2−α−Xmin2−αRα−2[γ ( 2−α , RXmax ) −γ ( 2−α , RXmin ) ] ( 12 ) where γ is the partial gamma function . We validated the accuracy of our analytical equation using simulation ( S10 Fig ) . This function decreases slowly as a function of sample size R , which explains the high degree of population coverage achieved even with limited discovery of all haplotypes in the population . Additionally , this value could be used to assign an expected frequency to missing haplotypes , which in turn could be used in HSCT matching to assign haplotype frequencies to newly discovered haplotypes not covered by the EM haplotype frequencies . A similar analysis can be performed on the observed portion of A . Each haplotype is composed of alleles at 6 genetic loci: ‘A’ , ’B’ , ’C’ , ’DQB1’ , ’DRB1’ , ’DRBX’ . We separated the haplotype data and generated allele distributions for each of the 6 loci . Our assumptions about the alleles are just the same as for the haplotypes , and a similar analysis was performed on each loci . We fit the expected number of alleles U ( R ) as a function of the sample size R to the actual number of unique alleles in partial samples of the data we have for alleles ( see Methods ) to obtain the values of Xmin , Xmax and α . The allele distribution by loci is very flat with powers approaching 1 ( Table 4 ) . Such powers give most of the importance to the frequent alleles and reduce the total number of alleles for a given population size . Indeed , estimates of A according to the current sample for the data of the five combined populations are not much higher than the observed number of alleles ( Fig 6 left plot ) . Consequentially , the fraction of uncovered population is much smaller for the alleles than for the haplotypes reaching 0 . 1% in the worst case ( Native Americans for HLA—B which is the most diverse locus ( Fig 6 right plot ) ) . As expected by their reported diversity [9] , Class I HLA alleles have a slightly higher power exponent ( 1 . 2–1 . 3 ) than class II ( around 1 ) ( Table 4 ) . The class II power exponents , which are all close 1 , are expected from neutral evolution model [23] , highlighting that if selection occurs , it is mainly focused on class I or it is occurring at the haplotype level . Surprisingly , the discovery rate for new alleles in the IMGT database appears to be increasing , in contrast with the conclusion that most alleles are known , raising the suspicion that the total number of existing alleles is much larger than current estimates . However , the estimates above propose that most alleles are currently known . To better understand this contradiction , we computed allele sample sizes during the years 1987–2011 and calculated the expected number of alleles based on these sample sizes , using the α , Xmin and Xmax values obtained for the largest sample . We plotted the expected number of alleles as a function of time and compared it to the number of alleles which were identified by that date in the IMGT ( Fig 7 ) . In analyzing the entire US population as a whole , a comparison between the expected number of alleles U ( R ) and those actually observed presents a contradiction: Initially , the number of observed alleles was significantly less than expected based upon sample size; presently , it is much larger than expected . The explanation for this contradiction appears to be two-fold: First , false allele discovery represents a known source of error common with PCR when crossover products occur [22] and could contribute to the observed discrepancy . Second and more importantly , the transition to sequencing-based typing ( SBT ) changed the fundamental nature of allele definitions and improved capacity for allele discovery . With a much larger definition of alleles and a higher discovery rate , the fundamental power law relationship would be expected to change at some point in time as the data sources move towards SBT versus older oligo or serology based methods . For example , a recent data submission from a single SBT laboratory resulted in a 30% increase in the number of observed alleles [24] . Thus , our current power-law methodology appears flawed for providing accurate estimates for the number of unique alleles and requires future modifications to accommodate the mixed data sources .
In this analysis , we developed a quantitative methodology for measuring haplotype and allele diversity with a particular application in mind of estimating A and H in the United States population to inform HSCT matching . Quantitative estimates of A and H are important for assessing whether typed donor cohorts have sufficient sampling depth to capture the clinically relevant HLA variation among individuals . Estimates of A and H for the Be The Match Registry suggest that current registry sampling depth is sufficient to represent only 52–100% of A and 21–30% of H across the 5 main racial populations assessed . The number of “rare” alleles and haplotypes , however , is small compared to the number of “common” ones suggesting that the current diversity of our registry is sufficient to describe 99 . 94–100% and 93 . 4–99 . 4% of the overall United States population with respect to allele and haplotype coverage . Further , the estimation methods developed in this article quickly converged—with limited sampling depth—to accurate estimates of allele and haplotype diversity in the presence of heavy-tailed distributions . Popular methods intended for measuring diversity with bell shaped distributions were shown to require deep samples before achieving meaningful convergence , as was the case for capture-recapture models , or failed outright to converge toward accurate estimates . A detailed comparison of our methodology with state of the art species richness estimates , as performed by the CatchAll software , show a much faster and more precise convergence to the real haplotype number using our methodology . The methodology we developed is closely aligned with the broader fields of ecology , computation , and statistics as they relate to describing species richness , language diversity , data base storage , and occupancy . The common thread among these problems is the need to understand the behavior of “rare” events in a population with a heavy tail distribution: this can be problematic because “rare” events represent critical patterns that have minimal , if any , representation within the sample . Appropriate statistical methodologies for “rare” events recognize that light tailed distributions ( often having the characteristic bell shape ) are inappropriate for adequately describing population diversity . The assumption of a light tailed distribution implies that a relatively small sample is enough to detect and fully describe the population ( i . e . captures all species or haplotype diversity ) . This intuition stems from the absence of extremely rare events , which is an untrue assumption for a heavy tail distribution . For example , in most realistic species distributions , the majority of species are very rare , and thus a very deep sample is required to detect most species . Moreover , if the tail is flat enough , another element should be incorporated , which is the extreme contribution of the most frequent species to the distribution . In such cases , most of the population will belong to a very small number of species . Therefore , we built our model around a truncated power-law for estimating the properties of infinite discrete distributions with regularly varying heavy tails [7 , 8 , 14] . The properties of our data fit these assumptions , which require an index of variation α ∈ ( 1 , 2 ) for the PDF of frequencies: our index of variation ranged from 1 . 45 to 1 . 85 across the 21 populations . We note that to support the interpretability of our results , we modeled the PDF of frequencies whereas others have modeled the CDF of frequencies where values of α will be α - 1 units smaller . It is important to note that the convergence to the proper values of α occurred for a limited sample size . Thus , our method may be applicable for other surveys where the sample size is much smaller . Moreover , beyond the need to estimate the total number of different haplotype , the proposed method can be used to properly estimate α for small population sizes . In the infinite case , one ignores the right hand side behavior because the “common” events offer a negligible contribution to overall diversity . Eventually , one might consider a more complex model where we define vx = Cx-αL ( x ) , with L ( X ) being a general function with the property that v ( x ) = Cx-αL ( x ) . Such a function would account more smoothly for the different slopes in the left hand part and the right hand part of the distribution . Note however , that in realistic situations , the most frequent haplotypes ( or species ) are well known , and their frequency is well estimated . Thus , a simpler approach could be to split the population into frequency known haplotypes and rare haplotypes , and focus on the distribution of the rare haplotypes , taking into account that their contribution to the population coverage may be very limited , even if they constitute the vast majority of the diversity . While we have here focused on the HLA haplotype distribution , the same formalism can be applied to other domains of ecology and population genetics , where the species richness distributions have heavy tails [25 , 26 , 27 , 28] , see [29] for a review . In ecological scenarios , since most species are very rare , the current shrinking of the wildlife population is equivalent to a dramatic reduction in the total diversity . While in normal distributions we expect species to disappear and diversity to be significantly reduced only when the total population is close to extinction , the heavy tailed distribution of species frequencies leads to opposite conclusions . For example , assuming parameters similar to the ones obtained here , a 10 fold reduction in the total population would be equivalent to a 17–18% reduction in the number of species . Note that in ecological scenarios , it is not appropriate to assume that all species are equally detectable at the level of the individual . Certain species are less detectable based on experimental or behavioral elements ( e . g . speed of movement ) . A future development of such methods should be the introduction of a frequency bias in the detection probability . Regarding validation , we compared our estimates of A and H—based upon subsamples of the data—to counts obtained by enumerating the complete data set . Although our predictions aligned reasonably with the absolute frequencies , we are concerned with the fact that the underlying data may be biased towards over-representing common alleles and haplotypes; thus , our estimates of A and H may under-represent the true population diversity . Our concern is based upon the mechanics of the EM algorithm , which trim the tail of the haplotype frequency distribution numerous times during the estimation process to generate a parsimonious and tractable candidate set; without trimming the number of potential haplotypes that greatly exceeds the number of donors . Further , expected changes in typing technology ( that are already underway ) from SBT to next generation sequencing ( NGS ) may also re-define allele definitions and increase the capacity for allele discovery . For example , NGS has the ability to detect allelic variants previously “hidden” behind genotypes that only differ by recombination [30] . In short , EM and typing technology factors could increase estimates for A and H by changing the fundamental nature of the power law relationship .
Six-locus high resolution HLA A~C~B~DRBX~DRB1~DQB1 ( where DRBX = DRB3/4/5 ) haplotype frequencies were estimated across 21 race groups using an EM algorithm and 6 . 59 million donor HLA typings from the Be The Match Registry: complete details regarding the data and estimation are provided by Gragert [11] . In brief , an EM algorithm was utilized to resolve uncertainties in allele and chromosomal phase for a mixed resolution set of donor HLA typings ( serology , sequence-specific oligonucleotide/primer , and sequence based typing ) . The only notable change from the Gragert methodology was that in the last iteration of the EM algorithm , a winner-take-all approach was applied where each donor contributed 1 unit of probability mass to their most likely pair of haplotypes; this is opposed to each donor assigning 1 unit of probability mass across a range of haplotype pairs—consistent with their HLA typing—in proportion to their conditional likelihood . Allele frequencies were derived as marginal sums of the haplotype frequencies . We have performed simulations with chosen α and Ntotal . For each set of α and Ntotal values , we set Xmin to be 1Ntotal , and calculated the maximum value possible for Xmax using numerical solution of x2 ( 1-xα-2 ) -C = 0 where C = ( 2-α ) Ntotal ( see Eq B6 in S2 Text ) . We then calculated the number of unique haplotypes expected in the total population ( H ) . Next , we generated H pj’s for the distribution ( see S1 Text ) by inverting the CDF , obtaining: pj= ( Xmax1−α−Xmin1−α ) rand ( 0 , 1 ) +[Xmin1−α]11−α where rand ( 0 , 1 ) is a uniform random number between 0 and 1 . We checked that the sum of all pj’s indeed equals 1 , and very limited deviations were observed . We normalized the results to overcome the imprecise sum . We then produced a random population with these a priori probabilities , and compared the expected frequency of a haplotype unobserved in a sample ( Z ( R ) ) , the expected number of unique haplotypes in the population ( U ( R ) ) and the fraction of the population covered by these haplotypes to the simulations results . Code for the estimation of U ( R ) and a test set are available in the Supp . Mat . Given an observed population's haplotype frequency distribution , we assume that the frequency distribution is a scale free distribution with upper and lower cutoff values: Xmin≤x≤Xmax . We assume a zero probability of observing haplotypes with frequencies beyond these values . In order to estimate the properties of the distribution , the values of α , Xmin and Xmax should be estimated . We provide boundaries for Xmin and Xmax using Eq B1 and B6 in S2 Text . We then perform a numerical fit of the observed and expected ( Eq 8 ) unique haplotypes , with a cost function of ∑R' ( log ( U ( R' ) -log ( observed ( R' ) ) 2*log ( observed ( R' ) ) for different values of sample sizes R’ . We use an initial guess of the parameter above using Eq B1 for Xmin , the highest observed frequency X0max , and the estimate of α , based on Eq 9 . We then extract the optimal parameters ( α , Xmax , Xmax ) , calculate the expected total number of unique haplotypes H and produce an estimate of the number of unique haplotypes for any population size U ( R ) .
|
The distribution of haplotypes and species tend to be heavy tailed . The heavy tail is expected from theoretical considerations and is observed in most populations . Accurate measures of diversity are difficult to achieve given that a limited number of common haplotypes represent the majority of the population , whereas the major contributor to haplotype diversity comes from unique haplotypes that are “rare” and present in only a fraction of the population . A major issue for unrelated HSCT donor registries is estimating population coverage with respect to servicing the public need . We here use a power-law methodology that accommodates heavy-tails to estimate both the population coverage by ethnicity in the US and the genetic diversity of alleles and haplotypes . For the European American population , which has the deepest sampling amongst ethnicities , we show that registry population coverage is better than 99% , but the diversity of this sample only represents 40% of the unique haplotypes expected to be found in the population . Population coverage for other ethnicities was poorer and ranged down to 92% as was the case for Native Americans that had the worst coverage . We further show that the formalism developed here produces better estimates of the population properties than existing methods .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Power Laws for Heavy-Tailed Distributions: Modeling Allele and Haplotype Diversity for the National Marrow Donor Program
|
Mycobacterium ulcerans disease ( Buruli ulcer ) is the most widespread mycobacterial disease in the world after leprosy and tuberculosis . How M . ulcerans is introduced into the skin of humans remains unclear , but it appears that individuals living in the same environment may have different susceptibilities . This study aims to determine whether frequent contacts with natural water sources , family relationship or the practice of consanguineous marriages are associated with the occurrence of Buruli ulcer ( BU ) . Case control study . Department of Atlantique , Benin . BU-confirmed cases that were diagnosed and followed up at the BU detection and treatment center ( CDTUB ) of Allada ( Department of the Atlantique , Benin ) during the period from January 1st , 2006 , to June 30th , 2008 , with three matched controls ( persons who had no signs or symptoms of active or inactive BU ) for age , gender and village of residence per case . Contact with natural water sources , BU history in the family and the practice of consanguineous marriages . A total of 416 participants were included in this study , including 104 cases and 312 controls . BU history in the family ( p<0 . 001 ) , adjusted by daily contact with a natural water source ( p = 0 . 007 ) , was significantly associated with higher odds of having BU ( OR; 95% CI = 5 . 5; 3 . 0–10 . 0 ) . The practice of consanguineous marriage was not associated with the occurrence of BU ( p = 0 . 40 ) . Mendelian disorders could explain this finding , which may influence individual susceptibility by impairing immunity . This study suggests that a combination of genetic factors and behavioral risk factors may increase the susceptibility for developing BU .
Mycobacterium ulcerans disease , commonly named Buruli Ulcer ( BU ) , is the most common mycobacterial disease in the world after leprosy and tuberculosis . This emerging disease has been reported in more than 30 countries in Africa , Latin America , Oceania and Asia [1] causing immense human suffering [2] , [3] , while its prevalence in most endemic countries is uncertain [4] . How exactly M . ulcerans is introduced into the skin of humans remains unknown , but in contrast to tuberculosis ( TB ) or leprosy , the infection is acquired directly or indirectly from the environment and not through contact with other patients [5] . There is now evidence that M . ulcerans is an environmental pathogen transmitted to humans from aquatic sources , and the first isolation and characterization of an M . ulcerans strain from an aquatic Hemiptera ( water striders , Gerris sp . ) from Benin was reported by Portaels et al . [6] . The individuals living in the same environment appear to have different susceptibilities to the disease . Indeed , most individuals exposed to M . ulcerans never develop BU disease [7] . The reason why some individuals , but not others , exposed to M . ulcerans develop BU is unknown but it could be linked to individual differences in innate and acquired immune responses to infection by this bacterium . Furthermore , the susceptibility to the development of BU may be determined by genetic factors as well . Several studies have been conducted to determine the genetic and/or immunological factors affecting BU disease [8]–[10] , but no studies have examined family relationship as a factor of presumptive susceptibility to BU . The aim of our study was to determine whether the occurrence of BU is associated with family relationship or the practice of consanguineous marriage , in addition to daily contact with natural water sources .
A case control study was carried out during the period of January 1st to June 30th , 2008 . The patients included in this study as cases were diagnosed and followed up at the BU detection and treatment center ( CDTUB ) of Allada ( Department of the Atlantique , Benin ) or at various health care centers ( HCC ) involved in the treatment of BU under the supervision of the CDTUB of Allada . From January 2006 to June 2008 , the cases with active BU lesions ( nodule , edema , plaque , ulcer or osteomyelitis ) were recruited [11] and confirmed by at least one laboratory test ( direct smear examination showing acid-fast bacilli , positive culture or IS2404-PCR ) [12] . The individuals who were no longer hospitalized were located by using the addresses in their medical file . Eligible cases who had moved or were not found during the data collection were excluded . All eligible cases who were not living in the Atlantique department were also excluded . An eligible control was defined as a person who had no signs or symptoms of active or inactive BU . The eligible controls who were suffering or had suffered from any mycobacterial disease ( leprosy , TB or BU ) were excluded as well . Three controls , matched by age , gender and village of residence , were selected for each case . The controls were randomly sampled from within the village of the case according to the matching criteria . A door-to-door systematic procedure was used for control selection from the center of the village of each related case . We used power calculation tools to determine the sample size . We set alpha equal to 0 . 05 and power equal to 80% . We assumed a rate of three controls per case . Because we lacked data on the frequency of consanguineous marriages , we assumed a rate of 50% in controls . The minimum of the odds ratio ( OR ) for the association between cases and controls was set equal to two . We obtained a sample size of 396 participants , including 99 cases and 297 controls . A standard questionnaire was administered to eligible cases and matched controls ( or their guardians ) by trained investigators . Structured interviews were conducted with the participants during home visits using the pre-tested questionnaire translated in Aïzo and Fon ( the most commonly spoken languages in the region ) . Interviews with current in-patients were conducted in the hospital . The questionnaire was filled out by the interviewer . If required , subsequent interviews were conducted until all the required data were obtained . The participant's identification data ( age , gender , geographical origin ) , family history regarding any disease ( especially sickle cell disease , diabetes and arterial hypertension ) , marital status ( single or not ) and habits regarding daily contact ( contacts from professional or domestic activities and , in the case of children , from play activities ) with natural sources of water ( e . g . , river , lake , lagoon , swamp ) were collected . The data relating to the illness ( clinical form , site and categorization of the lesion based on the World Health Organization ( WHO ) definition [11] ) were also collected . The family history of BU in the participant was investigated and if present , the number of family members who had BU was recorded . For each family member who had BU , the data were collected on the degree of the relationship with the participant ( grandparents , parents , collaterals and descendants ) , the residence at the time he/she was ill ( same house , same village but not the same house or other village or town ) and whether or not he/she had daily contact with a natural water source during his/her daily activities . The data were collected on the existence or practice of consanguineous marriages . When found , the type of relationship between the married couple ( brother/sister , cousins , parents/children , uncle/niece or aunt/nephew ) and the degree of the relationship in the consanguineous married couple to the participant were determined . The pedigree of each participant was determined ( with the help of the parents or guardians ) using an in depth interview . The pedigrees went as far back or forward as the 3rd generation before or after the participant , when applicable , and included the collaterals . Every parent who had BU was carefully specified . The data were recorded and analyzed using epiinfo 3 . 5 . 1 ( Database and statistics software for public health professionals , Centers for Disease Control and Prevention ( CDC ) , Atlanta , USA ) . First , a descriptive univariate analysis was conducted on the characteristics of the participants , using Pearson's chi-square test or Fisher test . Second , the cases and the controls were compared using both univariate and multivariate analysis to determine the odds ratio ( OR ) and its 95% confidence interval ( 95% CI ) . To examine the association between family relationships and the occurrence of BU , all variables were included in a multiple conditional logistic regression model , followed by a step by step backward elimination based on the likelihood ratio in which only significant predictors were retained” . The participant status ( case or control ) was the dependent variable; all other variables were used as independent variables . Non Respondents have been excluded from the multiple conditional logistic regression analysis . The study enrollment was voluntary . A written informed consent was obtained from the cases and the controls or from their parents or guardians ( for patients younger than 15 years ) . All BU cases had received or were currently receiving free treatment for BU according to the WHO's recommended protocol [11] . The study protocol was authorized by the Ministry of Health of Benin .
There were 416 participants in the study , including 104 cases and 312 controls . Among the cases , the median age was 12 years ( 2 to 68 years ) . A total of 62 patients ( 59 . 6% ) were <15 years and 58 . 7% ( 61 out of 104 ) were male . A total of 75 ( 72 . 1% ) patients came from the Zê district , while 10 ( 9 . 6% ) came from Allada , 9 ( 8 . 7% ) from Toffo , 6 ( 5 . 8% ) from So-Ava , 3 ( 2 . 9% ) from Abomey-Calavi and 1 ( 1 . 0% ) from Tori-Bossito . The age , gender and geographical origin were similar for the cases and the controls because of the matching protocol used in control selection . There were no statistical differences between the cases and the controls with regard to marital status , hereditary disease history and daily contact with natural sources of water ( Table 1 ) . With respect to clinical form , 56 cases ( 53 . 8% ) had ulcerative lesions and 48 cases ( 46 . 2% ) had non-ulcerative lesions . Lesions were on the lower limbs for 55 cases ( 52 . 9% ) , the upper limbs for 34 ( 32 . 7% ) and on the trunk for 10 ( 9 . 6% ) . There were four cases with lesions on multiple sites ( 3 . 8% ) and one case ( 1 . 0% ) with a lesion on the face . The categorization of patient lesions based on the WHO definition [10] placed 10 ( 9 . 6% ) in category 1 , 59 ( 56 . 7% ) in category 2 and 35 ( 33 . 7% ) in category 3 ( data not shown ) . Table 2 shows the association between family relationships and the occurrence of BU based on the univariate and multivariate analyses . The univariate analysis showed that BU history in the family and daily contact with natural water source were strongly associated with an increased risk of BU ( OR; 95% CI = 5 . 07; 2 . 81–9 . 14 and 2 . 31; 1 . 18–4 . 53 respectively ) . A consanguineous marriage in the family was not associated with the occurrence of BU ( p = 0 . 33 ) . In the multivariate conditional logistic regression model including participant characteristics , two main factors were retained and associated with the occurrence of BU: ( 1 ) daily contact with natural source of water ( OR; 95% CI = 2 . 7; 1 . 3–5 . 5 ) ; ( 2 ) BU history in the family ( OR; 95% CI = 5 . 5; 3 . 0–10 . 0 ) . Table 3 shows the degree of the relationship between the cases and any other family member with BU . Associations between family relationships ( parents , siblings , cousins , sons , daughters , nieces and nephews ) and the occurrence of BU were not found . But , the BU cases were more likely to have grandparents with BU than the controls ( p = 0 . 06 ) . However , there was a lack of precision , the 95% CI being too large ( 95% CI = 0 . 85–64 . 08 ) . Seven grandparents had a history of BU , six from cases and one from a control . The grandfather was involved three times and the grandmother four times . From seven affected grandparents , two were currently living in the same house , two were living in the same village ( but not in the same house ) , two were living outside the village ( including the grandparent of the control ) and one was deceased . At the time of the disease , three of the involved grandparents were living in the same house as the patient , two in the same village and one outside the village . The place of residence at the time of the disease was not known for one grandparent . There was no statistical difference with regard to the living places of the grandparents involved between cases and controls ( p = 0 . 30 , Fisher test ) . All grandparents related to a BU case had contact with a natural source of water during their daily activities ( data not shown ) .
The objective of this study was to investigate whether or not daily contact with natural sources of water combined with family relationships was associated with an increased susceptibility to the development of BU . Our major finding was that the odds ratio of having BU was three times higher in the cases than the controls for those having a daily contact with natural sources of water , and five times higher for those who had a history of BU in their family . Many publications have reported a relationship between BU and neighborhoods in humid environments ( reviewed in [13]–[15] ) . Lunn et al . in 1965 [16] and also Barker in 1972 [17] described cases that were primarily from the Nile Valley and bordering marshes in Uganda . Ravisse in 1975 described cases in Cameroon originating from the Nyong River and surrounding swamps [18] . In 1976 , Oluwasanmi described cases in Nigeria that were located in an area near an artificial lake bordering the University of Ibadan [19] . In 2005 , Johnson et al . showed that there was an inverse relationship between the prevalence of BU and the distance from the Couffo River in Benin [20] . In 2008 , Kibadi et al . reported three patients originating from villages near the Cuango and Kwango River in Angola and the Democratic Republic of Congo , respectively [21] . Thus , it is clear from these studies that M . ulcerans disease occurs mainly in areas located near rivers , lakes or swamps . Nontuberculous mycobacteria ( NTM ) can be found everywhere in nature and at all latitudes . In BU endemic areas of Benin , M . ulcerans DNA has been detected in several aquatic organisms [22]–[24] . Humans and animals are regularly in contact with the environmental mycobacteria . Consequently , the colonization of healthy individuals by NTM is fairly common [7] . It is thus important to know why some individuals develop the disease while others do not . Our results have shown that daily contact with a natural water source is a risk factor for BU , which confirms previous studies that investigated behavioral factors associated with BU [25]–[30] . In addition , we have shown , for the first time , the existence of family associations with BU . Previous studies have incorporated BU history in the family among the factors tested but did not find a statistically significant association [25] , [29] , [30] . In contrast to the other studies , we included several generations ( e . g . , grandparents , parents , children ) in our study , which might explain the discrepancies . The observed association of BU history in the family and the occurrence of the disease provide new evidence with regard to the susceptibility to BU . This association may be due to the fact that members of the same family have perhaps same habits and same exposure to common environmental sources . However , we did not find any association between family members of the contemporaneous generations ( collaterals , parents or children ) . This leads us to suggest that genetic factors may increase susceptibility to the BU disease . Several studies have shown that susceptibility to other mycobacterial infections , such as TB and leprosy , involve a major genetic component that determines the susceptibility of animals and humans to these infections [31]–[38] . In these studies , the development of TB or leprosy upon exposure to the mycobacteria and the pattern of clinical manifestations displayed by patients ( pulmonary TB , paucibacillary or multibacillary leprosy ) were highly dependent on human genes [32]–[35] , [37] , [38] . Genetic susceptibility to the development of BU was demonstrated by Stienstra et al . [39] , [40] . A similar pattern had previously been found in Leprosy and TB [32] , [41] . Awomoyi et al . showed that SLC11A1 ( NRAMP1 ) influenced TB susceptibility by regulating immunosuppressive cytokines such as interleukine-10 ( IL-10 ) [41] , subsequently reducing the Th-1 immune response in the active disease [10] . Several studies have also demonstrated the genetic origin of cytokine deficiencies observed in mycobacterial infections [42] , [43] . In addition , it is known that subjects with past or current M . ulcerans infection mount a dominant Th-2 type response ( as also observed in advanced TB [10] , [44] , [45] ) following stimulation with M . ulcerans , while unaffected contacts responded mainly with a Th-1 type response [10] . Thus , the various clinical lesions ( nodule , plaque or edema ) , as well as the resistance to BU , may rely on host factors , such as the type of immune response , that depend on genetic factors . This suggests that patients who develop the clinical disease and those who develop a severe form of the disease appear to have an inherent inability to generate a strong Th-1 response to mycobacterial antigens [10] . Characterizing these primary immunodeficiencies helped to define the Mendelian susceptibility to mycobacterial infections syndrome ( MSMIS ) [36] , [42] , [46] , [47] . Further studies should explore which human genetic factors play a role in BU infection per se , and in the development of its different clinical forms . This pattern may be of great therapeutic importance . A gene is inherited in a dominant , recessive or co-dominant mode . Depending on the mode of transmission , the susceptibility to infection can be maintained from one generation to the next in a continuous , discontinuous or even random way . In our study there could be an association between the existence of BU in grandparents and the occurrence of the disease in the study cases ( since the p value was at the level of significance ( p = 0 . 06 ) ) , whereas there was no association for parents , collaterals or progeny . However , our study lacks precision ( 95% CI = 0 . 85–64 . 08 ) . Consanguineous marriage is a practice that could promote an imbalance in the transmission of certain genes and thus the development of anomalies . Asha Bai et al . showed that developmental anomalies were significantly more frequent ( p<0 . 001 ) among the progenies of consanguineous parents [48] . La Rosa , in 2008 [49] , proposed the hypothesis that ethnic endogamy could explain the focal distribution of BU as described in Benin [20] . Lyons et al . showed that consanguinity was an important risk factor in susceptibility to infectious diseases in humans [50] . In particular , they found that cases of TB and hepatitis were more common among inbred individuals , but only in populations where consanguineous marriages are common [50] . Our study does not show any statistical difference in the frequency of consanguineous marriages between the cases and the controls . The overall frequency of the consanguineous marriage practice in our study was only 10 . 3% . In the cohort of Asha Bai et al . in India , it was 41 . 4% [48] . However , there is no information on the overall incidence of consanguineous marriages in Benin , sub-Saharan Africa or our study area . Thus , this is an area where further research is needed .
This study confirmed the role of water contact as a risk factor and also suggests that the combination of genetic factors may constitute risk factors for the development of BU . Further studies should explore which human genetic or epigenetic factors play a role in BU infection and the development of the disease .
|
Mycobacterium ulcerans disease ( Buruli ulcer ) is the most widespread mycobacterial disease in the world after leprosy and tuberculosis . How M . ulcerans is introduced into the skin of humans remains unclear , but it appears that individuals living in the same environment may have different susceptibilities . This case control study aims to determine whether frequent contacts with natural water sources , family relationship or the practice of consanguineous marriages are associated with the occurrence of Buruli ulcer ( BU ) . The study involved 416 participants , of which 104 BU-confirmed cases and 312 age , gender and village of residence matched controls ( persons who had no signs or symptoms of active or inactive BU ) . The results confirmed that contact with natural water sources is a risk factor . Furthermore , it suggests that a combination of genetic factors may constitute risk factors for the development of BU , possibly by influencing the type of immune response in the individual , and , consequently , the development of BU infection per se and its different clinical forms . These findings may be of major therapeutic interest .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion",
"Conclusion"
] |
[
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"public",
"health",
"and",
"epidemiology/environmental",
"health"
] |
2010
|
Family Relationship, Water Contact and Occurrence of Buruli Ulcer in Benin
|
Clostridium difficile is a global health burden and the leading cause of antibiotic-associated diarrhoea worldwide , causing severe gastrointestinal disease and death . Three well characterised toxins are encoded by this bacterium in two genetic loci , specifically , TcdB ( toxin B ) and TcdA ( toxin A ) in the Pathogenicity Locus ( PaLoc ) and binary toxin ( CDT ) in the genomically distinct CDT locus ( CdtLoc ) . Toxin production is controlled by regulators specific to each locus . The orphan response regulator , CdtR , encoded within the CdtLoc , up-regulates CDT production . Until now there has been no suggestion that CdtR influences TcdA and TcdB production since it is not carried by all PaLoc-containing strains and CdtLoc is not linked genetically to PaLoc . Here we show that , in addition to CDT , CdtR regulates TcdA and TcdB production but that this effect is strain dependent . Of clinical relevance , CdtR increased the production of TcdA , TcdB and CDT in two epidemic ribotype 027 human strains , modulating their virulence in a mouse infection model . Strains traditionally from animal lineages , notably ribotype 078 strains , are increasingly being isolated from humans and their genetic and phenotypic analysis is critical for future studies on this important pathogen . Here we show that CdtR-mediated toxin regulation did not occur in other strain backgrounds , including a ribotype 078 animal strain . The finding that toxin gene regulation is strain dependent highlights the regulatory diversity between C . difficile isolates and the importance of studying virulence regulation in diverse lineages and clinically relevant strains . Our work provides the first evidence that TcdA , TcdB and CDT production is linked by a common regulatory mechanism and that CdtR may act as a global regulator of virulence in epidemic 027 strains .
C . difficile antibiotic-associated diarrhoea is a toxin mediated disease [1 , 2] . During infection , TcdA , TcdB and CDT are secreted into the colonic epithelium by this bacterium , leading to diarrhoea that can progress to serious , life threatening inflammatory diseases , including pseudomembranous colitis and toxic megacolon [3] . The production of these toxins varies between strains . TcdB is the most commonly encoded toxin and is most often co-located with the TcdA gene in the PaLoc region [4] , both toxins act as monoglucosyltransferases that irreversibly modify Rho family members [3] . PaLoc variants that produce TcdB and not TcdA are , however , becoming increasingly common , for example , they represented 23% of strains in one recent study of human strains in China [5] . CDT is encoded in a specific locus , CdtLoc ( Fig 1 ) [6] the carriage of which has also increased significantly over the last decade; in 2004 6% of clinical isolates encoded CDT whereas 33 . 5% now encode this toxin [7 , 8] . CDT is an ADP-ribosyltransferase that is not essential for disease , but may be important for colonisation during an infection [9] . CdtLoc is not carried by all PaLoc-containing strains and it is not linked genetically to PaLoc . Regulation of toxin production in C . difficile is somewhat strain dependent , suggesting that toxin regulatory mechanisms have evolved independently to modulate pathogenesis [10–13] . The TcdR alternative sigma factor and TcdC anti-sigma factor , which are encoded with tcdA and tcdB in the PaLoc , act as the primary mechanism controlling the production of these toxins [14 , 15] . TcdA and TcdB regulation has also been linked to many important cellular processes in the C . difficile life cycle , including sporulation , by Spo0A , the master sporulation regulator , motility , via the flagella regulator SigD , and nutrient acquisition , by the regulators of carbon and amino acid metabolism , CcpA and CodY [10 , 11 , 16–20] . The ribotype 027 strains associated with epidemics of severe CDI appear to be more virulent than strains previously isolated , a phenotype that has been partly attributed to increased TcdA and TcdB production [21–23] . By comparison , little is known about the regulation of CDT production beyond the involvement of CdtR , and until now , no link had been identified between the control of CDT , TcdA and TcdB production . The difficulty in genetically manipulating strains from different C . difficile clonal lineages has also prevented a broader analysis of the role of this regulator across different strain types . In this study , we investigated the role of CdtR in different strains of C . difficile including two epidemic ribotype 027 strains . As expected , CdtR was found to regulate the production of CDT . Surprisingly , however , CdtR also regulated the production of the PaLoc encoded toxins , TcdA and TcdB , in the two ribotype 027 strains . Further analysis showed that regulation occurred at the transcriptional level and probably resulted from indirect regulation of the positive regulator of PaLoc gene expression , TcdR . Importantly , further analysis showed the importance of CdtR for C . difficile pathogenesis , with cdtR mutants causing less severe disease than the wild type strain in a mouse infection model . To determine whether CdtR function is conserved across evolutionarily diverse isolates , ribotype 078 ( JGS6133 ) and 012 ( 630 ) strains were investigated . CdtR regulated CDT production in the ribotype 078 strain; the ribotype 012 strain does not encode CDT . Notably , and in contrast to the ribotype 027 strains , CdtR did not regulated TcdA or TcdB production in either strain background , highlighting the regulatory variation of key virulence factors between C . difficile strains . These results highlight the importance of investigating regulatory mechanisms in clinically important strains of C . difficile and suggest that CdtR-mediated toxin regulation is an important virulence mechanism in the epidemic ribotype 027 strains .
To investigate the role of CdtR in the regulation of CDT production in the epidemic ribotype 027 strains , we constructed two independent cdtR mutants in the Canadian isolate M7404 and a cdtR mutant in the UK isolate R20291 and confirmed their genotype by Southern hybridisation ( S1 Fig ) . Western blot analysis showed that the cdtR mutants produced less CDTa and CDTb compared to the wild type and that complementation with cdtR in trans resulted in over-expression of both toxin subunits ( Fig 2A and 2B ) . Consistent with these results , ADP-ribosyltransferase assays demonstrated that the cdtR mutants had significantly reduced levels of CDT activity compared to the wild type , while the complemented cdtR mutants showed CDT activity greater than that of the wild type strain ( Fig 2D and 2E ) . Overall , these data show , for the first time , that CdtR is important for regulating CDT production in epidemic ribotype 027 C . difficile strains . Unexpectedly , Western blots performed using TcdA- and TcdB-specific antibodies showed that the cdtR mutants produced less TcdA and TcdB than the wild type , while the complemented cdtR derivatives expressed high levels of both toxins ( Fig 3A and 3C ) . Cytotoxicity assays were performed using HT29 and Vero cells to measure the activity of TcdA and TcdB , respectively , in the culture supernatants from the isogenic panel of M7404 and R20291 strains . The TcdA and TcdB activities of all of strains increased over time and , consistent with the Western blot results , showed lower activity in supernatants from the cdtR mutants compared to the wild type , confirming that the cdtR mutants were less cytotoxic in vitro ( Fig 3B and 3D ) . TcdA and TcdB activity of the complemented cdtR mutants was consistently higher than the cdtR mutants and the wild type across all time points ( Fig 3B and 3D ) . These data show that CdtR regulates the production of TcdA and TcdB in both M7404 and R20291 , which is the first demonstration of a common regulator modulating the expression of all three toxins in C . difficile . To investigate the molecular mechanism of regulation , we determined if CdtR controlled toxin production at the transcriptional level . Using the isogenic panel of M7404 strains , reverse-transcription droplet digital PCR ( RT-ddPCR ) analysis was employed to quantitatively compare the level of expression of each of the toxin encoding genes ( tcdA , tcdB , cdtA ) and the PaLoc-encoded toxin regulators ( tcdR , tcdC ) . The relative transcription of all three toxin genes and tcdR was significantly decreased in both cdtR mutants compared to the wild type ( Fig 4A–4D ) . Over-expression of cdtR in the complemented strains resulted in a dramatic over-expression of tcdA , tcdB , cdtA and tcdR ( Fig 4A–4D ) . Although TcdC is predicted to be non-functional in the 027 strains [12] , we analysed the expression of tcdC and found it to be similar in the isogenic panel of strains ( Fig 4E ) . Previous work in strain 630Δerm , a derivative of the non-027 historical strain 630 which belongs to ribotype 012 , showed that the flagella synthesis regulator , SigD , is an important regulator of TcdA and TcdB production via the regulation of tcdR expression [24] . Analysis of our isogenic panel of M7404 strains showed no change in sigD transcription , suggesting that CdtR does not influence the expression of sigD in this strain background and that therefore the modulation of tcdA and tcdB expression does not occur via SigD ( Fig 4F ) . Collectively , these results indicate that regulation of all three toxins by CdtR in M7404 occurs at a transcriptional level and that the regulation of tcdA and tcdB occurs via the upregulation of tcdR transcription . Our observation that CdtR increases toxin production in ribotype 027 strains has important implications for the virulence capacity of these clinically important strains . To determine if the modulatory effect of CdtR on toxin production influences C . difficile disease , we examined whether cdtR inactivation altered virulence in our mouse infection model [2] . It was previously shown that infection with a cdtA mutant of M7404 resulted in disease that was indistinguishable from the parent strain [2] . This mutant no longer produced CDT , but had an intact cdtR gene and continued to produce TcdA and TcdB at wild type levels . A reduction in CDT levels is therefore not likely to have a major effect on disease in our animal model . All of the mice infected with the isogenic cdtR-series of M7404 derivatives were colonised with C . difficile at similar levels ( S2 Fig ) . Wild type-infected mice rapidly lost weight and the majority were euthanized 40 to 48 hours post infection in accordance with animal ethics guidelines , with a mean time to death of 48 ± 5 . 1 hours and a survival rate of 13% ( Fig 5 ) . Mice infected with either of the two independent cdtR mutants had significantly higher survival rates ( Mantel-Cox log rank test , P < 0 . 0001 ) of 100% and 96% for cdtR1 ( V ) and cdtR2 ( V ) , respectively , and showed no overt signs of disease nor significant weight loss ( Fig 5 ) . By comparison , mice infected with either complemented mutant , cdtR1 ( R+ ) or cdtR2 ( R+ ) , had a wild-type virulence phenotype , with marked weight loss , and a mean time to death of 31 . 5 ± 2 . 5 hours and 57 . 6 ± 7 . 5 hours , respectively , reflected in survival rates of 0% and 33% ( Fig 5 ) . Damage to the colon and caecum of C . difficile-infected mice results from the production of TcdA and TcdB [2] . We therefore performed histopathological analysis to assess the damage to colonic and caecal tissues collected from the groups of infected and uninfected mice under study here . All tissues were de-identified and independently scored using a previously defined set of scoring parameters that included overall tissue damage , polymorphonucleocyte ( PMN ) influx , crypt damage and oedema [2] . Tissues of uninfected mice only had minimal surface damage to the intestinal epithelia resulting from the disruption of microbiota by antibiotic pre-treatment and tissue processing ( Fig 6A ) with low colon and caecum damage scores of 4 . 7 and 3 . 8 , respectively ( Fig 6B and 6C ) . By comparison , wild type-infected mice had severely inflamed tissues , with extensive damage to the epithelial surface , crypt branching and hyperplasia , goblet cell loss , significant PMN influx and mucosal and sub-mucosal oedema ( Fig 6A ) . These histopathologies were reflected in the high damage scores of 12 . 9 and 13 . 6 for their colonic and caecal tissues , respectively ( Fig 6B and 6C ) . Mice infected with either cdtR mutant had tissue architecture similar to that seen with the uninfected mice . Very little colonic and caecal damage was observed , with some surface epithelial damage to crypts occurring and no apparent crypt hyperplasia , and little PMN influx into the mucosa ( Fig 6A ) , with correspondingly low colonic ( 6 . 0 and 5 . 3 ) and caecal ( 6 . 6 and 7 . 1 ) damage scores ( Fig 6B and 6C ) . However , mice infected with either cdtR-complemented mutant had similar levels of tissue damage to mice infected with the wild-type strain . Severe crypt damage was observed in the majority of these tissues , particularly the caeca , with crypt hyperplasia , loss of goblet cells , PMN influx and severe oedema in the mucosa and sub-mucosa ( Fig 6A ) and high histopathological damage scores were determined for both the colon ( 11 . 8 and 9 . 6 ) and the caecum ( 14 . 5 and 11 . 4 ) ( Fig 6B and 6C ) . To confirm that the reduced virulence of the cdtR mutants can be attributed to reduced levels of TcdA and TcdB production in vivo , the cytotoxicity of the intestinal contents collected from mice infected with the panel of C . difficile strains was assessed . No cytotoxicity was observed against HT29 or Vero cells by faecal samples collected from uninfected mice whereas mice infected with the wild type strain showed high levels of cytotoxicity against HT29 and Vero cells , indicating the production of TcdA and TcdB in vivo ( S4 Fig ) . By comparison , samples collected from mice infected with the two cdtR mutants ( cdtR1 ( V ) and cdtR2 ( V ) ) showed reduced cytotoxicity against HT29 or Vero cells , indicating decreased levels of TcdA and TcdB production in vivo compared to the wild type strain ( S4 Fig ) . Intestinal contents collected from mice infected with the two complemented cdtR mutants ( ( cdtR1 ( R+ ) and cdtR2 ( R+ ) ) showed increased levels of cytotoxicity against HT29 and Vero cells , indicating restored in vivo TcdA and TcdB production . Collectively , the survival data , histopathological and in vivo cytotoxicity analysis support the hypothesis that CdtR modulates the virulence of the C . difficile ribotype 027 strain M7404 due to the role that it plays in regulating TcdA and TcdB production . To investigate whether TcdA and TcdB regulation by CdtR occurs in other strains , we assessed this phenotype in a ribotype 078 animal isolate , JGS6133 . Ribotype 078 strains are commonly isolated from animals , but have also been implicated in severe human infections [25 , 26] . Our sequencing analysis revealed that the cdtR gene in JGS6133 contains a naturally occurring stop codon mutation at codon 322 , which has been described previously in other 078 strains [27] , and results in a 142 amino acid truncation that is likely to result in a non-functional protein ( Fig 1B ) . Western blots showed that JGS6133 complemented with the full length cdtR gene in trans , and hence producing functional CdtR , had significantly more CDTa and CDTb than the wild-type and vector control strains , which produced similar amounts of both proteins ( Fig 2C ) . Increased CDT production by the JGS6133 cdtR+ derivative was reflected in increased toxin activity as assessed by ADP-ribosyltransferase assays ( Fig 2F ) . By contrast , production of TcdA and TcdB was not altered in the JGS6133 cdtR+ strain compared to the wild type ( Fig 3E ) . This result suggests that CdtR-mediated regulation of TcdA and TcdB production is not conserved within all strains of C . difficile . Derivatives of a ribotype 012 strain , designated 630 , have been routinely used for the analysis of gene regulation and other phenotypes in C . difficile due to the relative ease of their genetic manipulation . Many strains of C . difficile , including strain 630 , contain cdtAB pseudogenes within the CdtLoc ( Fig 1C ) . These genes contain multiple frameshift mutations and stop codons and do not encode a functional CDT , but still encode a full length CdtR protein [28] . The possibility that CdtR functions to regulate other important processes , such as TcdA and TcdB production , may provide a rationale for the retention of functional CdtR in strains that carry the cdtAB pseudogenes . We were therefore interested in investigating CdtR-mediated toxin regulation in a strain with these pseudogenes . To determine whether CdtR regulates TcdA and TcdB production in strains of C . difficile with cdtAB pseudogenes we transferred the cdtR complementation vector , pJIR4218 , into strain 630 . Western blot analysis showed that derivatives of 630 over-expressing cdtR had similar levels of TcdA and TcdB production as the isogenic vector control strains ( S3A and S3B Fig ) . RT-ddPCR analysis confirmed that cdtR was over-expressed in these strains relative to the wild type and the vector control ( S3C Fig ) . Taken together these results suggest that CdtR is not important for the regulation of TcdA and TcdB production in strain 630 . Since CdtR is also not important for TcdA and TcdB regulation in a ribotype 078 strain , regulation of these toxins by CdtR may be specific to the ribotype 027 strains .
The emergence of epidemic ribotype 027 strains over a decade ago prompted several investigations of the genetic and phenotypic characteristics that may have led to the global dominance of these strains . These features may include higher sporulation rates , resistance to key antibiotics and unique aspects of toxin regulation [11 , 12 , 29–31] . The presence of a full length CdtLoc was also initially considered to be important in this regard because it encodes CDT [21 , 23] , however , despite numerous studies , the importance of this toxin in virulence remains undefined [2 , 32 , 33] . The results of the work presented here suggest that CdtLoc , and specifically CdtR , may play an indirect but significant role in disease pathogenesis of the ribotype 027 strains by regulating TcdA and TcdB production . Our work confirms that CdtR enhances CDT production . Strikingly , ribotype 078 strains produce CDT even though they contain a conserved mutation in cdtR and they have CDT activity that is not significantly different from that of strains without this mutation [27 , 34] . Our data supports these observations since CDT production was detected from the ribotype 078 strain JGS6133 which contains this naturally occurring cdtR mutation . However , CDT production from JGS6133 was enhanced when functional CdtR was expressed in this strain . Even though CdtR is not essential for CDT production , our work suggests that the presence of this regulator increases the expression of this toxin . Although CdtR was previously shown to be an important regulator of CDT production [6] there was no evidence to suggest that it played a role in TcdA or TcdB production , particularly since the pathogenicity loci encoding these toxins are not genomically linked . Our data clearly show that CdtR is an important regulator of TcdA and TcdB production , as well as CDT , in two ribotype 027 strains , and that this regulatory capacity plays a role in virulence since inactivation of cdtR attenuated the virulence of strain M7404 in a mouse infection model . The results obtained with the mouse infection experiments are directly relevant to the disease-causing capacity of these strains . This is a significant finding as it is the first report of a regulatory link between the two pathogenicity loci , PaLoc and CdtLoc . The ability to regulate toxin production through a variety of mechanisms may provide a selective advantage to C . difficile since it may allow virulence factors to be produced in response to different and specific environmental cues . Ribotype 027 strains appear to have evolved to differ in their regulatory responses in comparison to other C . difficile lineages . C . difficile regulates toxin production in response to many environmental stimuli , including metabolisable carbon sources and quorum signalling molecules , through several different regulatory proteins [20 , 35 , 36] . TcdR is the primary positive regulator of TcdA and TcdB production , while TcdC represses toxin production [12 , 15] . TcdR is highly conserved between strains and many regulators directly influence its expression [20 , 24 , 27 , 35] . Our data suggest that CdtR may regulate TcdA and TcdB production by controlling tcdR expression . CdtR belongs to the LytTR family of DNA binding response regulators and may function by binding to the tcdR promoter , thereby regulating TcdR expression and , consequently , tcdA and tcdB expression [6] . However , we could not identify canonical LytTR DNA binding sites upstream of the tcdR , tcdA or tcdB genes . Similarly , LytTR DNA binding sequences with the conserved sequence and spacing could not be identified within the promoters of other genes identified to regulate toxin production in C . difficile in other studies , including sigD , codY or ccpA . It may be that the CdtR binding sites are too dissimilar to typical LytTR sites to be identified or that CdtR does not directly bind to these regions; instead , an unidentified , CdtR-controlled intermediate regulator may be modulating PaLoc toxin gene expression . CdtR-mediated TcdA and TcdB regulation may have specifically evolved in ribotype 027 C . difficile strains since the co-regulation of these toxins with CDT appears to be ribotype specific . CdtR does not play a similar role in other ribotypes tested here to that seen in ribotype 027 strains and this phenotype does not appear to be conserved between divergent strain backgrounds . The strains included in our assessment belong to three of the five defined evolutionary clades of C . difficile , specifically , clade 2 for ribotype 027 ( M7404 and R20291 ) , clade 5 for ribotype 078 ( JGS6133 ) and clade 1 for ribotype 012 ( 630 ) [37] . Our observations in strain 630 are particularly relevant; many studies are performed using this isolate because it is relatively easy to genetically manipulate . It is clear from our research and other studies [10 , 11] that strain 630 characteristics may not always reflect those of other strains . Similar observations have been made for clade 5 ribotype 078 strains , which are genetically and phenotypically divergent from strains belonging to other clades [11 , 30 , 38 , 39] . Although the global regulators CodY and CcpA are conserved in strains of C . difficile , it has been shown that these regulators control toxin production experimentally only in a strain 630 background and their role in other strains , including the 027 and 078 strains , is not known [19 , 20] . Similarly , several flagella structural and regulatory genes , including SigD , have only been linked to toxin production in C . difficile in strain 630 [24] . While the genetic organisation of the flagella genes within the F1 and F3 flagella regions are similar in strain 630 and the 027 strains , the sequence variation in these regions is thought to contribute to their different motility phenotypes [30] . By comparison , the F3 region , which contains several genes involved in toxin regulation in strain 630 , is absent in the 078 strains and is thought to explain the lack of motility in these strains [30] . It has been shown that several of the conserved flagella structural proteins encoded in the F1 and F3 flagella regions , including FliC , FliD and FlgE , are important for toxin production in strain 630 but do not contribute to toxin production in the 027 strain , R20291 [10] . Further research is required to determine if other conserved flagella genes , known to regulate toxin production in strain 630 , play a similar role in 027 and 078 strain backgrounds . To date , only one study has investigated the regulation of toxin production in an 078 strain and showed that the master sporulation regulator , Spo0A , differentially regulates toxin production in an 078 strain , two epidemic 027 strains and a strain 630 derivative [11] . Dingle et al . [40] found that strains from clade 5 , including the 078 strains , carry a PaLoc similar to that found in other ribotypes but that genes outside of this region are highly divergent . It was suggested that the 078 strains may have originated from a divergent , non-toxigenic strain that obtained the PaLoc in a separate event in comparison to other toxigenic lineages . We present data supporting the concept that strains from this background have evolved different toxin regulatory mechanisms from the more commonly studied 027 strains and strain 630 derivatives . The results presented here clearly show that modulation of tcdA and tcdB expression by CdtR may be specific to the ribotype 027 strains and is likely to be an important factor contributing to their increased virulence . Furthermore , the fluidity of the regulatory systems that control gene expression in C . difficile , exemplified by the toxin gene expression studies presented here , reflect the plasticity and dynamic nature of the C . difficile genome [37] . In conclusion , we have provided the first evidence that TcdA and TcdB production is linked to the production of CDT by a common regulatory mechanism and that CdtR acts as a global regulator of toxin production and virulence in two ribotype 027 strains . The observed differences in virulence between the ribotype 027 strains and other historical isolates have been attributed , in part , to elevated toxin production , mainly as a result of mutations in the tcdC gene [21–23] . Another key genetic difference identified between these strains is the possession of a full length CdtLoc [30] . Our results suggest that possession of the CdtLoc in the 027 strains enhances virulence by the CdtR-mediated up-regulation of TcdA and TcdB production . Therefore , we postulate that the ability of the epidemic ribotype 027 strains to coordinate production of all known C . difficile toxins , CDT , TcdA and TcdB , by CdtR is a key factor in the increased virulence of these strains .
All bacterial strains are defined in Table 1 . Culture media were from Oxoid or Becton Dickinson ( BD ) and all antibiotics and supplements used are from Sigma-Aldrich , Amresco or Merck unless otherwise stated . E . coli and B . subtilis strains were cultured at 37°C in 2xYT media [41] supplemented with either chloramphenicol ( 25 µg/ml for E . coli; 5 μg/ml for B . subtilis ) or tetracycline ( 10 µg/ml ) . C . difficile strains were cultured in HIS broth[42] or on HIS agar supplemented with 0 . 1% ( w/v ) L-cysteine and 0 . 375% ( w/v ) glucose or TY broth[1] with D-cycloserine ( 250 µg/ml ) , thiamphenicol ( 10 µg/ml ) , lincomycin ( 50 µg/ml ) or anhydrous tetracycline ( 50 ng/ml ) , as required . C . difficile cultures were grown in a Don Whitley A35 Anaerobic Workstation in an atmosphere of 10% ( v/v ) H2 , 10% ( v/v ) CO2 and 80% ( v/v ) N2 at 37°C . All oligonucleotide primers are listed in Table 2 . PCR cycling conditions ( unless otherwise stated ) were as follows: initial denaturation step at 94°C for 4 min , followed by 30 cycles of denaturation at 94°C for 30 sec , an annealing step at 50°C for 30 sec and an extension step at 72°C for 1 min per 1 kb . A final extension step was performed at 72°C for 10 minutes . PCRs were performed with Phusion DNA polymerase ( New England Biolabs ) and 2x Failsafe PCR buffer E ( Epicentre Biotechnology ) . Splice-overlap extension ( SOE ) -PCR to re-target the Targetron was performed as described in the TargeTron Gene Knockout System users guide ( Sigma-Aldrich ) with modifications as previously described [43] . Plasmid DNA was isolated from E . coli , B . subtilis and C . difficile using QIAprep spin miniprep columns ( Qiagen ) following the manufacturer’s instructions . Genomic DNA was isolated from C . difficile as previously described [44] . Standard methods of DNA digestion , modification and ligation were used . DNA sequencing was carried out using BigDye Terminator v3 . 1 Ready Reaction Mix ( Applied Biosystems ) following the manufacturer’s instructions . Sequencing reactions were resolved on an Applied Biosystems 3730 DNA Analyzer . Sequences were analysed using ContigExpress ( Invitrogen ) . All plasmids are outlined in Table 1 . Construction of the cdtR TargeTron plasmid was performed as previously described , with some modifications [43] . Briefly , the group II intron from pDLL45 was retargeted by SOE-PCR to insert between nucleotides 288 and 289 of the cdtR gene using the primer pairs JRP5448 and JRP3867 and JRP5449 and JRP5450 ( Table 2 ) to generate a 350 bp product , which was digested with BsrGI and HindIII and cloned into the corresponding sites of pDLL45 , resulting in pJIR4135 . A StuI-HindIII fragment was then sub-cloned from pJIR4135 into the corresponding sites of pDLL55 , resulting in pJIR4153 . The cdtR complementation plasmid was constructed by PCR amplifying the cdtR gene and approximately 300 bp of its promoter region from C . difficile M7404 using the primers JRP5632 and JIR5633 ( Table 2 ) . The resulting 1 . 1 kb fragment was purified using a PCR purification kit ( Qiagen ) following the manufacturer’s instructions , digested with BamHI and PstI and cloned into the corresponding sites of pDLL24 , resulting in pJIR4218 . Plasmid DNA was introduced into the B . subtilis conjugative donor strain BS34A as previously described [45] . The resulting strain was used as the donor for the conjugative transfer of plasmid DNA into C . difficile strains as before [11] . C . difficile cdtR mutants were isolated using the method previously described [11] and confirmed by PCR and Southern hybridisation analysis . Complementation of the mutation was achieved using the cdtR complementation plasmid , pJIR4218 . The cloning vector , pDLL24 , was transferred into the cdtR mutant and the wild-type strain to construct vector ( v ) controls . Toxins were partially purified and concentrated eight-fold from 72 hour C . difficile TY culture supernatants by methanol-chloroform precipitation [11] . Protein concentrations were determined using the BCA protein assay kit ( Pierce ) as per the manufacturer’s instructions . Concentrated supernatant proteins ( 10 µg ) were separated by 10% ( v/v ) sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) [46] and transferred onto a nitrocellulose membrane ( Whatman ) . Membranes were analysed as previously described [43] . TcdA and TcdB were detected using TcdA-specific monoclonal and TcdB-specific polyclonal antibodies ( tgcBIOMICS ) , respectively . CDTa and CDTb were detected , respectively , using a CDTa-specific antibody and C . perfringens Ib-specific antibody that is cross reactive with CDTb [47] . CDTa , CDTb and TcdB-bound antibodies were detected using horseradish peroxidase conjugated anti-rabbit goat antibodies ( Millipore ) and TcdA-bound antibodies were detected using anti-mouse goat antibodies ( Millipore ) . The Western Lightning Chemiluminescence reagent kit ( Perkin-Elmer ) was used to detect the bands , which were visualised by exposure to X-ray film or on a BioRad ChemiDoc XRS+ system . Toxins were partially purified from culture supernatants by precipitation with 70% ammonium sulphate as described previously [6] . ADP ribosyltransferase assays were performed as previously described [48] . Briefly , precipitated supernatant protein ( 50 µg ) was incubated for 60 minutes at 37°C with 10 µg of actin in assay buffer ( 20 mM Tris-HCl pH 7 . 5 , 1 µM dithiothreitol ( DTT ) , 40 µM ATP , 40 µM CaCl2 , 5 µM MgCl2 ) and 10 µM of biotinylated NAD+ ( Trevigen ) . The reaction was heat inactivated at 95°C for 5 minutes in 4x SDS sample buffer ( 240 mM Tris-Cl ( pH 6 . 8 ) , 40% glycerol ( v/v ) , 8% SDS ( w/v ) , 5% ( v/v ) 2-mercaptoethanol , 0 . 05% ( v/v ) bromophenol blue and separated by 10% SDS-PAGE . Proteins were transferred onto a nitrocellulose membrane and biotinylated proteins were detected with horseradish peroxidase-conjugated streptavidin ( GE Healthcare Life Sciences ) and the Western Lightning Chemiluminescence reagent kit ( Perkin-Elmer ) , following the manufacturer’s instructions . Relative band intensities were determined by densitometry using Image Lab Software ( Bio-Rad ) . Data were analysed using GraphPad Prism 6 and statistical significance assessed using an unpaired t-test with a 95% confidence interval . C . difficile strains were grown overnight in 20 ml of HIS broth with thiamphenicol and lincomycin , as required . The cultures then were used to inoculate 50 ml of TY broth with selection , such that each culture had a starting OD600 of approximately 0 . 05 . Aliquots ( 5 ml ) were taken at 12 , 24 , 48 and 72 hours , pelleted by centrifugation ( 10 , 000 g , 10 min , room temperature ) and the supernatants filter sterilised through 0 . 45 µM and 0 . 2 µM filters ( Sartorius ) . Supernatants were stored on ice until use . Vero cell and HT29 cell cytotoxicity assays were performed using the filtered C . difficile supernatants as previously described [43] . The levels of TcdA and TcdB produced by the C . difficile strains in vivo was assessed by determining the cytotoxicity of the intestinal contents collected 24 hours post infection against HT29 and Vero cells . Intestinal samples were resuspended in 100 mg/ml in PBS , diluted one in eight , filter sterilised and applied to Vero and HT29 cells , as described previously [2] . The endpoint ( toxin titre ) was scored as the last dilution with 100% cytopathic effect ( CPE ) . Data were analysed using GraphPad Prism 6 and statistical significance assessed using an unpaired t-test with a 95% confidence interval . Total RNA was extracted using TRIzol® ( Life Technologies ) following the manufacturer’s instructions . Forty ml of C . difficile TY broth cultures with an OD600 of approximately 0 . 3 for tcdC and cdtA expression analysis , and 10 ml of TY broth culture grown to OD600 of approximately 1 . 8 for tcdA , tcdB and tcdR expression analysis , were used . A total of 200 ng of RNA was converted to cDNA using SuperScript III Reverse Transcriptase ( Life Technologies ) , following the manufacturer’s instructions . Transcript levels were quantified using the QX200 Droplet Digital PCR System ( BioRad ) using QX200 ddPCR EvaGreen Supermix and 0 . 1–5 µg of total cDNA and specific primers ( Table 2 ) at a concentration of 200 nM . Transcription levels of each gene was normalised to transcription levels of the housekeeping gene rpoA . Data were analysed using GraphPad Prism 6 and statistical significance assessed using a Mann Whitney U test . Groups of five male six to eight week old C57BL/6 mice were used in C . difficile virulence trials as previously described [2] , except mice were switched back to plain drinking water on the day of infection . Mice were administered 106 C . difficile spores by oral gavage and were humanely euthanised at the onset of severe disease or at the end of the experiment , as previously defined [2] . Animal handling and experimentation were performed in accordance with institutional guidelines ( Monash University animal ethics committee numbers MARP/2014/135 and SOBSB/M/2010/25 ) . Faecal samples were taken daily to monitor C . difficile shedding using HIS agar supplemented with 0 . 1% ( w/v ) cysteine , 0 . 1% ( w/v ) taurocholate , 0 . 375% ( w/v ) glucose , 250 µg/ml D-cycloserine , 8 µg/ml cefoxitin , 10 µg/ml erythromycin , 12 µg/ml norfloxacin , 32 µg/ml moxalactam . Data were analysed using GraphPad Prism 6 and statistical significance assessed using a log-rank ( Mantel-Cox ) test . The entire colon and caecum were collected from each mouse and Swiss-rolled [49] prior to fixation to allow for cross-sectional examination of the entire length of the colon . Tissues were stained with PAS-Alcian blue and histopathological assessment of damage and scoring of tissues was performed blind by independent observers using a previously defined set of parameters [2] .
|
Clostridium difficile is the leading cause of antibiotic-associated diarrhoea . The TcdB , TcdA and binary toxins produced by C . difficile are encoded in two genomically distinct loci: TcdB and TcdA in the Pathogenicity Locus ( PaLoc ) and binary toxin ( CDT ) in the CDT locus ( CdtLoc ) . Toxin production is primarily controlled by regulators specific to each locus . Because the presence of these loci varies amongst different strains of C . difficile , no rational link for their co-regulation has ever been proposed . Here we have shown that the regulator of CDT production , CdtR , also regulates production of TcdA and TcdB in a strain dependent manner . These results represent the first evidence that TcdA and TcdB production is linked to the production of CDT by a common regulatory mechanism . Collectively , our results establish CdtR as an important virulence regulator in two clinically important , epidemic strains of C . difficile , and further highlights the need to investigate regulatory mechanisms of important virulence factors in diverse strain backgrounds .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"complement",
"system",
"medicine",
"and",
"health",
"sciences",
"gut",
"bacteria",
"toxins",
"pathology",
"and",
"laboratory",
"medicine",
"immune",
"physiology",
"gene",
"regulation",
"pathogens",
"immunology",
"toxicology",
"toxic",
"agents",
"regulator",
"genes",
"gene",
"types",
"infectious",
"disease",
"control",
"bacteria",
"clostridium",
"difficile",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"proteins",
"gene",
"expression",
"immune",
"system",
"biochemistry",
"virulence",
"factors",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2016
|
CdtR Regulates TcdA and TcdB Production in Clostridium difficile
|
Polycomb group ( PcG ) proteins are essential for the repression of key factors during early development . In Drosophila , the polycomb repressive complexes ( PRC ) associate with defined polycomb response DNA elements ( PREs ) . In mammals , however , the mechanisms underlying polycomb recruitment at targeted loci are poorly understood . We have used an in vivo approach to identify DNA sequences of importance for the proper recruitment of polycomb proteins at the HoxD locus . We report that various genomic re-arrangements of the gene cluster do not strongly affect PRC2 recruitment and that relatively small polycomb interacting sequences appear necessary and sufficient to confer polycomb recognition and targeting to ectopic loci . In addition , a high GC content , while not sufficient to recruit PRC2 , may help its local spreading . We discuss the importance of PRC2 recruitment over Hox gene clusters in embryonic stem cells , for their subsequent coordinated transcriptional activation during development .
Polycomb group ( PcG ) proteins are essential for proper development of most eukaryotic organisms . The founding member ( Polycomb ) was characterized genetically as a repressor of Drosophila homeotic genes [1] and subsequent studies have established these proteins as key organizers of the epigenome ( e . g . [2] ) . For instance , they play important roles in the stable repression of genes via epigenetic mechanisms such as X-inactivation [3] and imprinting [4] , [5] , as well as during cell cycle regulation and differentiation ( e . g . [6] , [7] , [8] ) . The fine-tuned balance between the activities of PcG proteins and proteins from the trithorax families ( trxG ) , displaying opposing functions , was shown to maintain Hox gene expression during the entire life of Drosophila [9] . However , neither the exact process ( es ) whereby these proteins impose their repressive effect , nor the specific mechanism ( s ) involved in the recognition and tethering to target genomic loci are as yet fully understood . In mammals , PcG proteins are mostly found in two large complexes: the Polycomb Repressive Complexes 1 and 2 ( PRC1 , PRC2 ) . PRC2 carries a methyl-transferase activity that methylates the histone H3 tail at lysine 27 , a mark largely associated with gene silencing . While the initial deposition of this post-transcriptional modification is carried through by PRC2 , PRC1 maintains this methylated status and compacts chromatin , largely , though not solely [10] , by the ubiquitylation of lysine 119 of histone H2A [11] . It was shown that specific PRC1 type complexes are recruited by PRC2 [12] , which can also compact chromatin , though to a lesser extent . The importance of these complexes for various developmental and differentiation processes is reflected by the early lethality induced by the loss-of-function of several of their components such as Suz12 [13] , Ring1B [14] , Ezh2 [15] and Eed [16] . These components are usually well conserved throughout metazoans , suggesting that both the global operational mode of these complexes , as well as the way they are recruited to specific loci may be comparable between species . In Drosophila , PcG group proteins are recruited to chromatin through the recognition of polycomb response elements ( PREs ) . These DNA segments , which were generally identified by forward genetics , must satisfy three criteria: ( 1 ) they must bind polycomb when randomly inserted into the genome , ( 2 ) they must be able to induce a H3K27me3 domain and ( 3 ) they must repress a reporter construct , when associated with them [17] , [18] , [19] . Drosophila PREs are approximately 1 . 5 kb long in average and are usually located in proximal promoter regions . This is the case of the PREs associated with either engrailed [20] , [21] , Hedgehog [22] , [23] or polyhomeotic [24] , [25] . However , PREs have also been mapped in vivo several kilobases away from their target genes and their identification may be biased by genome wide chromatin immunoprecipitation ( ChIP ) studies . Recently indeed , it has become clear that only a fraction of polycomb enriched regions are direct targets of PRC , whereas others are indirect , via chromatin looping [26] , suggesting that polycomb enriched regions may not always reflect the genuine presence of polycomb at a particular locus . Instead , they may derive from the three-dimensional organization of the genome , which along with the technology employed , may lead to false positives . Furthermore , although high throughput studies have shown that the binding profiles of polycomb proteins correlate with both the transcription start sites ( TSSs ) and stalled RNA PolII [27] , [28] , they do not share any salient sequence homology such that no consensus motif has been identified thus far [29] . In mammals , the understanding of the general mechanism ( if any ) accounting for the recruitment of PRC2 is lacking too . PRC2 recruitment has been associated with the presence of binding sites for the Pho ortholog YY1 [30] , [31] , although the physical interaction between Pho and PRC remains controversial [32] , [33] , [34] , [35] , [36] . Other candidates include various transcription factor binding sites [37] , high density of unmethylated CpG dinucleotide regions or the presence of a TSS [38] , [39] , [40] , [41] , [42] , [43] . While these explanations apply individually to a range of particular situations , they cannot fully account for the apparent high specificity of PRC2 recruitment genome-wide . PcG proteins are found over developmental genes in pluripotent embryonic stem ( ES ) cells , including the four Hox gene clusters ( Fig . S1 and [44] , [45] , [46] ) . In this uncommitted state , a significant fraction of Pc target loci also carry trxG proteins and their H3K4me3 epigenetic marks . Such ‘bivalent domains’ displaying both H3K27me3 and H3K4me3 modifications are found over genes poised for transcription . During embryonic development , these domains lose either one of the two marks and thus acquire a univalent epigenetic status . In mammals , expression of Hox genes in the foremost anterior structures is tightly repressed and , accordingly , H3K4me3 is lost over the four Hox clusters , whereas the coverage by PcG proteins and H3K27me3 is re-enforced [44] . This emphasizes the necessity , for an organism , to properly and selectively secure the recruitment of PRC2 to the appropriate loci , at the right time . We have addressed the question of polycomb recruitment at Hox loci by using an in vivo approach , based on the large number of genomic re-arrangements associated with the HoxD locus [47] , a DNA region that is amongst the most heavily covered by H3K27me3 marks in ES cells and where one of the few vertebrate PREs has been previously identified [31] . In contrast to what is observed in the Drosophila Bithorax complex where a 300 kb large domain of trimethylated histones involves only few PREs , we show that the mouse HoxD locus implements a mechanism that can compensate for large and systematic deletions within the target DNA interval . These results indicate that PRC2 recruitment at this locus must rely upon a range of cooperating binding sites , rather than upon a few nucleation sites . By using isolated transgenes , we also show that in this particular context , CpG islands ( or a high GC content ) are not the prime factors in PRC2 recruitment , despite their potential importance for local spreading .
To study the recruitment of polycomb complexes and the resulting H3K27me3 histone modification in vivo , we used a genetic approach of the mouse HoxD gene cluster , which is a main target of Pc silencing in ES cells and adult tissues [44] , [48] . Also , this locus has been shown to contain one of the few defined mammalian PREs [31] . We assessed the binding profiles of different members of the polycomb complexes in both wild type and deletion alleles using chromatin immunoprecipitation ( ChIP ) . Should a given part of this gene cluster be of particular importance for recruiting PRC , its deletion may lead to modifications of the binding of PRC proteins and/or of the general H3K27me3 profile . Deletion of parts of the HoxD cluster did not seem to affect the binding profiles of PRC proteins when assessed by ChIP-qPCR ( Fig . S1 ) . We hybridized the ChIPed material to high density tiling arrays and the overall binding profiles remained largely unchanged throughout the cluster , including those peaks assessed by ChIP-qPCR ( Fig . 1A , B ) . As it was shown that the loss of PRC1/2 leads to the loss of H3K27me3 marks [49] and that H3K27me3 is considered a hallmark of PRC2-mediated gene repression [2] , we focused on the analysis of H3K27me3 mark . We used mutant mice carrying complementary deletions covering the entire HoxD cluster to try and detect if any of the deleted part would impact upon this epigenetic modification . We performed ChIP of H3K27me3 from E13 embryonic brains ( Fig . 2A , B ) , a tissue where Hox genes are both enriched in PRC1 and PRC2 and where Hox trancripts are virtually absent . Also the distribution in H3K27me3 marks over Hox loci in fetal brain resembles that found in embryonic stem cells [38] , [44] . This material was hybridized to tiling arrays covering the mouse HoxD cluster [50] . We first evaluated the impact of deleting a small DNA sequence located between Hoxd11 and Hoxd12 , which in human cells showed the hallmarks of PREs including the tethering of polycomb proteins and the silencing of an associated reporter gene [31] . This region includes the highly conserved region RX , whose deletion ( del ( RX ) ) in vivo had no effect upon gene expression within the HoxD cluster [51] , [52] , nor did it significantly change the H3K27me3 profile when compared to wild-type animals ( Fig . 2C ) . The PRE reported in this region [31] is slightly larger than region X itself and essential sequences may not have been included in this short deletion . We thus looked at a deletion removing the entire PRE as well as some flanking DNA ( Fig . S2; del ( 12 ) ) . Here again , we did not score any modification of the H3K27me3 profile throughout the gene cluster . We next scanned the entire locus with a set of adjacent deletions , including either Hoxd8 , Hoxd9 , Hoxd10 , Hoxd11 , Hoxd12 or Hoxd13 and found no significant change in polycomb-mediated silencing ( data not shown ) , using H3K27me3 as a proxy for PcG occupancy ( Fig . 2C and Fig . S2 ) . Altogether , and in agreement with results obtained in Drosophila [53] , our data suggest that PRC2 recruitment to mammalian Hox clusters does not rely upon few strong PREs , whose activities would then further spread over the rest of the locus . As PREs could be formed by the addition of several low affinity sites , we analyzed deletions of several contiguous genes . The profiles remained surprisingly unmodified , as illustrated by Del ( 10-13 ) and Del ( 9-12 ) ( Fig . 2C ) . These two deletions were of particular interest since they cover both the previously described PRE [31] and the region de-repressed in the absence of the LncRNA HOTAIR , which was proposed to bring PRC2 over the HOXD cluster [48] . Yet they did not change the H3K27me3 coverage , neither in the Evx2 locus , nor over the rest of HoxD . Other deletions involving several genes in cis gave the same result , with no obvious variation in the profiles of H3K27me3 ( Fig . 2C and Fig . S2 ) , suggesting that the mechanism recruiting Pc proteins over this locus is robust and can compensate for drastic genomic re-arrangements . We next asked whether the extremities of the Hox gene cluster were of particular importance to set up a platform for recruiting PRC2 . We used a large deletion ( del ( 1-10 ) ) , where two-thirds of the anterior part of the cluster were removed including its most anterior gene Hoxd1 . Again , the remaining mini-cluster was able to compensate for this significant trimming and the H3K27me3 pattern over the remaining loci was nearly identical to that found in wild type conditions ( Fig . 2C ) . In fact , even the deletion of the entire HoxD cluster , from Hoxd1 to Hoxd13 ( del ( 1-13 ) d11Lac ) , did not significantly affect the presence of these epigenetic marks over the remaining 5′ located Evx2 gene ( Fig . S2 ) . In this case , interestingly , the H3K27me3 profile covering the Evx2 region was similar , in terms of relative peak intensities , to that observed with the shorter del ( 4-13 ) , even though the latter displayed massive amounts of H3K27me3 over the anterior part of the cluster ( from Hoxd1 to Hoxd3; Fig . S2 ) . This result indicated that each piece of the gene cluster is rather independent in its ability to recruit PRC2 , regardless of what would happen over the neighboring loci . The poor impact of the neighboring sequences upon the coverage of any given Hox gene loci by H3K27me3 was confirmed by the comparative analysis of del ( 10 ) , del ( 13 ) and del ( 10-13 ) , which share the same breakpoints but in various configurations . In this set of deletions , the reconstitution of three different neighborhoods did not modify the methylation patterns . Single gene deletions did not markedly change the relative distance between transcription units and hence they may not affect PRC2 recruitment , should several PREs locate near each gene and synergize . We thus modified the distance between two genes by excising the longest gene free DNA segment within HoxD . This intergenic region ‘i’ is over 13 kb long and maps between Hoxd4 and Hoxd8 . Its deletion ( del ( i ) ) results in a further concentration of genes by bringing Hoxd1 , Hoxd3 and Hoxd4 closer to the centromeric ( posterior ) side including Hoxd8 to Hoxd13 . del ( i ) did not show any clear difference in the H3K27me3 profile over the remaining parts of the gene cluster ( Fig . 2C , del ( i ) ) . We next looked at the effect of introducing a gene free region within the cluster such as to increase the distance between neighboring genes . We duplicated region i to produce a cluster with a 26 kb large gene-free domain inside . Intriguingly , the duplicated configuration did not exhibit any change on region i , except for a weak gain of signal over the duplicated region suggesting that both copies are covered by H3K27me3 . The flanking DNA segments , however , displayed the same H3K27me3 profiles as in wild type brains ( Fig . 2C and Fig . S2 , del ( i ) , dup ( i ) ) indicating that the mechanism recruiting PRC2 over the mouse HoxD cluster can compensate for modifications in the distance between transcription units , emphasizing once more the robustness of this process . In the absence of any strong and discrete signal for PRC2 recognition and nucleation within the cluster itself , we asked whether Pc proteins may be targeted by elements localized within the regulatory landscapes flanking the HoxD cluster , which contain numerous cis-acting sequences . We deleted a 230 kb large piece of DNA , from eight kb upstream Evx2 to a breakpoint located within the flanking centromeric gene desert ( del ( R1-R5 ) -d9Lac ) [54] ( Fig . 3B , C ) . While this deletion did not alter the HoxD gene cluster per se , it removed the border of the H3K27me3 domain and hence it reconstituted a neighborhood between heavily H3K27 tri-methylated nucleosomes and nucleosomes not methylated at all . The deletion of this ‘epigenetic border’ did not elicit any loss of H3K27me3 marks over the HoxD cluster in the developing brain , nor did it induce any leakage over the centromeric DNA from the gene desert ( Fig . 3B , C , del ( R1-R5 ) -d9Lac ) . Therefore , as previously reported [41] , the reconstitution of this artificial boundary between two chromatin domains with and without H3K27me3 marks did not lead to any spreading , at least not towards the centromeric end . We conclude that the recruitment of PRC2 is likely a sequence-specific process and that the spreading of its enzymatic activity may require some specific DNA features . The GC-content , which is unusually high within the cluster itself , while low in the sequences reconstituting the border ( Fig . 2B , C ) , may contribute to this process . Another mechanism to set the PcG epigenetic borders may involve transcripts encoded by the opposite DNA strand , a feature found in the HoxA , HoxC and HoxD clusters . In the case of HoxD , the Evx2 gene is found ca . 10 kb upstream Hoxd13 on the opposite strand . This gene , which is covered by H3K27me3 marks and locates close to the epigenetic border , was however not removed in the del ( R1-R5 ) -d9Lac deletion . Therefore , we analyzed a second deletion , ca . 260 kb large , with the same upstream breakpoint into the gene desert ( see above ) , but with a telomeric breakpoint located between Hoxd10 and Hoxd11 . In this del ( 11-R5 ) -d9Lac mutant , the entire posterior part of the HoxD cluster was removed including Hoxd11 , Hoxd12 and Hoxd13 as well as the Evx2 transcription unit and the epigenetic border . In this configuration , the H3K27me3 profile remained unchanged when compared to the wild type pattern . In particular , the reconstituted epigenetic boundary was similar to that seen with the shorter del ( R1-R5 ) -d9Lac deletion , suggesting that additional transcriptional units encoded by either DNA strands are not necessary for the recruitment of PRC2 at the extremity of the HoxD cluster , nor for the fixation of a sharp epigenetic boundary ( Fig . 3B , C; del ( 11-R5 ) -d9Lac ) . In both deletions , however , a Hoxd9/Lac transgene was relocated at the breakpoint , raising the possibility that transgenic sequences would interfere with PRC2 recruitment and hence we used a final mutant configuration carrying a ca . 800 kb large deletion including the HoxD centromeric regulatory landscape . This del ( Nsi-Atf2 ) deletion not only removes the 5′ epigenetic border , but also most of the regulatory elements that contact Hoxd genes and impose a chromatin topology to the locus [54] , [55] . The H3K27me3 profile observed in such mutant brains was as in wild-type animals ( Fig . 3B , C ) . Furthermore , ChIP-qPCR analyses revealed that the spreading of H3K27me3 from the HoxD cluster towards the new centromeric neighboring sequences did not exceed a 800 bp large interval , which corresponds to twice the average length of the sonicated DNA fragments ( data not shown ) . These results further indicated that the capacity to recruit PRC2 is restricted to Hox genes themselves , without any contribution from the surrounding genomic sequences . To demonstrate this point , we produced a transgenic line containing the Hoxd10 gene , which had inserted into a genomic region of average GC density and poor in H3K27me3 marks . While H3K27me3 was scored on the entire transgene , this histone modification did not spread over flanking nucleosomes , as assessed by ChIP-seq ( Fig . 3D ) . The Hoxd10 transgene was defined by the two loxP sites previously used for the deletion of this locus in vivo ( see above ) . Therefore , when this transgenic stock was crossed back into a mouse carrying a homozygous deletion of Hoxd10 ( TgN/del ( 10 ) −/− ) , the H3K27 trimethylation profile ( or the lack thereof- ) over Hoxd10 reflected that of the ectopic Hoxd10 copy . The Hoxd10 locus was selected because the CpG island located upstream the promoter ( CpG32 from UCSC ) could be removed by using FRT sites and the Flip recombinase in vivo , without affecting the transcription start site ( TSS ) . To make sure that no additional CpG islands remained after deletion of CpG32 , we deleted another potential short island ( CpG26 ) from our starting transgenic construct . Transgenic animals were crossed with a Cre-deleter strain to adjust copy number to one and various transgenes were thus crossed over Hoxd10 null mice to assess their H3K27me3 status in developing forebrains ( Fig . 4A , B and Fig . S3 ) . When the 9 kb long Hoxd10 locus containing a LacZ reporter cassette was used as a transgene , H3K27 trimethylation was almost undistinguishable from wild-type littermate brains , with a strong enrichment of H3K27me3 over the entire DNA fragment ( Fig . 4B , TgNd10Lac ) . Similar results were observed when the CpG26 sequences had been removed ( Fig . 4B , TgNd10 ) . Moreover , similar amounts of H3K27me3 were scored when the transcription start site of TgNd10 was deleted , suggesting that the recruitment of PRC2 may be independent of transcription ( Fig . 4B , TgNd10∂TSS ) [56] . Finally , when the second CpG island was excised , the H3K27me3 profile again remained unmodified , showing that CpG rich regions are dispensable for the initial recruitment of PRC2 , at least for this DNA segment and in this tissue ( Fig . 4B , TgNd10∂CpG ) . However , when a four kb large transgene containing only the 5′ sequence upstream Hoxd10 was used , H3K27me3 marks were no longer detected , even though this construct still contained an annotated CpG island ( Fig . 4B , TgNd10∂3 ) and was globally GC-rich . Of note , a transgene containing the same four kilobases together with exon 1 of Hoxd10 showed no recruitment of PRC2 either ( Fig . 4B , TgNd10∂PREd10 ) , regardless whether or not the TSS was present ( Fig . 4B , TgNd10∂TSS∂PREd10 ) , suggesting that neither the TSS , nor the CpG32 are essential for recruiting PRC2 in this configuration . Mapping the insertion sites did not reveal any correlation between the presence of H3K27me3 on the transgenes and their insertion into either a H3K27me3-rich or a GC-rich DNA region . In fact , transgenes were found integrated at least 500 kb away from H3K27me3-rich spots and into DNA segments with rather average GC contents ( data not shown ) . These experiments thus defined a 1 . 4 kb large DNA segment , containing exon 2 and the 3′UTR of Hoxd10 , which was necessary for the deposition of H3K27me3 marks . This DNA segments is referred to as PREd10 below . While H3K27me3 marks covering the Hox clusters are twice as dense in differentiated tissues than in ES cells ( Fig . S4 and [38] , [44] , [45] ) , the extent in coverage is identical , suggesting the implementation of the same mechanism . Consequently , we concentrated on pluripotent stem cells for further analyses of PREd10 . However , because the HoxD cluster is a target of polycomb repression in ES cells [38] , [44] , [45] , we derived induced pluripotent stem ( iPS ) cells from mice carrying a homozygous deletion of Hoxd10 to eliminate all endogenous signals . iPS cells are in principle indistinguishable from ES cells [57] , [58] ( Fig . 5A , B ) and our iPSdel ( Hoxd10 ) were thus used to assess the H3K27 methylation status of distinct electroporated DNA elements , overlapping with the deleted Hoxd10 DNA segment . Various portions of the TgNd10 transgene were first cloned between two homologous arms ( Env ) flanking the transgenes , in the hope of comparing random and targeted integration sites . However , homologous recombination events were not found . When either the entire Hoxd10 fragment , including the TSS and both exons , or the 1 . 4 kb long PREd10 were introduced into our iPSdel ( Hoxd10 ) cells , they became H3K27 tri-methylated , in agreement with the results obtained using classical transgenesis . More surprisingly , when the 5′ sequence corresponding to that used in TgNd10∂3 was assayed , H3K27me3 was detected too , in contrast to the results obtained in transgenic mice . We checked the capacity of either the vector backbone , or the PGK-neomycin gene promoter to recruit PRC2 , by electroporating the neomycin cassette alone . The PGK promoter is ubiquitous and hence neomycin transcripts were detected in all conditions tested ( Fig . S5 ) . The gene body did not show any enrichment in H3K27me3 , regardless whether cells were grown with or without G418 selection ( Fig . S5 ) . We next assessed whether PRC2 recruitment by the Env-d10∂3′ DNA fragment was enhanced by the presence of large concatemers of the transgene . We treated Env-d10∂3′ cells with a CRE-expressing lentiviral construct leading to the reduction of the concatemers to a single transgene copy , devoid of selection cassette . However , after proper excision of the supernumerary transgenes , the single copy was still able to capture PRC2 , even in the absence of PREd10 ( Fig . 5C , Env-d10∂3-CRE ) . We verified if this recruitment was influenced by the presence of the DNA homology arms included for a potential recombination at the locus , which contained sequences from both the Hoxd11 and a portion directly 3′ to Hoxd10 , which could thus initiate a ‘spreading’ of H3K27me3 marks over the Env-d10∂3 fragment . Accordingly , we electroporated d10∂3 ( 42% rich in GC ) into iPSdel ( Hoxd10 ) without any other surrounding DNA sequences . While H3K27me3 was detected over a multimerized version of d10∂3 , this mark was lost after the CRE recombinase had reduced copy number to one ( Fig . 5C , d10∂3-CRE ) . In contrast , when CpG-island free PREd10 ( 44% rich in GC ) was introduced into iPSdel ( Hoxd10 ) , H3K27me3 marks were readily scored after CRE-excision of the multimers ( Fig . 5C , d10PREd10-CRE ) . PRC1/2 subunits were also detected over this exogenous , randomly integrated sequence , suggesting it contains all proper information necessary for PcG recruitment to ectopic sites ( Fig . 5D ) . To further narrow down potential PRE's within PREd10 , we split PREd10 into two smaller fragments , PREd10-800 ( 44% rich in GC ) and PREd10-600 ( 43% rich in GC ) , which were tested as individual transgenes . Unexpectedly , both fragments were decorated by H3K27me3 , when introduced into iPSdel ( Hoxd10 ) cells as single copy ( Fig . 5C ) . This suggested that , as in Drosophila , mammalian polycomb recruiting elements can hardly be narrowed down to a unique sequence . Moreover , substantially less H3K27me3 was scored on either fragment , suggesting that these low interacting sequences may synergize to form a robust PRE .
In many bilaterian species , Hox genes are found in one or several genomic clusters , an organization tightly associated with the necessity for these genes to properly coordinate their transcriptional activation and maintenance . In particular , animals ( vertebrates or invertebrates ) displaying a temporal sequence in the establishment of their segmented body plan systematically show a complete clustering of their Hox gene complement , whereas other animals following different strategies ( such as cell lineages ) usually have broken Hox clusters or even Hox genes scattered throughout the genome ( refs in [59] ) . It was recently proposed that the temporal sequence in Hox gene activation was associated with the progressive removal of H3K27me3 marks [50] and that these marks helped maintaining silent genes into a repressive spatial compartment [60] , [61] . This configuration may be necessary to impose a tight repression over Hox genes until their proper time of transcriptional activation , to avoid their precocious activity leading to homeotic transformations . In this view , the activation of the Hox gene family may rely upon a progressive and directional removal of the Pc repressive activity , which may have helped to select for gene clusters with a high density of genes and concomitant start sites and GC islands , leading to a global re-enforcement and tightening of PRC2 recruitment . A high concentration of- and co-operativity between the sequences recruiting PRC2 may readily compensate for the lack of some of them , explaining why none of our deletion mutants in vivo elicited a visible re-organization of the H3K27me3 profile . Recent studies have proposed that stalled polymerase could be involved in PcG tethering [62] , a proposal which could apply to the reported D11 . 12 PRE ( 48% GC ) [31] , since it contains an alternative start site for HOXD11 . However , we show that transgenic constructs can lack H3K27me3 marks even though both the start sites and coding sequences are present , whereas other transgenes displayed H3K27me3 marks despite the absence of TSS . While it is possible that stalled PolII is still present at cryptic or shadowed sites , or that TSS present on some transgenes are not functional , our data do not favor the view whereby a TSS can work as a PREs . In this view , gene repression via PcG proteins likely relies on a number of regulatory mechanisms , rather than being solely due to transcriptional interference mechanism [63] , [64] , [65] , [66] , [67] , [68] . As for many Drosophila PREs , PREd10 overlaps with both a DNase hypersensitive site and a CTCF binding site . However , these hallmarks are present neither in the previously identified d11 . 12 , nor in the MafB/Kreisler PREs . Moreover , it is noteworthy that , although they are both bound by PRC1 and PRC2 , PREd10 and PRE d11 . 12 are neither bound by Jarid2 , nor by KDM2B , two proteins found in some PRC2 complexes to target them to appropriate loci [69] , [70] , [71] . It is possible that different PRE sequences throughout the HoxD cluster have different operational modes . Also , the presence of GC rich sequences , and more specifically their unmethylated form [70] , [71] , has been proposed as a pre-requisite to establish Pc-dependent repression due to the correlation between Polycomb group proteins and CpG islands ( at least 50% GC over 200 bp ) [44] , [45] . Moreover , bacterial DNA sequences with high GC density are sufficient for PRC tethering in embryonic stem cells [40] , [41] and two thirds of all PcG bound targets contain GC rich fragments , either in their promoters or in their gene bodies . Because of their unusually high concentration of genes , the Hox clusters are amongst the genomic loci with the highest GC content . Here again however , our results do not support a high GC content as the major parameter in recruiting PRC2 . We show that DNA segments with a GC content similar to the average of the mouse genome ( 42% ) are still able to properly recruit PcG proteins and the deletion of CpG islands from our transgenic constructs did not abrogate the trimethylation of H3K27 . While these results suggest that CpG islands are neither sufficient , nor required , for the tethering of PcG proteins in the context of Hox gene clusters , they do not rule out their potential importance for the spreading or the re-enforcement of the coverage by PRC2 ( see below ) . The existence of CpG islands devoid of PcG , as well as PcG target DNA devoid of CpG islands , such as in the case of the first described mammalian PRE-like sequence regulating the mouse MafB/Kreisler , support this view . Moreover , sequences unable to recruit PRC when present as single copy transgenes may become H3K27me3 when concatamerized , suggesting that larger stretches of GC-high sequences can artificially recruit PRC , a possible explanation to the discrepancies observed between our results and those of others [40] . Our data are in agreement with an ad minima model whereby H3K27me3 is deposited on a series of low affinity PRC2 interacting sequences , which work synergistically between themselves and together with GC-rich sequences to confer robust silencing over target genes ( Fig . 6 ) . The minimal number of such sequences required to elicit Pc-dependent silencing is unknown , as well as the mechanism underlying their cooperativity . To date , three such minimal sequences have been spotted within HoxD , including PRE d11 . 12 , PREd10 and a sequence within the construct we used as homologous arms . Each Hoxd gene locus may thus carry at least one such sequence . Once PRC2 tethered to low affinity PREs , the GC density may help strengthening the interaction between the repressive complexes and the surrounding DNA , either by stabilizing PRC2 or by recruiting PRC1 ( Fig . 6 ) . In agreement with this view , nucleosome density , a feature that correlates with high GC content [72] , is important for the maintenance of H3K27me3 , whereas PRC2 may be activated by these initial repressive marks via an allosteric modification [73] . Accordingly , any DNA segment , regardless of its GC content , would become H3K27 trimethylated , if introduced into the HoxD cluster , as is the neomycin cassette in the rel0neo+ allele ( Fig . 2C , D ) . Moreover , GC rich DNA segments introduced in the vicinity of a PRE could stabilize the association with the PcG complex and become H3K27 trimethylated , as in our different transgenic constructs . This view would also accommodate the absence of H3K27me3 spreading near the integration site of TgNd10Lac , as well as in the case of the deletion of the 5′ epigenetic border . In such Hox loci , where chromatin compaction seems to be enhanced whenever the cluster is inactive , potential cross-linking artifacts may give the impression of a dense and continuous coverage by H3K27me3 , whereas some regions could be devoid of PRC2 . While this may indeed slightly bias the results , it remains from our genetic analyses that a range of PRE-like sequences must exist scattered within the HoxD cluster , instead of a few strong PRC2 airports , from which an enzymatic activity would spread , either via the spreading of the enzyme , or due to conformational proximity .
All experiments involving animals were authorized and carried out according to the Swiss law on animal experimentation ( LPA; No 1008/3482/0 to DD ) . All stocks of mice were kept as heterozygous and bred to homozygosity . Lines were all described and can be found in previous publications of the Duboule laboratory . Two additional lines were produced by recombination between the loxP site in the second exon of Hoxd1 and either the site telomeric to Hoxd8 ( del ( 1-i ) ) or the site centromeric of to Hoxd4 ( del ( 1-4 ) ) . The del ( i ) line was produced by TAMERE between the loxP telomeric of Hoxd8 and the one centromeric to Hoxd4 . Genotyping was performed on individual yolk sacs . Chromatin immunoprecipitation followed by quantitative polymerase chain reaction was performed as described in [74] . Briefly , cells were pre-plated 45 minutes to ensure no contamination from feeder cells . Cells or tissues were fixed for 10 and 15 minutes , respectively , in 1% formaldehyde , washed three times in cold PBS and stored at −80° before being processed using polyclonal anti-H3K27me3 antibody ( Millipore , 17–622 ) or H3K4me3 ( Millipore 17–614 ) . ChIPped DNA was either hybridized to customized tiling arrays ( see customized tiling array ) or deep sequenced using the Illumina Genome Analyzer . Reads were mapped onto the mouse mm8/mm9 genome using Tophat and visualized with the integrative genome viewer and RChiV . Mouse embryonic fibroblasts were derived from heterozygous crosses of E13 . 5 embryos using standard protocols . Cells were cultured in standard MEF/ES cell culture conditions . MEF/ES media contained DMEM supplemented with 10% FBS and LIF ( ES media only ) . Isolated MEF lines were first genotyped using embryonic tissues and subsequently confirmed with DNA extraction procedures . Passage three MEFs were used for iPS derivation experiments . Human Oct4 , Klf4 and Sox2 were cloned and separated by bacterial 2A sequences , in a single lentiviral backbone ( 3F ) . Virus was produced in 293T cells using FuGENE HD transfection reagent ( Promega , E2311 ) and ultracentrifuged . Induced pluripotent ( iPS ) stem cells were derived following standard protocols [57] . Colonies were picked at d16–d18 and expanded before genotyping . Pluripotency of clones was confirmed by their ability to grow indefinitely , the expression of pluripotency markers ( SSEA1 , Nanog , Oct4 and Sox2 by immunohistochemistry and western blot , standard protocols ) , a non-aberrant chromosome count ( by chromosome spread , standard protocols ) and re-establishment of bivalent domains ( K27me3 and K4me3 , see ChIP ) . Electroporation of induced pluripotent stem cells was performed using an Amaxa Nucleofector I and the Lonza mouse embryonic stem cell kit ( Lonza VPH-1001 ) . Briefly , 25 µg of DNA were digested overnight , phenol-chloroform purified and resuspended in 10 µl H2O . Media was changed 4 hours before electroporation . Cells were washed twice with Mg ( 2+ ) -Ca ( 2+ ) -Free PBS , trypsinized and aliquoted to 2×106 . Electroporated cells were plated on 10 cm dishes coated with DR4 resistant feeders . G148 selection ( 200 µg/ml ) ( Sigma G8168-10ML ) was started 24 hours after electroporation and was continued until individual colonies were picked and genotyping . CRE treatment of iPS cells was done as follows: 3×105 cells were plated overnight and transduced with a PGK-CRE lentiviral construct at MOI 100 . Individual colonies were picked 5 days post-transduction , expanded and genotyped . Affymetrix custom-made tiling arrays covering two megabases surrounding the mouse HoxD cluster were spotted with 25-mer oligonucleotides at 15b bp resolution ( Genome Assembly 2006 NCBI36/mm8: chr2:73 , 709 , 304–75 , 470 , 233 ) . Fragmentation , labeling and hybridization of ChIPed DNA were done following standard protocols . Cells were first disrupted and homogenized using a Polytron ( kinematic ) before RNA was extracted using the RNeasy Microkit ( Qiagen , 74034 ) . qRT-PCR was performed with SYBR Green . Two biological replicates , processed in triplicates and normalized to a housekeeping gene ( Rps9 ) were used to derive mean values . Primers are given in Table S1 . Southern blotting was performed using standard protocols . Different probes were DIG-labeled using the PCR DIG Probe Synthesis kit ( Roche , 11 636 090 910 ) . Genomic integration mapping of transgenes/constructs was performed using the inverse PCR ( iPCR ) method from the Molecular Cloning Manual ( third edition ) . Briefly , DNA was digested , phenol-chloroform precipitated , self-ligation 4 hours at room temperature and ethanol-precipitated . A first round of PCR was done with 50 ng of template before proceeding to a second round of PCR , using nested primers . Finally , distinct amplicons were purified using the QIAquick Gel Extraction kit ( Qiagen , 28704 ) and sent for sequencing . Raw hybridization data was extracted using the two-sample comparison analysis and quantile normalized using Tiling Analysis Software ( TAS ) from Affymetrix . Data was exported as plain text using a log2 or −10log10 scale for the signal , respectively the p-value . Files were visualized in RChiV , an in-house developed genome browser , which takes into account the deleted segment and normalizes the signal using a sliding window approach .
|
Hox genes are essential for the proper organization of structures along the developing vertebrate body axis . These genes must be activated at a precise time and their premature transcription is deleterious to the organism . Early on , Hox gene clusters are covered by Polycomb Repressive protein Complexes ( PRCs ) , which help keep these genes silent . However , the mechanism ( s ) that selectively recruit PRCs to these particular genomic loci remains elusive . We have used a collection of mutant mice carrying a set of deletions inside and outside the HoxD cluster to try and detect the presence of any DNA sequence of particular importance in this mechanism . We conclude that a range of low affinity sequences synergize to recruit PRCs over the gene cluster , which makes this process very robust and resistant to genetic perturbations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
A Genetic Approach to the Recruitment of PRC2 at the HoxD Locus
|
Salmonella Typhimurium has evolved a complex functional interface with its host cell largely determined by two type III secretion systems ( T3SS ) , which through the delivery of bacterial effector proteins modulate a variety of cellular processes . We show here that Salmonella Typhimurium infection of epithelial cells results in a profound transcriptional reprogramming that changes over time . This response is triggered by Salmonella T3SS effector proteins , which stimulate unique signal transduction pathways leading to STAT3 activation . We found that the Salmonella-stimulated changes in host cell gene expression are required for the formation of its specialized vesicular compartment that is permissive for its intracellular replication . This study uncovers a cell-autonomous process required for Salmonella pathogenesis potentially opening up new avenues for the development of anti-infective strategies that target relevant host pathways .
Bacterial pathogens that have sustained long-standing associations with their hosts have developed complex functional interfaces shaped by the concerted activities of molecules from the host and the pathogen [1]–[4] . For many pathogens , this functional interface is largely dependent on the activity of specialized protein secretion machines known as type III secretion systems ( T3SSs ) , which deliver bacterial effector proteins into host cells to modulate a variety of cellular processes [1] , [5] . One example of such a pathogen is Salmonella enterica serovar Typhimurium ( S . Typhimurium ) , a cause of human gastroenteritis , which interacts with host cells through the activities of two T3SSs encoded within its pathogenicity islands 1 ( SPI-1 ) and 2 ( SPI-2 ) [6]–[8] . The SPI-1 T3SS mediates bacterial entry into non-phagocytic epithelial cells , while the SPI-2 T3SS is required for the building and maintenance of a specialized membranous compartment that harbors the intracellular bacteria . Bacterial internalization is mediated by the SPI-1 T3SS effectors SopE , SopE2 , and SopB , which activate the Rho family of GTPases Rac1 , Cdc42 and RhoG [9] , [10] . In addition these bacterial effectors stimulate a transcriptional reprogramming in host cells , which leads to the production of pro-inflammatory cytokines believed to be essential for the initiation of the inflammatory diarrhea that characterizes acute Salmonella infection [11] , [12] . The early transcriptional responses stimulated by Salmonella upon infection of intestinal epithelial cells exhibit many of the hallmarks of the responses seen after the stimulation of innate immune receptors [13] . However , the Salmonella-induced responses are unique in that this pathogen is capable of stimulating them independently of innate immune receptors [12] , which are largely inactive in intestinal epithelial cells due to robust negative regulatory mechanisms [14]–[16] . Indeed , inflammation is essential for Salmonella growth in the intestine since without inflammation this pathogen cannot gain access to essential nutrients [17] and cannot effectively compete with the normal microbial flora [18] . Therefore , despite exhibiting the fingerprints of an innate immune response , the early transcriptional responses stimulated by Salmonella can be best characterized as a pathogen-driven process triggered by specific adaptations to cope with the host environment rather than as a hard-wired host defense response to conserved bacterial products . Although there is ample evidence for a role of this pathogen-induced transcriptional reprogramming of epithelial cells in the initiation of the inflammatory response to Salmonella through the production of pro-inflammatory cytokines [19] , [20] , it is unknown whether these changes in gene expression influence other aspects of Salmonella biology . In fact , despite the widely demonstrated ability of many pathogens to stimulate transcriptional responses in infected cells [13] , there is surprisingly little evidence for a potential influence of these responses in cell autonomous processes that may affect intracellular pathogen biology . Here we have characterized the transcriptional responses of cultured epithelial cells stimulated by S . Typhimurium during its intracellular stage and have dissected the signaling pathways that lead to these responses . Our results demonstrate an important role for the pathogen-induced changes in host-cell gene expression in the establishment of an intracellular niche suitable for Salmonella replication . This study uncovers a previously unknown strategy utilized by an intracellular bacterial pathogen to promote its replication within host cells , which represents a remarkable example of pathogen modulation of host responses for its own benefit .
We have previously shown that cultured epithelial cells undergo a significant transcriptional reprogramming shortly after infection with S . Typhimurium that is strictly dependent on the function of its SPI-1 T3SS [11] , [12] . In the present study , we have examined the transcriptional profile of epithelial cells at a much later time point following infection with wild-type S . Typhimurium . We found that the majority of the genes whose expression increased 10 h after infection were not upregulated early in infection ( Fig . 1A , 1B and Table S1 ) . Twenty hours after infection , the transcriptional response was even more distinct , and only ∼25% of the upregulated genes at this time point were also upregulated early in infection ( Fig . 1A , 1B and Table S1 ) . Interestingly , while genes associated with transcriptional innate immune responses dominated the initial response to bacterial infection ( Table S1 ) [12] , later in infection genes associated with other biological processes were more prevalent . For example , genes associated with cell adhesion , G protein signaling , lipid metabolism , vesicle traffic , protease inhibition , and processes associated with the general maintenance of cell homeostasis were uniquely induced later in infection . These results indicate that later in infection there is a distinct transcriptional program that results in a pattern of gene expression that is very different from that observed early in infection . However , like the early transcriptional responses , stimulation of the responses later in infection was also strictly dependent on the SPI-1 T3SS ( Fig . 1B and Table S1 ) . This observation suggests that the unique features of the late transcriptional responses to Salmonella may be simply the result of the intrinsic dynamics of the responses triggered early in infection . In fact , early in infection S . Typhimurium induced the expression of a number of transcription factors ( e . g . FOS , FOSB , FOSL1 , JUN , JUNB , EGR1 , EGR4 , ATF3 , STAT3 ) ( Table S1 ) [12] , which could significantly modify and amplify the transcriptional responses later in infection . Alternatively , this pathogen may utilize specific mechanisms to further modulate host responses later in infection . To discriminate between these two possibilities we examined the transcriptional responses of cells infected with a S . Typhimurium Δasd mutant [21] . Although its levels of infection are indistinguishable from those of wild-type S . Typhimurium ( Fig . S1 ) [21] , the S . Typhimurium Δasd mutant is only able to survive and grow in the presence of L-diaminopimelic acid , an essential component of the cell wall that is absent in mammalian cells . Therefore , shortly after infection , the SPI-1 T3SS of the S . Typhimurium Δasd mutant becomes non-functional ( Fig . S1 ) and the bacteria mutant dies shortly after entering into mammalian cells ( Fig . S1 ) [21] . Cells infected with the Δasd mutant strain showed a gene expression profile that overlapped significantly with that observed in cells infected with wild type , particularly when considering genes whose expression as a consequence of infection changed the most ( Fig . S2 and Table S1 ) . These observations suggest that the pattern of gene expression observed in infected cells at later times after infection is defined in large measure by responses triggered early in infection . We examined in infected cells the levels of a subset of proteins ( EGR1 , c-FOS , TTP , SerpinB3 and Serpin B4 ) whose genes were most significantly upregulated after S . Typhimurium infection . Consistent with the transcriptional response , we found much higher levels of these proteins in infected cells ( Fig . 1C ) . However , we observed very significant kinetic differences in their expression . For example , although the levels of EGR1 and c-FOS were markedly increased shortly after infection , 10 h after infection these proteins were not detectable in lysates of Salmonella infected cells ( Fig . 1C ) . In contrast , expression of TTP was observed throughout bacterial infection , while SerpinB3 and SerpinB4 were not detected until 10 h after infection ( Fig . 1C ) , despite the fact that their mRNA levels were drastically increased shortly after Salmonella infection ( Table S1 ) . In all cases , and as predicted from the mRNA measurements , stimulation of expression of these proteins was strictly dependent on the presence of a functional SPI-1 T3SS ( Fig . 1C ) . These results indicated that at least for a subset of genes there are post-trancriptional regulatory mechanisms that further modulate the gene expression changes stimulated by Salmonella . Taken together , these results revealed a complex host cell response to Salmonella infection leading to profound changes in gene expression To gain insight into the signal transduction pathways triggered by Salmonella that result in the observed transcriptional responses we examined the profile of transcription-binding sites in the genes that were induced early and late in infection . Although binding sites for transcription factors known to be activated by MAP kinases or NF-κB signaling are prevalent among genes induced early after infection ( Table S2 ) , binding sites for signal transducer and activator of transcription 3 ( STAT3 ) [22] are the most highly represented among genes whose expression was increased late in infection ( Table S3 and Table S4 ) . We also examined potential relationships among the Salmonella-induced genes using the STRING database , which curates data from multiple sources including physical as well as functional association between proteins [23] . This analysis also placed STAT3 as a central interaction node among the genes whose expression was stimulated by Salmonella late in infection ( Fig . 2A ) . We therefore hypothesized that STAT3 may play a central role in the orchestration of the transcriptional response to Salmonella . Consistent with this hypothesis , we observed phosphorylation of STAT3 at both Y705 and S727 , a measure of its activation , shortly after S . Typhimurium infection of cultured epithelial cells ( Fig . 2B-2E and Figs . S3 and S4 ) . STAT3 phosphorylation was observed throughout infection , with continuously increasing levels until at least 20 h post-infection ( Fig . 2B and 2C ) . Consistent with its activation , immunofluorescence analysis detected phosphorylated STAT3 in the nucleus of ∼70% of Salmonella infected cells ( Fig . 2D and 2E ) while none was detected in uninfected cells . We then tested whether STAT3 was required for the transcriptional responses stimulated by S . Typhimurium infection . We examined the changes in mRNA levels of several genes whose expression had shown among the largest fold-change as a consequence of Salmonella infection ( Table S1 ) . We found that addition of the STAT3 inhibitor S31-201 [24] effectively blocked expression of the reporter genes in infected cells ( Fig . 2F ) . These results indicate that Salmonella induces transcriptional responses through the activation of STAT3 . It has previously been reported that flagellin is capable of activating STAT3 [25] . However , a S . Typhimurium mutant unable to produce flagellin was equally capable of stimulating STAT3 compared to wild type ( Fig . 3A ) . Instead , we found that STAT3 activation was strictly dependent on the SPI-1 T3SS system , since cells infected with a S . Typhimurium invA mutant , which is defective in this system [26] , showed no STAT3 activation ( Fig . 3B and Fig . S3 ) . More specifically , we found that , like the transcriptional responses [12] , STAT3 activation was specifically dependent on the SPI-1 T3SS effector proteins SopE , SopE2 and SopB ( Fig . 3C and Fig . S5 ) , which in a redundant manner activate Rho-family GTPases [10] , [27] . Although mutations in each one of these effectors individually did not result in a significant reduction in bacterially-induced STAT3 phosphorylation , the simultaneous removal of these three effectors completely abolished Salmonella's ability to activate STAT3 ( Fig . 3C and Fig . S5 ) . Since these three effector proteins also mediate bacterial internalization [28] , we tested whether the presence of intracellular bacteria per se was sufficient to stimulate STAT3 activation . We expressed the Yersinia pseudotuberculosis invasin protein , which mediates bacterial uptake via the α4-β1 integrin receptors [29] , in invasion-deficient S . Typhimurium mutant strains defective in the SPI-1 T3SS system ( i . e . ΔinvA ) , specifically lacking the SPI-1 T3SS effectors SopE , SopE2 and SopB , or lacking all the known effectors of this T3SS ( i . e . “effectorless mutant” ) . We found no STAT3 activation in cells infected with any of the S . Typhimurium mutant strains expressing invasin ( Fig . 3D ) despite the presence of equivalent numbers of intracellular bacteria ( Fig . S6 ) to those that in the case of wild-type Salmonella , were sufficient to stimulate STAT3 activation ( Fig . 2B and 2C ) . Taken together , these results indicate that S . Typhimurium stimulates the activation of STAT3 through specific signaling pathways that are dependent on the presence of the SPI-1 T3SS effector proteins SopE , SopE2 and SopB . STAT3 activation most often results from the binding of various ligands to specific cell surface receptors of the Janus tyrosine kinase ( JAK ) family , which activate different members of the STAT family of cytoplasmic transcription factors [30] . It is well documented that STAT3 in particular can be potently activated by secreted cytokines such as IL-6 [30] . In fact , infection of macrophages by S . Typhimurium , which results in the activation of Toll like receptors and the production of cytokines , leads to STAT3 activation [31] . We therefore tested whether the STAT3 activation we observed in epithelial cells in response to Salmonella infection was the result of an autocrine and/or paracrine signaling pathway . We treated uninfected cultured epithelial cells with supernatants obtained from S . Typhimurium infected cells at different times after infection and examined STAT3 phosphorylation in the treated cells . We found no detectable STAT3 phosphorylation in cells treated with any of the infected cell supernatants ( Fig . 4A and Fig . S7 ) indicating that STAT3 activation as a consequence of Salmonella infection is not likely to be the result of autocrine or paracrine pathways . These results suggested that the activation of STAT3 by Salmonella infection does not involve the canonical pathways dependent on the JAK tyrosine kinases . To test this hypothesis we examined the effect of a specific JAK tyrosine kinase inhibitor on Salmonella-induced STAT3 phosphorylation . We found that addition of the JAK inhibitor Tofacitinib ( CP-690550 ) [32] had no effect on S . Typhimurium-induced STAT3 phosphorylation , even when supplied at a concentration 100 fold higher than that required for its inhibitory activity after stimulation by the addition of cytokines to cells that are known to activate this kinase in response to cytokines ( Fig . 4B and Fig . S8 ) . These results indicate that the Salmonella-induced STAT3 activation may be the result of a non-canonical pathway involving a tyrosine kinase ( s ) other than members of the JAK family . We reasoned that such a kinase must act downstream of Rho-family GTPase signaling since activation of STAT3 by S . Typhimurium required the SPI-1 T3SS effector proteins SopE , SopE2 and SopB ( Fig . 3C ) , which exert their function by redundantly activating Rac1 , Cdc42 and RhoG [10] , [27] . A candidate to play this role is the Abelson tyrosine kinase ( c-Abl ) [33] , which can be activated by S . Typhimurium infection [34] in a SopE/SopE2/SopB-dependent fashion ( Fig . S9 ) , and can directly phosphorylate STAT3 [35] , [36] . We therefore tested whether Salmonella-induced STAT3 activation required this tyrosine kinase . We found that addition of imatinib , a specific inhibitor of c-Abl [37] , significantly inhibited Salmonella-induced STAT3 activation ( Fig . 4C ) . These results indicate that c-Abl is a component of the signaling cascade triggered by Salmonella that leads to STAT3 activation . Salmonella-induced transcriptional reprogramming requires the activation of Cdc42 and Rac1 by the SPI-1 T3SS effector proteins SopE , SopE2 and SopB [10] , [12] , indicating that a downstream effector ( s ) of these Rho-family GTPases must be involved in the Salmonella-induced activation of c-Abl . The p21-activated kinases ( PAKs ) are well-characterized Rac and Cdc42 effector proteins , which are involved in a variety of signaling pathways [38] . PAK2 has been shown to phosphorylate and activate c-Abl [39] . Furthermore , Salmonella infection has been shown to robustly activate PAK2 [40] . We therefore reasoned that members of the PAK family would be good candidates to link the signaling events triggered by the Salmonella SPI-1 T3SS effector proteins with STAT3 activation . We found that addition of IPA-3 , a specific inhibitor of PAK1/2/3 [41] , effectively blocked Salmonella-induced STAT3 phosphorylation ( Fig . 4D ) . Similar results were obtained after expressing a dominant-negative form of PAK3 ( Fig . S10 ) . These results indicate that members of the PAK family are the link between Rho family GTPase signaling stimulated by the Salmonella SPI-1 T3SS effectors and STAT3 activation , most likely through the activation of c-Abl . We then tested whether the PAK and Abl kinases were required for the Salmonella stimulated transcriptional responses . We found that addition of PAK and ( to a lesser extent ) Abl kinase inhibitors blocked the transcriptional responses stimulated by S . Typhimurium infection ( Fig . 4E ) . In contrast , addition of JAK or Src kinase inhibitors did not ( Fig . 4E ) . Taken together , these results indicate that S . Typhimurium triggers pathogen-specific signaling pathways that lead to STAT3 activation and the reprogramming of gene expression in infected cells . The delineation of the signaling pathways stimulated by Salmonella leading to changes in host-cell gene expression provided us with an opportunity to evaluate the potential influence of these responses in Salmonella biology . We investigated whether specific inhibitors of the signaling pathways leading to transcriptional responses altered the intracellular growth of this pathogen . S . Typhimurium replication and survival within cells have been correlated with its ability to form a specialized bacteria-containing membrane-bound compartment that is characterized by the presence of tubular-like membranous structures known as SIFs ( for Salmonella induced filaments ) that can be stained with the endosomal protein LAMP1 [42] . We found that addition of the specific STAT3 inhibitor S31-201 blocked the formation of SIFs ( Fig . 5A and 5B ) . Furthermore , addition of the STAT3 inhibitor ( Fig . 5C ) or RNAi-mediated depletion of STAT3 ( Fig . 5D and 5E ) significantly impaired S . Typhimuirum intracellular growth without affecting host cell survival ( Fig . S11 ) . We made the observation that , in epithelial cells , only a subset of Salmonella-infected cells undergo transcriptional reprogramming , as indicated by the ability of these cells to express SerpinB3 as well as other genes whose expression levels were shown by microarray measurements to be increased at later times after Salmonella infection ( see Text S1 ) . Although the in vivo relevance of this observation is unknown , it provided us with an opportunity to investigate potential differences in the biology of Salmonella when localized within cells that have ( SerpinB3-positive ) or have not ( SerpinB3-negative ) undergone changes in their gene-expression profile as a consequence of bacterial infection . We first investigated potential differences between the fitness of bacteria present within these populations of host cells . We used a S . Typhimurium strain that expresses the red fluorescence protein ( dsRed ) under the control of an arabinose-inducible promoter and measured fitness as the capacity of Salmonella to produce the red fluorescence protein upon addition of arabinose . We found that most Salmonella in SerpinB3-positive cells were able to express dsRed ( Fig . 5F ) . In contrast , a significantly higher proportion of bacteria in SerpinB3-negative cells were unable to activate dsRed expression ( Fig . 5F ) . Differences in dsRed expression could not be explained by differences in the accessibility of the inducer in SerpinB3-positive or SerpinB3-negative infected cells since the total ratio of dsRed+ vs . dsRed- bacteria was maintained in experiments in which the inducer was added to bacteria that had been released from cells ( Fig . S12 ) . We also found that the number of Salmonella associated with SIFs was significantly higher in cells expressing SerpinB3 than in cells that were not ( Fig . 5G ) . Consistent with the increased bacterial fitness and the increased number of SIFs , we found a significantly higher number of Salmonella within cells that have undergone gene expression changes ( i . e . SerpinB3 positive ) than within cells that have not ( i . e . SerpinB3-negative ) ( Fig . 5H and Fig . S13 ) . Taken together these results indicate that Salmonella triggers signaling events that lead to changes in host-cell gene expression and render the host cell more permissive for bacterial growth and survival .
It is well established that the interaction of microbial pathogens with mammalian cells often leads to significant changes in host-cell gene expression [13] . These responses are most often the result of the stimulation of innate immune receptors by conserved bacterial products , which lead to “hard-wired” transcriptional outputs . Salmonella , however , has evolved the additional ability to stimulate transcriptional responses independent of the activation of innate immune receptors [11] , [12] , [43] . This specific adaptation , which requires effectors of the SPI-1 T3SS , allows Salmonella to potentially trigger this response in cells , such as those of the intestinal epithelium , that are subject to robust regulatory mechanisms to prevent the stimulation of innate immune receptors [14]–[16] . In this study we have defined a host-cell signaling pathway leading to Salmonella-induced changes in gene expression in epithelial cells and found that STAT3 plays a central role in their orchestration . We have found that Salmonella activates STAT3 through a non-canonical pathway that does not require JAK kinases . Instead , this pathway is triggered by the SPI-1 T3SS effectors SopE , SopE2 and SopB , which through Rho-family GTPases , stimulate PAK and Abl tyrosine kinases leading to STAT3 activation ( Fig . 6 ) . Our examination of the host cell gene expression in Salmonella-infected cells has revealed a very complex pattern that changes significantly over time . However , our results suggest that the pattern of gene expression is largely established by mechanisms operating shortly after infection . In fact , infection of cells with a conditionally lethal mutant of Salmonella that dies shortly after infection led to a transcriptional response that showed significant overlap with that observed in cells infected with wild-type Salmonella . We therefore hypothesize that a significant proportion of the changes in the transcriptional response to Salmonella late infection is defined by the intrinsic dynamics of the responses triggered shortly after infection . In support of this premise , the expression of several transcription factors was seen elevated shortly after infection ( although not late in infection ) . It is therefore possible that the early stimulation of expression of these transcription factors may be central to the orchestration of changes in gene expression later in infection . The potential mechanism by which these transcription factors may influence the cellular response to Salmonella infection is unknown but it is unlikely to involve autocrine or paracrine mechanisms since we found that culture supernatants of infected cells did not stimulate transcriptional responses in non-infected cells . We speculate that at least some of the changes in gene expression stimulated by Salmonella are also under post-transcriptional regulatory control . For example , the SerpinB3 and SerpinB4 proteins were not detected until 10 h after infection despite the fact that we detected a very significant increase in their mRNA levels shortly after Salmonella infection . One of the proteins whose expression is induced and maintained throughout infection is the post-transcriptional regulator tristetraprolin ( TTP ) . TTP is an RNA-binding protein that controls gene expression by modulating mRNA decay of messages containing a consensus adenylate and uridylate-rich ( ARE ) element in the 3′-untranslated region [44] . The extent to which TTP modulates the gene expression changes resulting from bacterial infection is not known but it is intriguing to hypothesize that by accelerating the decay of certain messages TTP may help to shape the nature of the response stimulated by Salmonella late in infection . There is abundant evidence for a role of the transcriptional responses to Salmonella as well as to other pathogens in the stimulation of inflammation [19] , [20] . In fact , studies have shown that early responses to a variety of different pathogens often lead to the production of pro-inflammatory cytokines [13] . However , there is remarkably little evidence to support a role for pathogen-induced transcriptional responses to modulate cell autonomous processes that may actually help pathogen replication . We found that cells that have undergone changes in gene expression , support more bacterial growth and harbor fitter bacteria than cells that have not . We found that the formation of Salmonella-induced filaments ( SIFs ) , membranous structures harboring Salmonella that characterize the replication-competent intracellular compartment [42] , was more efficient in cells that have undergone changes in gene expression . The mechanisms by which changes in gene expression contribute to Salmonella replication and the formation of its intracellular niche are likely to be multifactorial and multigenic . Indeed among the genes whose expression was increased as a consequence of Salmonella infection there are several that could potentially contribute to the formation of Salmonella's intracellular niche such as genes involved in G protein signaling , lipid metabolism , vesicle traffic , or protease inhibition . However , more studies will be required to understand the mechanism by which the transcriptional re-programing aids the intracellular growth of Salmonella . We have described here a mechanism by which S . Typhimurium renders infected cells more permissive for its survival and replication by stimulating changes in gene expression through a pathogen-specific mechanism . The stimulated signaling mechanisms involve pathways for which suitable inhibitors are available and at advance stages of clinical development for a variety of applications [45]–[47] . Therefore these findings may provide the bases for the development of novel therapeutic strategies to combat Salmonella infection that would target relevant host pathways . The results described here constitute a remarkable example of how pathogens modulate cellular functions for their own benefit .
All bacterial strains used in this study are derived from S . Typhimurium strain SL1344 and are listed in Table S5 . For trans-complementation studies , expression plasmids derived from pBAD24 encoding sopE , sopE2 , or sopB were introduced into the effectorless strain SB1011 . S . Typhimurium was grown under conditions that increase expression of the SPI-1 T3SS [48] , and when required , 0 . 1% arabinose was added to the medium to induce the expression of genes under the control of the arabinose-inducible paraABC promoter . To grow the S . Typhimurium Δasd mutant , the growth media was supplemented with 50 µg/ml sodium diaminopimelic acid ( L-DAP , Sigma-Aldrich ) . Bacterial growth curves in Luria-Bertani ( LB ) broth supplemented with the different inhibitors used in this study were obtained simultaneously using a multi-well plate and a plate reader ( TECAN infinite M1000 ) . Dominant negative mPak3 has been previously described [40] . Antibodies and other reagents were purchased from the indicated companies: rabbit-anti-TTP and rabbit-anti-SerpinB3/B4 ( Abcam ) ; rabbit-anti-Salmonella O Group B Antiserum ( Becton Dickson ) ; rabbit-anti-Phospho-STAT3 ( Ser727 ) , rabbit-anti-Phospho-STAT3 ( Tyr705 ) , and rabbit-anti-EGR1 ( Cell Signaling Technology ) ; mouse-anti-LAMP1 clone H4A3 , ( Developmental Studies Hybridoma Bank ) ; rabbit-anti cFOS , mouse-anti-SerpinB3 , mouse-anti-SerpinB4 , and mouse-anti-STAT3 ( Santa Cruz Biotechnology ) ; rabbit-anti-actin and mouse-anti-tubulin ( Sigma-Aldrich ) ; secondary antibodies ( Molecular Probes ) ; 4′ , 6-diamidino-2-phenylindole ( DAPI ) and PP1 ( both Sigma-Aldrich ) ; S3I-201 ( Calbiochem ) ; Imatinib ( Enzo Life Sciences ) ; Tofacitinib ( SYN KINASE ) ; IPA-3 ( Santa Cruz Biotechnology ) , and λ-Phosphatase ( NEB ) . The human epithelial cell lines Henle-407 and HeLa , the human embryonic kidney epithelial cell line HEK-293T , and the mouse macrophage RAW cells were cultured in antibiotic free Dulbecco's Modified Eagle Medium ( DMEM , Gibco ) supplemented with 10% bovine calf ( Henle-407 ) or bovine fetal ( HeLa , HEK-293T and RAW ) sera . The human liver epithelial cell line HepG2 was cultured on antibiotic-free minimal essential medium ( MEM , Gibco ) supplemented with 10% FBS . For bacterial infections , cells at a confluency of 80% were washed with Hank's buffered salt solution ( HBSS ) and allowed to equilibrate in HBSS for 15 min at 37°C . Cells were then infected for 1 h with S . Typhimurium strains at the multiplicities of infection ( MOI ) indicated in the Figure legends . To determine ABL1 activation , 24 h serum starved HEK-293T cells were lifted and resuspended in 500 µl HBSS , gently rocked for 2 . 5 h in the incubator before infected in suspension . In the case of infection with the flagellar mutant , cells were centrifuged for 5 min at 2 , 000 rpm after addition of the bacteria to facilitate bacteria/host cell contact . Infected cells were then washed once with HBSS and were incubated for 2 h in DMEM containing 50 µg/ml gentamicin , washed again once with HBSS and further incubated in DMEM containing 10 µg/ml gentamicin for the indicated times . Bacterial internalization and survival within host cells was examined using a gentamicin protection assay as described before [49] . Briefly , cultured epithelial cells grown in a 12 well plate were infected for 1 h and incubated in the presence of gentamicin as described above . Cells were washed twice with HBSS and then lysed in 300 µl 0 . 1% Sodium Deoxycholate ( DOC ) in HBSS to release their bacterial content . Multiple dilutions in buffered saline with gelatine ( BSG ) were plated onto LB medium , which in case of the Δasd strain was supplemented with DAP , to determine the concentration of living bacteria by quantifying colony forming units ( c . f . u . ) . Statistical significance was calculated by a one-tail distributed paired Student's t-test . Resulting p-values of less than 0 . 05 were considered significantly different . Epithelial cells were pretreated for 1 h in HBSS containing the different inhibitors at the indicated concentrations . Cells were then infected for 1 h and further chased for the indicated periods always in the presence of the respective drugs before harvested either for Western blot analysis , immunofluorescence or qRT-PCR . None of the compounds exhibited any inhibitory effect on bacterial growth ( Fig . S14 ) . Henle-407 cells grown in a 6 well plate were pretreated with DMSO or S3I-201 for 1 h in HBSS , infected with S . Typhimurium for 1 h ( MOI = 5 ) , chased for 2 h in DMEM containing 50 µg/ml gentamicin and further incubated in DMEM containing 10 µg/ml gentamicin for another 6 h , always in the presence of drugs . Subsequently cells were trypsinized and fixed with 4% PFA/PBS for 15 min at RT . Apoptosis was detected in a TUNEL reaction using the “in situ Cell Death Detection Kit , Fluorescein” ( Roche ) according to the manufacture's protocol in combination with a fluorimetric analysis employing a FACSCalibur ( BD Biosciences ) . Epithelial cells were lysed in 2× SDS sample buffer . Proteins contained in equal volumes of the cell lysates were separated by SDS-PAGE and subsequently transferred to PVDF membranes . After blocking with 3% BSA ( phospho specific blots ) or 5% milk in Tris buffered saline ( TBS ) , membranes were probed with the respective primary and secondary antibodies in blocking solution supplemented with 0 . 02% SDS and 0 . 1% Tween 20 . The blots were analyzed using either the Odyssey LI-COR system together with the LI-COR Odyssey application software or visualized by enhanced chemiluminescence ( ECL ) . When indicated , samples were dephosphorylated by incubating them in the presence of 1 , 200 units λ-Phosphatase ( NEB ) for 2 hs at 30°C prior to SDS-PAGE and Western blot analysis . Endogenous STAT3 was silenced by using the pSUPER RNAi System ( oligoengine ) . Briefly , a 60 mer oligonucleotide ( 5′-GATCCCCGGCGTCCAGTTCACTACTATTCAAGAGATAGTAGTGAACTGGACGCCTTTTTC-3′ ) , including the STAT3 target site described elsewhere [50] , was cloned into the BglII and HindIII side of the pSUPER vector . Cells were transfected using Lipofectamine2000 ( Invitrogen ) , infected ( MOI = 2 . 5 ) 48 h or 72 h later with S . Typhimurium for 1 h and then incubated in gentamicin ( 50 µg/ml ) -supplemented DMEM . At the indicated time points cells were lysed with 0 . 1% DOC in HBSS to release intracellular bacteria as described above . Cell lysates were examined for the presence of c . f . u . and analyzed by western blot for the presence of endogenous STAT3 ( total and phospho-Y705 ) as well as tubulin as a loading control . Human epithelial cells grown on glass coverslips were infected with different S . Typhimurium strains as described above , washed once with HBSS and fixed in 4% PFA/PBS for 15 min at RT . For SerpinB3 staining , cells were permeabilized with 0 . 1% Triton X-100 in PBS for 2 min at RT , washed once with PBS and then incubated in blocking solution ( 1% BSA , 0 . 01% Triton X-100 in PBS ) for 20 min at RT . Staining of phosphorylated proteins was done overnight at 4°C in a wet chamber . For staining of Salmonella-induced filaments ( SIFs ) , cells on glass coverslips were directly treated with blocking solution ( 3% BSA , 0 . 1% Saponin , 50 mM NH4Cl in PBS ) for 20 min at RT . Subsequently , glass coverslips were incubated in the respective blocking solution containing primary antibodies against the protein of interest for 1 h at RT , washed 3 times in blocking solution and incubated another 30 min at RT in blocking solution containing a mixture of secondary antibodies coupled to Alexa dyes and DAPI for DNA stain . Finally , glass coverslips were washed twice with blocking solution , PBS and water before they were mounted on glass slides and examined by epifluorescence microscopy ( Nikon Diaphot ) or spinning disk confocal microscopy ( Improvision ) using a Nikon TE2000 microscope , a Hamamatsu EM-CCD digital camera and Volocity software ( Improvision ) . Infected epithelial cells on glass coverslips were fixed and immunostained for SerpinB3 , LPS ( to stain bacteria ) and DAPI ( to stain DNA ) as described above . Cells were randomly imaged using a combination of epifluorescence and bright field microscopy to simultaneously detect SerpinB3 production , presence of Salmonella and the outlines of the epithelial cell . These images were then analyzed with ImageJ software ( http://rsbweb . nih . gov/ij/ ) to quantify the proportion of SerpinB3-positive or negative Salmonella infected cells and to quantify the number of intracellular bacteria . SerpinB3-positive and negative cells were then categorized by the number of intracellular bacteria that they harbor . Bacterial fitness was evaluated by the ability of S . Typhimurium to produce the dsRed fluorescence protein . Henle-407 cells were infected as described above for 1 h with a strain of S . Typhimurium carrying a plasmid that encodes the dsRed gene whose expression is under the control of an arabinose-inducible promoter . After 17 h incubation in arabinose free DMEM , expression of dsRed was induced by addition of 0 . 1% arabinose to the culture medium for 3 h . Cells were fixed , and immunostained for SerpinB3 and LPS , and with DAPI for DNA . The proportion of bacterial cells expressing dsRed in SerpinB3-positive and SerpinB3-negative cells was calculated by evaluating the number of dsRed-positive and dsRed-negative bacteria in a minimum of 100 cells in randomly taken images obtained by epifluorescence microscope using ImageJ software . To control for the accessibility of arabinose in SerpinB-positive or negative cells , bacteria were released from infected cells after 17 h incubation in arabinose free DMEM by addition of 0 . 1% DOC , pelleted , and resuspended in HBSS or HBSS containing 0 . 1% arabinose . Bacteria suspensions were incubated for additional 3 h at 37°C , recovered by centrifugation , fixed in 4% PFA/PBS for 15 min at RT and immunostained with an antibody against LPS . The proportion of bacteria producing dsRed was determined by epifluorescence microscopy . At least 100 bacteria per condition were quantified . Formation of SIFs was evaluated by immunofluorescence microscopy as follows . Inhibitor-treated or untreated cultured epithelial cells were infected ( MOI = 10 ) for 1 h , fixed after the indicated periods of incubation in gentamicin containing DMEM , and immunostained for SIFs with a monoclonal antibody to LAMP1 , a marker for this compartment . When required , cells were also stained with an antibody to SerpinB3 . The number of infected cells exhibiting SIFs was determined by analyzing randomly taken images obtained by epifluorescence microscopy using the ImageJ software . At least 100 infected cells per condition were evaluated . Uninfected and infected Henle-407 cells were washed once with PBS and treated with trypsin/EDTA for 5 min at 37°C . Detached cells were collected in DMEM containing 5% bovine serum albumin ( BSA ) , centrifuged at 1000 rpm for 5 min , fixed in 1 ml RNase free PBS containing 4% PFA for 15 min at RT , transferred into a fresh 2 ml reaction tube and washed once with RNase free PBS . Cells were resuspended in 1 . 5 ml RNase free blocking buffer ( 3% BSA , 0 . 1% Saponin , 50 mM NH4Cl in PBS ) containing primary antibodies against SerpinB3 and were incubated for 1 h at 4°C . Cells were then washed three times with blocking buffer and were incubated for additional 30 min at 4°C in blocking buffer containing the secondary antibody coupled to Alexa-488 . Cells were washed again , resuspended in 3 ml 1% RNase free BSA in PBS , filtered through a 30 µm nylon mesh and finally subjected to FACS sorting . FACS sorting was done at the Yale Cell Sorting Core Facility using a BD FACSAria . Sorted cells were collected in RNase free PBS containing 2 ml 1% BSA in PBS . Total DNA and RNA from equal numbers of sorted Henle-407 was isolated using the “DNAeasy Blood and Tissue” kit ( QIAGEN ) , following the manufacture's description for cultured animal cells . Briefly , pelleted cells were resuspended in 20 µl proteinase K and 200 µl AL buffer ( QIAGEN ) and incubated for 10 min at 56°C . After addition of 200 µl ethanol , the lysate was loaded onto a “DNeasy Mini spin column” , bound nucleotides subsequently washed with 500 µl AW1 buffer and AW2 buffer ( QIAGEN ) and finally eluted with 100 µl RNase free water . Remainining contaminant DNA was removed by treatment with DNAse I ( Roche ) for 15 minutes . RNA isolation , in vitro transcription and quantitative real-time PCR was carried out as described elsewhere ( Bruno et . al . ) . Briefly , 24 h serum starved Henle-407 cells were infected with wild type S . Typhimurium for 1 h and chased for additional 3 h . Total RNA was isolated using TRIzol ( Invitrogen ) reagent according to the manufacture's protocol . RNA was then further purified using the “RNeasy Mini Kit” ( QIAGEN ) . Following DNAse treatment , RNA was transcribed using the iScript reverse transcriptase ( BIO RAD ) . Transcript levels were determined using gene specific primers sets ( Table S6 ) , that have been designed by PrimerBank ( http://pga . mgh . harvard . edu/primerbank/ ) , and the iCycler real time PCR machine ( BIO RAD ) . Total RNA was isolated from infected ( MOI = 30 ) , serum starved Henle-407 cells at the indicated time points as described for quantitative real-time PCR . Samples were submitted to the Yale University W . M . Keck facility , where further sample preparation and hybridization to the Affymetrix HG U133 Plus 2 . 0 gene arrays was performed , using an Affymetrix GeneChip Instrument System according to the manufacturer's recommendations . Imaging was done on an Affymetrix GeneChip scanner 3000 according to Affymetrix standard protocols ( GeneChip Expression Analysis Technical Manual , Affymetrix , 2004 ) , while the raw data was processed and normalized using “affy” package in Bioconductor 2 . 9 [51] . For each experimental group , fold changes in gene expression observed for each strain were calculated relative to the uninfected control .
|
Essential for the ability of Salmonella Typhimurium to cause disease is the function of a type III secretion system ( T3SS ) encoded within its pathogenicity island 1 ( SPI-1 ) , which through the delivery of bacterial effector proteins modulates a variety of cellular functions . This study reports that the infection of mammalian cells with Salmonella Typhimurium results in a profound reprogramming of gene expression that changes over time . The stimulation of this response requires the activity of a specific subset of bacterial T3SS effector proteins , which stimulate unique signal transduction pathways leading to STAT3 activation . We found that the Salmonella-stimulated changes in host cell gene expression are required for its intracellular replication . Targeting the mechanisms described in this study may lead to the development of novel anti-infective strategies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Salmonella Modulation of Host Cell Gene Expression Promotes Its Intracellular Growth
|
Modularity analysis offers a route to better understand the organization of cellular biochemical networks as well as to derive practically useful , simplified models of these complex systems . While there is general agreement regarding the qualitative properties of a biochemical module , there is no clear consensus on the quantitative criteria that may be used to systematically derive these modules . In this work , we investigate cyclical interactions as the defining characteristic of a biochemical module . We utilize a round trip distance metric , termed Shortest Retroactive Distance ( ShReD ) , to characterize the retroactive connectivity between any two reactions in a biochemical network and to group together network components that mutually influence each other . We evaluate the metric on two types of networks that feature feedback interactions: ( i ) epidermal growth factor receptor ( EGFR ) signaling and ( ii ) liver metabolism supporting drug transformation . For both networks , the ShReD partitions found hierarchically arranged modules that confirm biological intuition . In addition , the partitions also revealed modules that are less intuitive . In particular , ShReD-based partition of the metabolic network identified a ‘redox’ module that couples reactions of glucose , pyruvate , lipid and drug metabolism through shared production and consumption of NADPH . Our results suggest that retroactive interactions arising from feedback loops and metabolic cycles significantly contribute to the modularity of biochemical networks . For metabolic networks , cofactors play an important role as allosteric effectors that mediate the retroactive interactions .
Hierarchical modularity has emerged as an organizational principle of biochemical networks , where larger less cohesive clusters of network components ( e . g . metabolic enzymes or signaling molecules ) comprise functionally distinct sub-clusters [1] , [2] . For example , Ihmels and coworkers analyzed the co-expression patterns of metabolic genes in Saccharomyces cerevisiae to find coordinated regulation of individual pathways as well as higher-order functions such as biosynthesis and stress response that require multiple feeder pathways [3] . Hierarchical organization was also observed by Gutteridge and coworkers for metabolic regulatory networks , where hub metabolites regulating many enzymes connect to modules of spoke metabolites that are chemically similar and/or regulate functionally related enzymes [4] . In recent years , observations on modularity have prompted metabolic engineers and synthetic biologists to consider whole pathways , rather than individual genes , as modular building units for cellular design [5] . An emerging design rule is to assemble and express a coherent set of genes that encode the desired biochemical pathway along with regulatory mechanisms that modulate the activity of the pathway [6] . Modularity analysis also offers a route to build practically useful , simplified models of complex biological systems . The size and complexity of biochemical networks reconstructed from genome databases has greatly increased over the years [7] , [8] , [9] , rendering the estimation of kinetic or regulatory parameters either impractical or outright infeasible . In this regard , the modularity of a biochemical network should allow the system to be partitioned into minimally interdependent parts , enabling systematic derivation of coarse-grained , yet comprehensive models . Such coarse-grained models could greatly simplify the parameter estimation problem by substituting detailed reaction kinetics with less detailed module kinetics [10] . While there is general agreement that a biochemical module should represent a group of connected network components , and that the arrangement of modules in the network is hierarchical , there is less consensus on the criteria that should be used to systematically extract biologically meaningful modules [11] , . One recent argument was to focus on cyclical , or ‘retroactive , ’ interactions between network components , as opposed to simple connectivity [15] . Biochemical pathways operate with direction , where upstream components ( e . g . concentration of reactants ) influence downstream components ( e . g . concentration of products ) . In the case where a downstream component also influences an upstream component ( e . g . via a feedback regulatory mechanism ) , the two components participate in a cycle and thus interact retroactively . Placing such components into the same module reduces the interdependence between different modules , consistent with the intuitive definition of a biological module . Indeed , metabolic cycles and feedback loops have been shown to confer robustness [16] by isolating external perturbations and attenuating their propagation through the entire network [17] . In this paper , we extend the concept of retroactivity to account for cyclical interactions spanning distant parts of a biochemical network as exemplified by feedback loops of signaling and metabolic pathways . In earlier work [6] , retroactivity was only considered for interactions between nearest neighbors in a network . To investigate hierarchy , we adopted Newman's algorithm for community detection [18] to successively partition a network into modules containing cyclical interactions based on a round trip distance metric , which we call Shortest Retroactive Distance ( ShReD ) . Applied to test models of a signaling network [19] ( Figure 1 ) and a metabolic network ( Figure 2 ) , the ShReD-based partitions produced hierarchically arranged modules that confirm biological knowledge . In addition , the partitions also revealed modules that are less intuitive . For the metabolic network , we also examined the role of allosteric regulators and cofactors as network elements that determine the number of cyclical interactions and the hierarchical depth of modules .
To examine the effect of cyclical , i . e . retroactive , interactions on modularity , we compared the partitions of the EGFR signaling network obtained using Newman's connectivity ( Figure 3a ) and the ShReD metric ( Figure 3b ) . Several qualitative similarities between the two partitions are evident . In both partitions , modules that possess a large fraction of reactions from phosphatidylinositol polyphosphate ( PIP ) signaling coupled to either intracellular Ca2+ signaling ( CAS ) or small guanosine triphosphatase ( SGTP ) were identified . Quantitatively , both partitions reach a hierarchical depth of 6 and become more homogeneous closer to the terminal nodes of the partition tree . From the root to terminal nodes , the canonical group compositions of the modules ( represented by the pie colors ) trend toward a single , dominant group ( Figure 4a ) . At the terminal nodes ( height zero ) , the fraction of reactions in a module belonging to a single canonical group , on average , exceeds 80% for both Newman and ShReD partitions . There are also notable differences between the two partitions . While both partitions extract modules predominantly consisting of G-Protein coupled Receptor ( GPCR ) activation reactions , the ShReD partition identifies greater hierarchy stemming from those modules . In the Newman partition , there are several terminal leaf nodes that predominantly comprise Mitogen Activated Protein Kinase ( MAPK ) reactions . Analogous terminal nodes are not present in the ShReD partition . The ShReD partition yields a large terminal node consisting of 99 reactions ( Supplementary Figure S1b , ID: 22219 ) , whereas the largest terminal node of the Newman partition consists of 36 reactions ( Figure S1a , ID: 22202 ) . The largest terminal node in the Newman partition ( ID: 22202 ) predominantly comprises GPCR transactivation reactions , whereas the largest terminal node in the ShReD partition ( ID: 22219 ) comprises several signaling functions , including MAPK cascade , endocytosis , and cell cycle . Another notable difference is that while the average number of cycles in a module decreases with increasing depth for both partitions , a larger number of cycles are preserved in the ShReD partition at greater depths ( Figure 4b ) . We next compared the Newman ( Figure 5a ) and ShReD partitions ( Figure 5b ) for the liver metabolic network , complete with regulatory edges and cofactors . As was the case for the EGFR network , both partitions lead to modules that generally increase in homogeneity from the root node to the terminal nodes ( Figure 6 ) . However , unlike the EGFR network , the arrangement and compositions of the two partitions are drastically different ( Figure 5 ) . In contrast to the Newman partition , the ShReD partition generates modules with hierarchical depth , similar to the GPCR dominated modules of the EGFR network . In the case of the metabolic network , hierarchical depth was greatest for modules comprising reactions in and around glycolysis ( GLYCO ) . Moreover , the terminal node modules of the ShReD partition reach greater homogeneity compared to the Newman partition ( Figure 6 , Figure S2 ) . The impact of metabolic regulation on ShReD-based modularity was investigated by comparing the partitions for the metabolic network model with ( Figure 5b ) and without the allosteric interactions ( Figure 7a ) . The two models yield qualitatively similar hierarchical partitions with subtle differences in the placement of reactions into modules ( Dataset S1 ) . These differences include the placement of reactions coupled to the pyruvate kinase reaction , which is subject to a high degree of allosteric regulation relative to other reactions in the network . The quantitative impact of regulation is observed by comparing the number of ShReDs present in the network prior to the partition . At depth zero , there are approximately 250 additional ShReDs in the model with allosteric regulation compared to the model without regulation ( Figure 8a ) . However , there is no obvious difference in the number of ShReDs between the two models at greater depths . There is also no obvious difference in the average ShReD at most depths , with the exception of depth zero , where the average ShReD is approximately 7% shorter for the model with allosteric regulation compared to the model without regulation ( Figure 8b ) . We next assessed the impact of cofactors such as ATP , NADH , and NADPH on ShReD-based modularity by comparing the partition generated for the complete metabolic model ( Figure 5b ) to the partition for a partial model with regulatory edges , but lacking any interactions resulting from cofactors ( Figure 7b ) . Qualitatively , the partitions reveal similar canonical groupings . Both partitions identify modules predominantly characterized by glucose metabolism ( GLYCO ) and modules predominantly characterized by amino acid metabolism ( AA ) . Both partitions also group together reactions of the TCA cycle ( TCA ) , urea cycle ( UREA ) and pyruvate metabolism ( PYRU ) . A major difference between the two partitions involves the reactions of lipid metabolism ( LIPID ) and detoxification ( DETOX ) . For the complete model , the ShReD partition identifies a module consisting of reactions from LIPID , DETOX , GLYCO , and PYRU ( Figures 5b and 9b: ID: 15995 ) , whereas no analogous module is identified for the model without cofactors . The reactions of module 15995 either produce or consume NADPH to support detoxification and lipid synthesis ( Figure 9b ) . Quantitatively , the number of ShReDs trends lower when the cofactors are absent , with the largest difference observed at zero depth ( Figure 10a ) . Conversely , the average ShReD of a module is generally larger when the cofactors are absent , with the largest difference again observed at zero depth ( Figure 10b ) . At greater depths ( >3 ) , the average ShReD plateaus to a value between 2 and 3 edges for both models . For completeness sake , we compared the partitions based on ShReD with partitions based on local , or nearest neighbor , retroactivity . To obtain local retroactivity partitions , the size of cycles was restricted to two edges , effectively eliminating all retroactive paths involving non-neighboring vertices . Algorithmically , ShReDij was set to infinity , if ShReDij was greater than 2 . Biochemically , a locally retroactive interaction represented either a reversible reaction catalyzed by a single enzyme or two irreversible reactions with opposite stoichiometry . For all cases , including the EGFR signaling network as well as various versions of the hepatocyte metabolic network , partitions based on local retroactivity failed to generate any modules .
In this paper , we introduce the use of ShReD as a round trip distance metric , which can be combined with a partition algorithm ( adapted from Newman's earlier work on community detection ) to systematically identify biochemical reaction modules that feature cyclical interactions . The notion of grouping together network components based on “retroactivity” was first proposed by Saez-Rodriguez and coworkers , who hypothesized that a strictly downstream component should have little impact on the activity of an upstream component unless there is a feedback or retroactive relationship [15] . It has been suggested that such feedback relationships contribute to robustness with respect to external perturbation , notably in signal transduction networks [16] . The ShReD metric accounts for cyclical interactions that span multiple reaction steps , and thus significantly extends on the prior work on retroactivity , which focused on local interactions between neighboring components . Previously , ( shortest ) path lengths between network components have been used to identify reaction modules by clustering , but without consideration of directionality and retroactivity [20] . To evaluate the performance of ShReD as a module-detection metric , we performed two sets of comparisons . One set of comparisons involved the community detection algorithm presented by Newman , which also formed the basis for our partitioning algorithm . Newman's original algorithm partitioned based on connectivity , and favored the placement of a pair of network elements ( vertices in the graph representation ) into the same module if the number of connections between the two elements exceeded the expected ( e . g . average ) number of connections assuming an equivalent network with edges placed at random . The second set of comparisons involved the special case of local feedback loops or cycles arising from reversible reactions . The results of these comparisons were used to investigate how multi-step signaling loops or metabolic cycles , as opposed to conventional connectivity or reaction reversibility , contribute to the modular organization of biochemical networks . Applied to a model network of EGFR signaling , the ShReD-based partitions generated modules with a greater number of cyclical interactions across all depths compared to Newman's connectivity-based partitions ( Figure 4b ) , consistent with the premise of the ShReD metric . Our results suggest that the total number of cyclical interactions in a network or module at least partially dictates the hierarchical depth of ShReD-based partitions . The ShReD-based partitions of the EGFR model generated one large terminal module with 99 reactions ( Figure 3b , ID 22219 ) , which could not be further modularized due to the relatively small number of cycles ( a total of 10 ) in the module ( ∼0 . 5 ShReDs per reaction ) . In contrast , the GPCR dominated module ( ID 22203 ) has 167 cycles and 774 ShReDs connecting just 39 reactions ( ∼20 ShReDs per reaction ) , and can be further partitioned to generate 4 additional levels of hierarchy . In the case of the liver metabolic network , which has a substantially greater number of cycles ( arising from allosteric feedback loops ) compared to the EGFR signaling network , the difference between ShReD and Newman partitions is more dramatic . ShRed partitions again lead to greater hierarchy , reaching a depth of 7 , whereas Newman's partition only reaches a depth of 3 . The greater hierarchy achieved using the ShReD metric is significant , because the partition algorithm is essentially identical to Newman's algorithm , i . e . the only difference is the metric used to calculate the modularity score Q . For both metrics , a module is further partitioned only if the Q score is positive after the partition . Indeed , the scoring criterion based on the ShReD metric is actually more stringent , because the algorithm performs an additional test to ensure that the modules resulting from a partition each has at least one cycle . In this regard , the greater hierarchy generated by the ShReD ( as opposed to the partition algorithm ) gives credence to the metric for being able to identify hierarchical modules based on the preservation of cycles . The retroactive interactions captured by ShReD include not only reaction reversibility ( as in previous work [15] ) , but also cycles and feedback loops involving multiple reactions and allosteric effectors . Feedback loops resulting from allosteric regulation of an upstream enzyme by a downstream product represent an important regulatory motif that is common to biochemical networks . To examine the impact of feedback loops on modularity , ShReD partitions were obtained for the metabolic network with and without allosteric regulation . While qualitatively similar , the partitions differed in the placement of highly regulated reactions . For example , biochemistry textbooks generally associate pyruvate kinase ( PK ) with glycolysis , where the enzyme catalyzes the terminal step . The enzyme's activity is subject to allosteric regulation by several sugar phosphates produced upstream in glycolysis . On the other hand , the enzyme's product , pyruvate , is highly connected to the TCA cycle and amino acid pathways through anaplerosis and transamination reactions . When regulatory edges are absent , ShReD partitions place PK in one of the terminal leaf nodes along with reactions of lipid metabolism , pyruvate metabolism , the TCA cycle , oxidative phosphorylation and ketone body synthesis ( Figure 7a , ID: 15982 ) . When regulatory edges are present in the model , however , the partitions place PK in a module dominated by reactions of sugar metabolism ( Figure 5b , ID 15939 ) , consistent with textbook biochemistry . In this regard , the ShReD metric captures the impact of both stoichiometric connectivity and feedback regulation in determining modularity . As many of the allosteric regulators were energy currency metabolites , we also examined the partitions for a partial metabolic model that lacks these cofactors . The resulting network contains fewer ShReDs , presumably reflecting an overall decrease in the total number of paths . Compared to the complete model , the corresponding ShReDs ( connecting the same reaction vertices ) of the partial model are ∼30% longer , indicating that allosteric feedback and other cofactor-dependent interactions more tightly couple the reactions in the network . In the present study , abstracting the metabolic network as a reaction-centric graph greatly facilitated the inclusion of cofactors in the modularity analysis , identifying both intuitive and non-canonical groupings that could not be identified by removing interactions effected by cofactors . For example , not including the cofactors in the model would completely isolate the oxidative phosphorylation reactions and carbamoyl phosphate production reaction from the rest of the metabolic network as disconnected components . Including the cofactors allows these reactions to be placed into modules; for the complete metabolic model , these reactions are kept together at a height of 2 ( Dataset S1 ) . Another example of cofactor-dependent modularity involves the association of NADH and FADH2 oxidation with different reactions in and around the TCA cycle ( Figure 9a ) . The partitions place NADH oxidation into a module ( ID: 15984 ) that also contains isocitrate and alpha-keto glutarate dehydrogenases , which are NADH producing reactions in the TCA cycle . Similarly , FADH2 oxidation is placed in a module ( ID: 15985 ) containing succinate dehydrogenase , which reduces FAD+ to FADH2 . The coupling between TCA cycle reactions and oxidative phosphorylation is intuitive . However , the TCA cycle reactions are also highly connected to reactions in glutamate metabolism and β-oxidation , associations that may be subjectively less intuitive . In this light , ShReD partitions reflect an emphasis on cyclical interactions mediated by the cofactors . A third example of an intuitive , yet non-canonical grouping involves the drug transformation reactions . In the present study , the metabolic model included reactions that are induced by troglitazone , a hydrophobic anti-diabetic compound withdrawn from the market due to severe hepaotoxicity . Module 15995 illustrates the cyclical interactions coordinating reactions of several different canonical pathways , including glutathione , lipid , glucose , and pyruvate metabolism ( Figure 9b ) . A dominant characteristic ( exhibited by seven of the nine reactions ) of this module is the production and consumption of NADPH , again underscoring the significance of the cofactors in determining the modularity . To examine whether the influence of the cofactors reflected the relatively small size of the model network ( comprising ca . 150 reactions ) , we also applied the ShReD-based modularity analysis to a larger model of the human liver ( comprising ca . 2500 reactions ) [21] . The analysis again identified cofactor modules centered on NADH and NADPH consumption and production , similar to the smaller liver model ( Figure S3 , Dataset S2 ) . Many of the terminal modules for the larger model comprised reactions that were grouped into analogous modules for the smaller model , suggesting that the size of the model did not qualitatively alter the structural organization of the metabolic network . Quantitatively , the maximum hierarchical depth was greater for the larger network , increasing from 7 to 16 . The increased depth was presumably due to the greater detail of the HepatoNet1 model , which includes many additional pathways of amino acid , lipid and nucleotide metabolism . In conclusion , this paper presents a novel methodology for modularity analysis that enables hierarchical partitions of biochemical networks by preserving feedback loops and other cyclical interactions . To the best of our knowledge , the present study is the first to build a module detection method that focuses on cycles or feedback loops as the key structural feature . The present study is also the first to account for cofactors in modularity analysis , further emphasizing the role of pathway regulation in network modularity . Previously , studies on modularity have generally ignored cofactors , citing methodological challenges arising from having to place these highly connected hub metabolites into particular modules [20] , [22] . It should be noted that the current analysis , which does not weight the edges in calculating the ShReDs , implicitly assumes that all reactions in the network are equally engaged . Clearly , the levels of engagement can be expected to vary across different reactions , and should ideally be weighted appropriately , by using quantitative activity data such as metabolic flux . For example , a high glycolytic flux may confer a larger weight to edges representing PK regulation , which in turn may impact the overall modularity of the network . Moreover , cells subjected to different chemical or genetic perturbations will likely exhibit different flux dynamics , which would need to be reflected in the metric to obtain partitions that meaningfully analyze the modularity of a dynamic system such as the biological cell . A thorough examination of the role of reaction engagements in modularity analysis is beyond the scope of this study , and warrants further work in a future study .
A common way to model a biochemical network using a graph is to represent the components as vertices and their interactions as edges . In this study , the focus is on understanding the hierarchical and modular relationship among reactions , treating metabolites as shared resources among modules . We therefore use a directed graph with vertices representing reactions and edges indicating a directional interaction between the connected reactions . Edges are drawn between two reactions ( Figure 11a ) if the product of one reaction is either a reactant ( Figure 11b ) or allosteric effector of another reaction ( Figure 11c ) . For reversible reactions , reactant-product relationships are considered in both directions . We utilize round trip distance as a metric , which we call Shortest Retroactive Distance ( ShReD ) , to characterize the connectivity between two vertices that interact retroactively . A retroactive interaction exists between two vertices i and j , if and only if there is a directional path from vertex i to j and a return path from vertex j to i . The retroactive interaction represents a mechanism for mutual feedback , and thus expresses interdependence . The ShReD of vertices i and j ( ShReDij ) is the sum of the shortest path distance from node i to j and the shortest return path distance from node j to i . In the example network of Figure 12 , ShReD1 , 3 is 3 because there are two edges along the shortest path from R1 to R3 and there is one edge from R3 to R1 . There is another cycle connecting the two reaction vertices , which also involves R4 , R5 and R6 . This cycle , however , is not the ShReD , as its length of 6 exceeds the ShReD value of 3 . For a given network ( or sub-network ) a ShReD value is computed for every pair of vertices in the network ( or sub-network ) . To compute the ShReD values , we first calculated the shortest distances between all pairs of vertices using the Floyd-Warshall algorithm [23] . The resulting all-pairs shortest path matrix was then added to its own transpose to generate a symmetrical ShReD matrix . When there is no path or no return path between two vertices , the ShReD value between these two vertices is infinity . The ShReD between a node and itself is zero . For the example network in Figure 12 , the ShReD matrix is as follows: ( 1 ) Partitions were obtained by adapting Newman's community detection algorithm [18] , which was modified to generate partitions based on the ShReD metric , as opposed to simple connectivity . An overview of the algorithm flow is shown in Figure 13 . The initial step is to find the connected subnetworks in the parent network using a breadth-first traversal algorithm [24] , as it is possible that the parent network , represented as a reaction centric graph , may not be connected . For the search , the network is represented as an undirected graph , as we are interested in identifying the connectivity of vertices , regardless of direction . Each connected subnetwork is then partitioned into two daughter subnetworks to maximize a “modularity score” while ensuring that each subnetwork resulting from a partition retains at least one retroactive interaction , i . e . cycle . Applied recursively , the algorithm produces a hierarchical tree of binary partitions . In Newman's algorithm , the modularity score was computed as the difference between the actual and expected number of connections between two components . In this study , we computed the difference between the actual and expected ShReD to determine the modularity score . The expected ShReD between i and j , Pij , is computed as the arithmetic mean of the average of all non-zero and non-infinite ShReDs involving i and the average of all non-zero and non-infinite ShReDs involving j: ( 2 ) where Di and Dj are the number of non-zero and non-infinite ShReDs involving i and j respectively , and n is the total number of vertices in the network ( or sub-network ) . We define a ShReD-based modularity matrix , G , as follows: ( 3 ) The diagonal entries of G are set to zero , because both the expected and actual ShReD between a vertex and itself are zero . An entry Gij is also set to zero , if ShReDij is infinity . For the example network in Figure 12 , the average ShReD of R1 and R2 are both 4 . 8 . The expected ShReD between R1 and R2 , P12 , is thus 4 . 8 , and G12 is 1 . 8 . The full matrix G for the example network is shown below . The ShReD-based modularity matrix differs from Newman's connectivity-based modularity matrix , which does not take into account the direction of an interaction . ( 4 ) Defining the modularity score Q based on the ShReD-based modularity matrix G , we wish to find a vector s , which assigns each vertex in the network to one of the two partitioned sub-networks to maximize Q: ( 5 ) where si is an element of a vector s . Each si has a value of either −1 or 1 . An increase in Q is obtained in two cases: if Gij is positive and the vertices i and j are assigned to the same sub-network ( si = sj = 1 or si = sj = −1 ) , or if Gij is negative and the two vertices are assigned to different subnetworks ( si = 1 and sj = −1 or vice versa ) . The vector s maximizing Q can be found using spectral partitioning methods [25] as described by Newman [18] . The solution to the maximization problem can be approximated by the leading eigenvector of G . For our example network ( Figure 12 ) , the leading eigenvector of G ( Equation 4 ) is given by v = [−0 . 41 , −0 . 41 , −0 . 41 , 0 . 41 , 0 . 41 , 0 . 41 , 0 , 0] , from which s is approximated as s = [−1 , −1 , −1 , 1 , 1 , 1 , −1 , −1] . All non-positive entries , including zero , in the eigenvector are assigned the value −1 . This partition assigns R1 , R2 , R3 , R7 and R8 to one module , and R4 , R5 and R6 to the other module . The reactions in the first module are not fully connected , which gives rise to two disconnected components , one comprising R1 , R2 and R3 and the other comprising R7 and R8 . In this example , a single binary partition generated three separate modules , each consisting of a single cycle . In Newman's original community detection algorithm , partitioning of a subnetwork continues if the modularity score Q is greater than zero and the leading eigenvector s of the modularity matrix G has at least one positive and one negative element; otherwise the subnetwork is not further partitioned . The algorithm terminates if there is no subnetwork that can be further partitioned . In our algorithm , we modified the termination criterion to also check that there is a cycle in each subnetwork resulting from a partition operation . The check for a cycle was performed using an algorithm similar to topological sort [26] . For a given module abstracted as a directed graph , the number of incoming edges is computed for each vertex . A vertex with zero incoming edges is removed from the graph along with its outgoing edges . The number of incoming edges is then recalculated for the remaining vertices . The process repeats until there are no more vertices , in which case the graph has no cycles , or until there are no vertices with zero incoming edges , indicating the presence of a cycle . In our example , the Q score for the first partition is greater than zero ( Q = 43 . 2 ) and each resulting subnetwork contained at least one cycle . Thus , the partitioned subnetworks are accepted as modules and the algorithm continues by finding the connected subnetworks in each module . The module comprising R1 , R2 , R3 , R7 and R8 is not fully connected , and two subnetworks are found , one comprising R1 , R2 , and R3 and the other comprising R7 and R8 . Neither subnetwork can be further partitioned , as every element in the leading eigenvector of the corresponding modularity matrix has the same sign . Similarly , the module comprising R4 , R5 and R6 cannot be further partitioned , as every element in the leading eigenvector of the corresponding modularity matrix has the same sign , and the algorithm terminates . The partitioning results are reported in the form of a hierarchical tree annotated with several properties . Each module is represented as a pie chart , where the size of each slice is proportional to the fraction of reactions that belong to the corresponding , pre-assigned canonical ( textbook ) grouping . The homogeneity index of a module corresponds to the fraction occupied by the largest slice in the pie chart . The homogeneity index therefore ranges from 0 to 1 , where a larger number indicates greater homogeneity in terms of composition based on the canonical group assignments . The black lines connecting the nodes in the hierarchical tree represent ShReD-based partitions , whereas the red lines represent the formation of components from partitions that include disconnected components . The depth of a module is determined as the number of black edges traversed from the root node to the module . The height of a module is determined as the largest possible number of black edges traversed from the module to a terminal leaf node . The number of cycles within a module is used to compare the partitions obtained based on the ShReD and Newman's connectivity metrics . While standard algorithms exist for counting the number of cycles in a graph [27] , the run time is proportional to the number of ( non-unique ) cycles . The number of cycles may be exponential in the number of vertices , and renders cycle counting as computationally inefficient . The cycle count is thus reported up to 1 , 000 unique cycles . Any count above 1 , 000 is effectively reported as 1 , 000 . In addition to cycles , we also determined the number of non-infinite shortest retroactive paths in a module as well as the mean ShReD of the module . The mean ShReD of a module is calculated by averaging the corresponding non-infinite entries in the corresponding ShReD matrix . As case studies , we examined two types of biochemical networks that feature directed interactions and feedback loops .
|
Mathematical models are powerful tools to understand and predict the behavior of complex systems . However , the complexity presents many challenges in developing such models . In the case of a biological cell , a fully detailed and comprehensive model of a major function such as signaling and metabolism remains out of reach , due to the very large number of interdependent biochemical reactions that are required to carry out the function . In this regard , one practical approach is to develop simplified models that nevertheless preserve the essential features of the cell as a complex system by better understanding the chemical organization of the cell , or the layout of the biochemical network . In this work , we describe a computational method to systematically identify closely interacting groups of biochemical reactions by recognizing the modular hierarchy inherent in biochemical networks . We focus on cyclical interactions based on the rationale that reactions that mutually influence each other belong in the same group . We demonstrate our method on a signaling and metabolic network and show that the results confirm biological intuition as well as provide new insights into the coordination of biochemical pathways . Prospectively , our modularization method could be used to systematically derive simplified and practically useful models of complex biological networks .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biology",
"computational",
"biology",
"metabolic",
"networks"
] |
2011
|
Identification of Biochemical Network Modules Based on Shortest Retroactive Distances
|
RNA turnover plays an important role in both virulence and adaptation to stress in the Gram-positive human pathogen Staphylococcus aureus . However , the molecular players and mechanisms involved in these processes are poorly understood . Here , we explored the functions of S . aureus endoribonuclease III ( RNase III ) , a member of the ubiquitous family of double-strand-specific endoribonucleases . To define genomic transcripts that are bound and processed by RNase III , we performed deep sequencing on cDNA libraries generated from RNAs that were co-immunoprecipitated with wild-type RNase III or two different cleavage-defective mutant variants in vivo . Several newly identified RNase III targets were validated by independent experimental methods . We identified various classes of structured RNAs as RNase III substrates and demonstrated that this enzyme is involved in the maturation of rRNAs and tRNAs , regulates the turnover of mRNAs and non-coding RNAs , and autoregulates its synthesis by cleaving within the coding region of its own mRNA . Moreover , we identified a positive effect of RNase III on protein synthesis based on novel mechanisms . RNase III–mediated cleavage in the 5′ untranslated region ( 5′UTR ) enhanced the stability and translation of cspA mRNA , which encodes the major cold-shock protein . Furthermore , RNase III cleaved overlapping 5′UTRs of divergently transcribed genes to generate leaderless mRNAs , which constitutes a novel way to co-regulate neighboring genes . In agreement with recent findings , low abundance antisense RNAs covering 44% of the annotated genes were captured by co-immunoprecipitation with RNase III mutant proteins . Thus , in addition to gene regulation , RNase III is associated with RNA quality control of pervasive transcription . Overall , this study illustrates the complexity of post-transcriptional regulation mediated by RNase III .
Bacteria are highly adaptive organisms that are able to rapidly alter their gene expression in response to environmental changes . In addition to transcriptional control , regulation of RNA decay has emerged as a major pathway in fast adaptive processes . Changes in RNA turnover facilitate stress responses , growth phase transitions , and virulence factor production [1]–[3] . Over the past decades , the knowledge on key ribonucleases that act in processing and turnover of RNAs in Escherichia coli and Bacillus subtilis has increased considerably [1]–[3] . Degradation of mRNA can follow several pathways involving a combination of exo- and endoribonucleases , and differs substantially between Gram-negative and Gram-positive bacteria [3] , [4] . For instance , E . coli uses the single-strand-specific RNase E to catalyze the initial rate-limiting cleavage of a large number of mRNAs [1] , while mRNA decay in B . subtilis involves the action of the endoribonuclease RNase Y and the bi-functional RNases J1/J2 , which are endowed with 5′ exoribonuclease and endoribonuclease activities [5] , [6] . Among the endoribonucleases , ribonuclease III ( RNase III ) is a member of a highly conserved and universal family of double-stranded-RNA ( dsRNA ) -specific enzymes with essential roles in RNA processing and decay [1] , [3] , [7] . The discovery that RNase III-type enzymes generate eukaryotic microRNAs and short interfering RNAs has triggered interest in defining the mechanisms of action of this family [8] , [9] . Crystal structures of Aquifex aeolicus RNase III in complex with different dsRNAs indicated that this protein contains a long RNA-binding surface cleft denoted the catalytic valley [9] , [10] . Bacterial RNase III is a homodimer that forms a single processing center with each subunit contributing to the hydrolysis of one RNA strand . Each monomer contains four RNA binding motifs that make extensive contact with the ribose-phosphate of the dsRNA up to 10 base pairs from the cleavage site , while conserved acidic amino acids and Mg2+ are responsible for catalysis [9] , [11] . Biochemical studies have identified the determinants of the dsRNA substrate and RNase III that are required for substrate specificity and catalytic activity . RNase III cleavage produces RNA fragments with 5′-phosphate and 3′-hydroxyl termini and a two-nucleotide 3′-overhang [11]–[14] . Aside from the universal function of RNase III in the maturation of ribosomal RNAs [15] , E . coli RNase III plays a broad role in gene regulation . Not only does RNase III autoregulate its own synthesis [16] , it also contributes to regulation by small RNAs [17] , [18] . In addition , recent genomic analyses revealed that the absence of RNase III in E . coli [19] and B . subtilis [20] affects the abundance of numerous mRNAs and non-coding RNAs ( ncRNAs ) . Did the cellular functions and substrate specificity of the ubiquitous RNase III diverge in Gram-positive bacteria ? In Streptococcus pyogenes , RNase III was identified as an essential host factor for the prokaryotic CRISPR/Cas immunity system [21] . In B . subtilis , the rnc gene is essential suggesting that RNase III-dependent maturation of one or several critical mRNAs is required for protein synthesis [20] , [22] . In Staphylococccus aureus , an rnc mutant strain showed compromised virulence in a murine peritonitis model [23] , while rnc deletion did not impair cell growth [23] , [24] . Our previous studies in S . aureus have shown that RNase III coordinates the repression of mRNAs encoding virulence factors and a transcriptional regulator via the quorum-sensing-dependent regulatory RNA , RNAIII [24]–[26] . The RNAIII-target mRNA complexes adopt various topologies , such as imperfect duplexes and loop-loop interactions that are efficiently recognized and cleaved by RNase III , thus leading to irreversible repression [27] . In addition , a very recent study has shown an unprecedented role of RNase III in antisense regulation restricted to Gram-positive bacteria [28] . Deep sequencing of short S . aureus RNAs revealed numerous 22-nt RNA fragments generated by RNase III digestion of sense/antisense RNAs and almost 75% of the cleaved mRNAs had corresponding antisense RNAs [28] . These data are indicative of pervasive antisense regulation by RNase III . Collectively , the previous studies in E . coli [19] , B . subtilis [20] , and S . aureus [28] evaluated the role of RNase III at a genome-wide scale . However , these analyses of transcriptome changes by tiling array or RNA-seq are not per se suitable to identify direct RNase III substrates because they also score indirect regulatory effects . This prompted us to more precisely analyze the functions and direct targets of S . aureus RNase III in gene regulation . We present here the first global map of direct RNase III targets in S . aureus . To this end , we used deep sequencing to identify RNAs associated with epitope-tagged wild-type RNase III and two catalytically impaired but binding-competent mutant proteins . Newly identified RNase III targets were validated by a combination of in vivo and in vitro approaches . Our analysis revealed an unexpected variety of structured RNA transcripts as novel RNase III substrates . In addition to rRNA operon maturation , autoregulation of rnc mRNA decay , degradation of structured RNA transcripts , and antisense regulation , we propose novel mechanisms by which RNase III activates the translation of mRNAs through cis- or trans-acting elements . Overall , our study explores the broad function of RNase III in gene regulation of S . aureus .
Biochemical and structural studies performed on RNase III in A . aeolicus and E . coli demonstrated a stepwise hydrolysis mechanism of the phosphodiester bonds mediated by two Mg2+ ions , involving mutual conformational changes of the RNA and the enzyme [10] , [29] . The nuclease domain of RNase III is characterized by two clusters of conserved acidic amino acids , in which the side chains of E41 , D45 , D114 , and E117 ( in E . coli ) are coordinated to Mg2+ ions [13] , [30] , [31] . Two of these residues , E117 and D45 , are essential for catalysis , as their substitution by alanine strongly compromised cleavage without affecting RNA binding [11] , [14] , [29] , [31] . Although S . aureus RNase III ( Sa-RNase III ) shares only 33% amino acid identity with the E . coli enzyme , the acidic amino acids are strictly conserved ( Figure 1A ) . To obtain catalytically inactive but binding-proficient variants of Sa-RNase III , amino acids E135 and D63 ( corresponding to E117 and D45 in E . coli , respectively ) were changed to alanine ( Figure 1A ) . A histidine epitope tag was added to the N-terminus of the mutant and wild-type ( WT ) proteins , and the proteins were purified to homogeneity following expression in E . coli [27] . The activities of the mutant enzymes were compared to that of the WT protein using spa mRNA , a well-characterized Sa-RNase III substrate [24] . Terminally labeled spa mRNA was used to map the cleavage sites for WT and mutant S . aureus proteins and cleaved products were resolved by polyacrylamide gel electrophoresis under denaturing conditions ( Figure 1B ) . As expected , WT RNase III cleaved both sides of a helix at U70 , C98 , and G110 in the coding sequence ( CDS ) of spa mRNA . The E135A mutation very strongly compromised the activity of the enzyme , while the effect of D63A was less pronounced ( Figure 1B ) . Gel retardation assays were used to monitor the binding of the mutant enzymes to terminally labeled spa mRNA in a buffer containing Ca2+ instead of Mg2+ ( Figure 1C ) . Ca2+ inhibits the catalytic activity of E . coli RNase III but does not affect RNA binding [32] . In our study , the mutant E135A ( Figure 1C ) and D63A ( result not shown ) enzymes bound spa mRNA in a manner similar to the WT RNase III . Hence , the two mutations uncoupled catalytic activity from RNA binding capacity in a manner similar to that described for E . coli RNase III [11] , [14] , [29] . These two mutant proteins were used to capture RNA substrates in vivo . To identify RNase III targets in vivo , co-immunoprecipitation ( coIP ) assays were carried out with Flag epitope-tagged WT and mutant proteins expressed from a plasmid-borne Cd2+-inducible promoter in a Δrnc S . aureus background . The Flag-epitope was added to the C-terminus of the proteins . A control coIP experiment was performed with the untagged WT RNase III expressed from the chromosome ( strain RN6390 ) . Bacteria were harvested at two time points ( 4 and 6 h of growth at 37°C ) corresponding to exponential and late exponential phases of growth , respectively . The growth curves of RN6390 ( WT strain ) , Δrnc mutant strain , and Δrnc complemented with WT RNase III , E135A , or D63A mutant proteins , were similar in the BHI medium ( data not shown ) . Western blot analysis showed that the two mutant proteins accumulated at comparable levels while the WT protein was expressed at a lower level ( Figure 1D ) , indicative of a possible autoregulatory event on the rnc mRNA . RNAs isolated from the coIP experiments with the four strains were converted to cDNA libraries and analyzed by high-throughput pyrosequencing as previously described [33] . Recovered sequences ranged from 1 to 145 nt , but sequences below 18 nt were discarded in later analyses to increase the accuracy of mapping ( Table S1 ) . In agreement with the impaired catalytic activity of the mutants , we obtained more reads of ≥18 bp with the E135A ( 77% of the total reads ) and D63A ( 86% ) mutant proteins than with WT RNase III ( 51% ) . Mapping of the cDNA reads to the genome of S . aureus N315 revealed that the mutant enzymes primarily recovered RNA fragments arising from rRNA and tRNA operons ( 80 to 90% of the total number of mapped reads ) ( Table S2 ) . However , a high number of reads were mapped to 58 different ncRNAs including the housekeeping ncRNAs tmRNA , RNase P , and the RNA component ( 4 . 5S RNA ) of the signal recognition particle ( SRP ) ( Tables S2 , S3 ) . Furthermore , in the coIPs of the mutant enzymes , reads were recovered for almost 1 , 500 individual mRNAs of the 2 , 653 annotated ORFs in the S . aureus genome , but considerably fewer were recovered for polycistronic mRNAs ( mnhA-G , gapR , pdhA-D , and qox; Table S4 ) . Moreover , a significant number of reads corresponded to antisense RNAs ( asRNAs ) that were assigned to the opposite strand of 1 , 175 mRNAs ( Tables S1 , S5 ) . Given the limited number of sequenced cDNAs ( ∼1×105 ) ( Table S1 ) , the actual number of RNase III targets may be underestimated . Based on comparison to sequences from the control coIP ( WT strain expressing untagged RNase III ) , which should represent RNAs that are unspecifically bound during the coIP , we only considered transcripts as potential RNase III substrates if they were significantly enriched in the coIP samples of the tagged variants ( Tables S2 , S3 , S4 , S5 ) . Not surprisingly , the RNAs that were unspecifically bound in the control coIP reflect the abundance of the transcripts in the cell , i . e . , most of these reads were derived from rRNAs that represent >90% of the transcripts in the cell ( Table S2 ) . We note that the co-immunoprecipitated RNAs identified with the E135A and D63A mutant proteins were very similar , supporting the reliability and reproducibility of the method ( Tables S2 , S3 , S4 , S5 ) . Moreover , many of the target RNAs were detected at both time points of cell growth . Representatives of each RNA class were then selected for experimental validation using in vitro and in vivo approaches ( Table 1 ) . We first performed gel retardation assays to validate a direct interaction between various classes of RNAs and the mutant E135A protein . The data showed that the E135A protein bound to a plethora of structured RNAs including cis-acting regulatory elements of mRNAs ( e . g . , the flavin mononucleotide ( FMN ) sensing riboswitch ) , ncRNAs , structured mRNAs , and small ORF-containing RNAs ( Figure S1A ) . Competition binding assays were also performed to monitor the specificity of RNase III binding on cspA mRNA . Two forms of cspA mRNAs were analyzed: cspAL containing a long 5′UTR ( 113 nt ) , which was recovered with the two mutant proteins and cspAS containing a short 5′UTR ( 52 nt ) , which was pulled down with the WT enzyme . We also used SA2097 mRNA , which was not co-immunoprecipitated with the mutant and WT RNase III . The experiments were carried out with the 5′ end-labeled cspAL mRNA bound to E135A mutant protein in the presence of increasing concentrations of cold cspAL , cspAS or SA2097 mRNA ( Figure S1B ) . The experiments showed that the concentrations of cspAS and SA2097 necessary to compete for binding were 10 times higher than that for cspAL suggesting that the interaction of RNase III with cspAL is specific . Overall , the data strongly suggest that the immunoprecipitated RNAs resulted from a direct interaction with RNase III . The molecular mechanism of RNase III action on several target RNAs was then studied in more detail both in vivo and in vitro ( Table 1 ) . A high number of reads were mapped to the five rRNA operons and several isolated tRNA operons . The most highly enriched RNA fragments , pulled down with the mutant proteins , corresponded to the intergenic regions of the rRNA operons ( Figure 2A ) . This strongly suggests a role of RNase III in rRNA and tRNA processing as it was previously demonstrated in E . coli [34] and B . subtilis [22] . We probed one of the five rRNA operons in WT and mutant strains using antisense oligonucleotides complementary to different tRNA and rRNA intergenic sequences ( Figure 2B ) . As expected in the case of impaired rRNA processing , 16S precursor transcripts were observed in the Δrnc strain and in the same strain complemented with either the E135A or D63A mutant enzyme , but not in the RN6390 ( WT ) strain or in the Δrnc strain complemented with WT RNase III . In addition , aberrant precursors from 5S rRNA and tRNAs were visible on Northern blots probed with a specific DIG-labeled riboprobe or the 5′ end-labeled oligonucleotide 278 , respectively , in Δrnc cells and in the same strain expressing the mutant E135A protein ( Figure 2B ) . Secondary structure analysis of the rRNA operon transcripts predicted that the termini of 16S rRNA and 23S rRNA might each base-pair within long helical domains , generating a typical RNase III substrate ( Figure 2D ) . RNase III cleavage assays were performed on an in vitro transcribed 16S rRNA containing the 5′ and 3′ end trailing sequences ( see Text S1 ) . Cleavage sites were identified by primer extension on the cleaved rRNA with reverse transcriptase using either the 5′ end-labeled oligonucleotide 405 ( Figure 2C ) or the 5′ end-labeled oligonucleotide 279 ( result not shown ) . Two specific RNase III cuts were identified at positions A-92 ( Figure 2C ) and U+64 ( result not shown ) , respectively . These cleavages produced a two-nucleotide 3′ overhang , a hallmark of processing by RNase III ( Figure 2D ) . Primer extension was also performed on total RNA extracted from the WT and Δrnc strains . Using the 5′ end-labeled oligonucleotide 405 that hybridizes within the 16S rRNA , a major reverse transcriptase ( RT ) stop at A-91 within the 5′ trailer of the 16S rRNA precursor was only seen in the WT strain . Thus , the in vivo and in vitro RNase III cleavages within the 16S rRNA precursor were congruent ( Figure 2C ) . Interestingly , in the Δrnc strain , several RT stops were detected upstream of A-92 in vivo ( Figure 2C , lane 6 ) , suggesting that another ribonuclease might target the same region in the absence of RNase III . Such alternative rRNA processing that permits the production of functional ribosomes tentatively explains why the Δrnc mutation in S . aureus has only minor effects on cell viability [23] , [24] . Reads , mapping to rnc mRNA , were consistently recovered with both the WT and the two RNase III mutants but not in the control coIP ( Figure 3A ) . These data suggest that RNase III of S . aureus specifically recognizes its own mRNA . This hypothesis is supported by the Western blot of the E135A , D63A , and WT proteins expressed in the Δrnc strain because the two mutant proteins accumulated to higher levels than the fully catalytically active WT enzyme ( Figure 1D ) . Prior to mapping the RNase III cleavage site , we determined the 5′ end of the rnc mRNA in vivo by primer extension ( Figure S2A ) . Two major reverse transcriptase ( RT ) stops were found , one located at G+306 in the coding sequence and the other >70 nt upstream of the AUG start codon ( Figure S2A ) . Several weaker stops were also observed , e . g . at position U+296 , after longer exposure of the autoradiography ( Figure S2A ) . Given the location in the CDS , the RT stop at G+306 represented an internal cleavage of rnc mRNA . We then mapped the RNase III cleavage sites on in vitro synthesized full-length rnc mRNA ( 843 nt ) and a truncated version ( 752 nt ) in which a large part of the 5′UTR had been deleted ( Figure 3B ) . The unlabeled RNAs were subjected to RNase III hydrolysis , and the RNA fragments were separated on agarose gels under denaturing conditions followed by staining with ethidium bromide ( for experimental details , see Text S1 ) . RNase III specifically cleaved the in vitro transcribed rnc mRNA and generated at least two main fragments in a Mg2+-dependent manner ( Figure 3B ) . Removal of the 5′UTR altered migration of the smaller fragment , identifying this fragment as 5′ proximal . This result suggests that RNase III recognizes and cleaves its own CDS . The RNase III cleavage sites were then mapped more precisely by reverse transcription in vitro ( Figure 3C ) . This experiment showed that position U+296 , located within the CDS of rnc mRNA , is the site of the major RNase III-dependent cleavage . Notably , this cleavage coincided with the 5′ end of the RNA fragment recovered by coIP with WT RNase III ( Figure 3A ) . It is surprising , however , that the primer extension performed on total RNA identified a potential RNase III-dependent cleavage at G+306 of rnc mRNA , 10 nucleotides downstream of U+296 ( Figure 3C and 3D ) . Although we do not exclude that RNase III cleaves its own mRNA differently in vivo , additional trimming of the cleaved RNA by an unknown ribonuclease could tentatively explain this difference . Structure probing of the rnc mRNA was performed using the single-strand-specific RNases T2 and T1 , and the double-strand-specific RNase V1 ( Figure S2B and S2C ) . Enzymatic reactions were restricted to less than one cut per molecule , and cleavages were mapped by reverse transcription [35] . The structure probing supported the formation of three long hairpins in the CDS , as indicated by numerous RNase V1 cleavages located in the arms and strong RNase T2/T1 cuts occurring in the apical loops ( I , II , and III ) and the internal loop regions ( Figure S2C ) . The long irregular helix III , in which the RNase III cleavage site at U+296 is located , appears to be the preferred RNase III binding site ( Figure 3D ) . Taken together , the data support a model wherein RNase III initiates decay of its own mRNA within the CDS , resulting in negative feedback regulation of its expression . The deep sequencing analysis additionally revealed several RNA fragments that were antisense to rnc mRNA ( Figure 3A ) . However , expression of these asRNAs was not detectable by Northern blot experiments in the RN6390 strain , indicating a very low abundance and/or low stability of these transcripts . The cspA mRNA , which encodes the major cold-shock protein and RNA chaperone , was a candidate RNase III substrate because the entire transcript was represented by reads from the coIP with the E135A mutant protein ( Figure 4A; Table 1 and Table S4 ) . To validate this target , we used Northern blots to first compare cspA expression in RN6390 and the Δrnc strain , in the presence or absence of RNase III WT and mutant proteins ( Figure 4B ) . Surprisingly , the absence of RNase III ( Δrnc strain ) led to the accumulation of a longer cspA mRNA ( cspAL ) than that observed in the WT strain ( Figure 4B ) . While complementation of the Δrnc strain by functional RNase III partially restored the WT pattern , the two mutant variants did not ( Figure 4B ) . Thus , RNase III appeared to process the cspA transcript into a shorter form ( cspAS ) . Northern blot analysis was then performed on RNA samples collected throughout growth from WT or Δrnc strains at 37°C , after cold-shock at 15°C ( at t0 , Figure 4C ) , and after shifting cultures back to 37°C ( at t3 , Figure 4C ) . Under all of these conditions , cspAL mRNA only accumulated in the Δrnc strain , suggesting that the maturation is not regulated by cold-shock but rather is a step in the normal biogenesis of cspA mRNA . Next , we performed primer extension on total RNA extracts for a comparative mapping of the 5′ end of cspA mRNA in WT and Δrnc strains ( Figure 4D ) . The 5′ end of the processed cspAS transcript ( WT strain ) mapped to U-52 , while that of the unprocessed cspAL mRNA ( in Δrnc ) was found 60 nucleotides upstream , at U-113 ( Figure 4D ) . Importantly , the 5′ end of the cspAL transcript precisely matched the 5′ boundary of RNA fragments recovered in the coIPs with the two mutant enzymes ( Figure 4A ) . Thus , the comparison of WT and mutant enzymes pinpointed an RNase III-mediated processing event . We then precisely mapped the RNase III cleavage sites on an in vitro synthesized unlabeled cspAL by reverse transcription ( Figure 4E ) , and by using 5′ end-labeled cspAL ( Figure 5D ) . RNase III hydrolysis of unlabeled cspAL followed by primer extension , revealed a major cleavage site at G-53 and a minor one at A-88 ( Figure 4E ) , generating the characteristic two-nucleotide 3′ overhang ( Figure 4F ) . The cleavage at position A-88 was also clearly detected with the 5′ end-labeled cspAL ( Figure 5D ) . Importantly , cleavage at G-53 matched the 5′ termini of cspAS in vivo ( Figure 4D ) . Hence , the RNase III cleavage assay in vitro faithfully recapitulated a major step of cspAL mRNA processing in vivo . Having confirmed that RNase III processing occurs within the 5′UTR of cspA mRNA , we set out to define the functional consequences of this event . The secondary structures of cspAL and cspAS were compared using single-strand-specific RNases ( RNases T2 and T1 ) and the double-strand-specific RNase V1 on in vitro synthesized mRNAs ( Figure S3A ) . The enzymatic cleavages were mapped by primer extension ( for experimental details , see Text S1 ) . The derived secondary structure model supports that cspAL mRNA is highly structured and starts at the 5′ end with several unpaired nucleotides followed by an almost perfect 32-bp helix ( Figure 4F and Figure S3B ) . This long 5′ hairpin structure resembles a typical RNase III binding site . Shortening of the 5′UTR led to the formation of a smaller but stable 5′ hairpin structure in cspAS ( Figure 4F , inset ) . Paired nucleotides at the 5′ end of mRNAs are known to protect against pyrophosphate removal by RppH and degradation by the 5′-3′ exoribonuclease activity of RNase J1 in B . subtilis [3] , [36] . To evaluate the effect of the short stable 5′ hairpin on transcript decay , we analyzed the in vivo RNA stability of cspA mRNA by Northern blot experiments after rifampicin treatment ( Figure 5A ) . Quantification of the data showed that the processing significantly stabilized cspA , increasing transcript half-life from <2 . 5 min in the Δrnc strain to >16 min in the WT strain ( Figure 5A ) . We then used toeprinting assays to monitor the formation of translation initiation complexes comprised of S . aureus 30S subunits , initiator tRNAfMet and cspA mRNA variants ( for experiment details , see Text S1 ) . The experiment showed that ∼50% of ternary initiation complexes were formed at 30S concentrations of 120 nM with cspAS and of 300 nM with cspAL ( Figure 5B ) . Thus , cspAS formed initiation complexes more readily than cspAL . Similarly , a differential proteomic analysis based on two-dimensional gel electrophoresis of cytoplasmic proteins prepared from WT and Δrnc bacteria showed that the synthesis of CspA protein was strongly reduced in the absence of RNase III-mediated processing ( data not shown ) . The increased initiation complex formation of the processed cspAS mRNA most likely reflects higher accessibility of the RBS as suggested by the enzymatic structure probing of cspAS mRNA . Indeed , single-strand-specific RNase cleavages were significantly enhanced in the region encompassing the SD sequence in cspAS ( Figure S3 ) . Thus , in the WT strain , the RNase III processing event in the 5′UTR of cspAL stabilizes the mRNA and facilitates ribosome binding to increase CspA synthesis ( Figure 5C ) . How the long 5′ hairpin of cspAL hampers ribosome binding remains to be studied . Interestingly , previous work showed that a stable hairpin structure located several nucleotides upstream of a SD sequence sterically interfered with translational initiation [37] . The deep sequencing data indicated the existence of an asRNA complementary to the entire 5′UTR of cspAL including the six first codons ( Figure 4A ) . Northern blot and primer extension experiments confirmed the presence of this asRNA in both WT and Δrnc strains grown at 37°C ( Figure 5D; Table S5 ) . However , the Northern experiments performed with DIG-labeled riboprobes , covering the same region of the genome , suggested that the yield of this asRNA was very low compared to that of cspA mRNA ( Figure 5D ) . We tested whether this asRNA guides RNase III cleavage of cspA . End-labeled cspAL mRNA was subjected to RNase III hydrolysis in vitro , in the absence or presence of the asRNA ( Figure 5D ) . Two conditions were used to form the asRNA-mRNA complexes: both RNAs were either denatured together and directly hybridized ( denaturing conditions ) , or were denatured and refolded separately before hybridization ( native conditions ) . After RNase III hydrolysis , the labeled RNA fragments were separated on a sequencing gel ( Figure 5D , lower panel ) . The results show that RNase III efficiently cleaved preformed mRNA-asRNA duplexes into short RNA fragments in vitro . Therefore , the asRNA suppressed rather than promoted the generation of stable cspAS mRNA . This regulation could contribute to fine-tuning of mRNA levels in vivo [28] . Overall , the RNase III-mediated processing step in the biogenesis of cspA mRNA is determined by the intrinsic structural properties of its 5′UTR alone . These data strongly suggest that RNase III cleavage activates the synthesis of the major cold-shock protein CspA at the post-transcriptional level . In addition to several mRNAs , the abundant housekeeping RNAs , tmRNA , RNase P , 4 . 5S RNAs , and the transcriptional regulator 6S RNA , were significantly enriched in the coIPs with the mutant proteins ( Table S2 ) . These ncRNAs are all processed from precursor transcripts by the concerted action of several endo- and exoribonucleases ( e . g . , [38] , [39] ) . However , the frequent recovery of such abundant and highly structured RNAs does not strictly imply their maturation by RNase III . For example , although B . subtilis 4 . 5S RNA maturation involves RNase III [39] , [40] , we did neither observe an altered processing pattern or precursor accumulation in the Δrnc mutant strains in Northern blot experiments ( Figure S4A ) , nor did we detect RNase III-dependent cleavages of 4 . 5S RNA in vitro ( results not shown ) . Likewise , the mature 230 nt product of 6S RNA was recovered by coIP , and its irregular hairpin structure was recognized by the RNase III mutant E135A ( Figure S1 ) . Nevertheless , we failed to observe RNase III-dependent processing on Northern blots ( Figure S4A ) and in vitro cleavage assays ( results not shown ) . As an aside , the 6S gene is located downstream of the aspS-hisS operon , which is controlled by a T-Box motif [41] , [42] ) . Whether 6S RNA expression responds to decreased pools of amino acids or uncharged tRNAs remains to be investigated . Many of the enriched RNA fragments recovered by coIP ( listed in Table S3 ) originated from full-length and bona fide ncRNAs of S . aureus , such as RsaA , C , E , H , I , and J [43] , [44] , the pathogenicity island-encoded ncRNAs SprA , SprA3 , SprB , SprC , and SprF3/SprG3 [42] , [44] , as well as RNAIII [45] . Several of these ncRNAs ( RsaA , RsaE , RsaX29/X39 , RsaI , RsaO , SprA ) were enriched with the mutant proteins suggesting that they are substrates of RNase III ( Figure S5 ) . These RNAs carry stable stem-loop structures and typical Rho-independent terminator hairpins ( Figure 6 , Figure S6 ) [46] . Experimental validation was performed on RsaA ( Figure 6 ) . In addition to RsaA , a second larger RNA ( RsaAL ) was detected on Northern blots , which likely originated from read-through at the transcriptional terminator . RsaA and RsaAL share a similar 5′ end as determined by RACE experiments [43] . Half-live measurements revealed a significantly higher RsaA stability in the Δrnc strain ( >60 min ) compared to WT strain ( ∼25 min for RsaA; Figure 6A ) . The longer RsaAL RNA was also more stable in the Δrnc strain ( 10 min ) than in the WT strain ( ∼2 . 5 min; Figure 6A ) . We also performed RNase III cleavage assays on in vitro transcribed RsaA followed by primer extension . Two main cleavages were identified in the bulged loop of the 5′ hairpin structure of RsaA ( Figure 6B ) and of RsaAL ( data not shown ) . Notably , these RNase III-specific cleavages coincided with the 5′ ends of two RNA fragments recovered by coIP with WT RNase III ( Figure S5 ) . Thus , RNase III contributes to the turnover of RsaA and RsaAL . This study identified novel ncRNAs such as RsaL , RsaN , and RsaO ( Table S3; Figures S4B and S6 ) . Northern blot analyses showed that RsaO was expressed in all strains tested ( Figure S4B ) , while RsaN was only detectable in the Δrnc strain ( Table 1 and Table S3 , data not shown ) . Other novel transcripts mapped to loci with multiple copies in the genome . For instance , two of the transcripts with the most abundant sequence reads corresponded to homologous and redundant ncRNAs ( RsaX29 and RsaX39 ) that originated from a partial duplication of the 5S rRNA genes . RsaX29 harbors a long helical structure that might be recognized by RNase III ( Figure S6; Table S3 ) . According to the deep sequencing data , several ncRNAs have associated asRNAs . The abundance of these antisense transcripts varied considerably according to Northern blot experiments ( Figure S4C and S4D ) . For instance , the putative asRNAs of RsaA or RsaH were solely detectable by deep sequencing ( results not shown ) . Conversely , several sense-antisense RNA pairs ( SAS028/teg102 , SprF3/SprG3 ) gave strong signals on Northern blots in WT and Δrnc strains ( Figure S4C , S4D ) . Teg102 has been previously identified as an asRNA complementary to SAS028 mRNA , which encodes a small hypothetical protein [44] , [47] . Its 5′ half was found in two copies in the same intergenic region of the genome ( Figure S4C ) . The levels of SAS028 mRNA were reproducibly lower in the Δrnc strain overexpressing the WT RNase III ( Figure S4C ) . It remains to be seen whether this RNase III-dependent effect is a consequence of asRNA regulation . SprF3/SprG3 , whose partial sequences are present in multiple copies in the genome [42] , may belong to the group I toxin-antitoxin systems , with SprG being the putative toxin [48] . Whether SprG3 encodes a peptide is yet unknown . Measurement of the half-lives in vivo showed that SprG3 ( >60 min ) is more stable than SprF3 ( <12 min ) ( Figure S4D ) . However , under the conditions of growth used in the experiment , the in vivo half-lives and the steady-state levels of SprF3 and SprG3 RNAs were similar in the WT and Δrnc strains ( Figure S4D ) even though RNase III efficiently cleaves the duplex formed in vitro ( data not shown ) . These surprising results are reminiscent of a recent study of a B . subtilis class I toxin ( bsrG ) -antitoxin ( SR4 ) system , which showed that the half-lives of bsrG and SR4 RNAs were increased only by 2-fold in a rnc mutant [49] . Deep sequencing of RNase III-associated RNAs recovered several mRNAs that encode proteins of various functions , including regulatory proteins that control the expression of virulence factors ( repressor of toxin Rot , transcriptional regulatory protein SarH , two component-system SrrA-SrrB ) , bona fide virulence factors ( protein A , the exotoxin Geh ) , and enzymes involved in various metabolic pathways ( Table S4 ) . In many cases , certain mRNA fragments were strongly enriched . This observation might be due to fragmentation occurring during the purification procedure , or alternatively reflect RNase III binding to structured mRNA fragments as a step in promoting their subsequent degradation . Many coIP mRNA fragments contained long hairpin structures which are typical RNase III binding sites , as it is observed for secY mRNA ( Figure 7A ) . In vitro RNase III cleavage assays were performed on in vitro transcribed and unlabeled secY mRNA followed by reverse transcription . Two cuts were located in a long hairpin structure within the CDS of one of the coIP fragments , generating typical RNA fragments with a two-nucleotide 3′ overhang ( Figure 7A ) . Many mRNA fragments corresponded to highly structured 5′UTRs of mRNAs , e . g . , ndrl and ptsG ( Figure S7 ) . These 5′UTRs were described as cis-acting regulatory elements of downstream genes with functions in the translational machinery or metabolic pathways ( Table 1 and Table S4 ) . They contain specific binding sites for diverse ligands such as metabolites , deacetylated tRNAs , or regulatory proteins ( ribosomal proteins , antitermination regulatory proteins ) [41] , [46] , [50] . A shared characteristic of most of these structured leaders is the presence of a long Rho-independent terminator structure , indicating that these RNA transcripts resulted from premature transcription termination ( Figure S7 ) . Other structured regions in the data sets corresponded to 3′UTRs of mRNAs that all carried stable Rho-independent terminators spanning at least one helical turn , i . e . the minimal substrate of E . coli RNase III [51] ( Table 1 and Table S4; Figure S7 ) . Several of these 3′UTRs are rather long ( >100 nts ) and two of them ( RsaM , RsaL ) correspond to ncRNAs ( Table 1 and Table S3 ) [44] , [47] . Overall , these examples illustrate that RNase III might affect the turnover of structured mRNAs , in addition to that of its own transcript and the cspA mRNA . The coIP strategy using two catalytically impaired RNase III mutant proteins facilitated the identification of asRNAs opposite to 44% of the annotated mRNA genes ( Table S5 ) . These asRNAs generally seem to be expressed at a very low level , or are rapidly degraded , since many of them were undetectable on Northern blots ( Table 1 ) . One example is hu mRNA and its asRNA ( Figure 7B , 7C ) . The stability of hu mRNA was measured in vivo in WT and Δrnc strains after rifampicin treatment ( Figure 7B ) . Quantification of the data showed that RNase III moderately affected the half-life of hu mRNA ( Figure 7B ) . Northern blot experiments performed with DIG-labeled riboprobes , covering identical region of the genome , suggested that the levels of the asRNA were significantly below that of hu mRNA ( Figure 7C ) . Moreover , in the Δrnc strain complemented with the WT RNase III , the signal of the asRNA was weaker than in the same strain complemented with the mutant enzymes ( Figure 7C ) . To evaluate whether the asRNA can induce mRNA processing , RNase III cleavage assays were performed on in vitro synthesized and 5′ end-labeled full-length hu mRNA either free or bound to the asRNA . The cleaved products were resolved on sequencing gels . While the free hu mRNA was not efficiently cleaved by RNase III in vitro ( Figure 7C ) , the pre-formed asRNA-hu mRNA duplexes were strongly cleaved into short RNA fragments ( Figure 7C ) . Thus , hu mRNA may be subject to rapid degradation by the combined action of the asRNA and RNase III . Several sense-antisense transcript pairs that were strongly enriched by coIP with the mutant proteins corresponded to overlapping UTRs of divergent genes , as illustrated with pdf1/SA0943 and tagG/tagH mRNAs ( Figure 8A; Tables S4 and S5 ) . While pdf1 encodes the essential peptide deformylase , the tagG/tagH genes encode the ABC transporter complex TagGH involved in the export of teichoic acids . Northern blot analysis was performed using specific labeled riboprobes complementary to the 5′UTR of SA0943 or to the CDS of pdf1 ( Figure 8A ) . In addition to full-length SA0943 mRNA , we observed a weak but reproducible signal for a ∼350 nt long RNA fragment that was only detected in the Δrnc strain ( Figure 8A ) . In contrast , among the three pdf1 mRNA species , the longest mRNA accumulated strongly in the Δrnc strain ( Figure 8A ) . A very similar pattern was observed for the tagG/tagH mRNAs . An RNA probe complementary to tagG mRNA detected three forms of the mRNA , the longest of which strongly accumulated in the Δrnc strains expressing the mutant proteins ( Figure 8A ) . Concomitantly , an RNA fragment ( <300 nts ) corresponding to the 5′UTR of tagG was detected in Δrnc cells , suggesting an additional RNase cleavage event . Mapping of the 5′ ends of the tagG/tagH mRNAs by primer extension and RACE confirmed that both mRNAs were processed by a mechanism that is partly dependent on RNase III ( Figure 8B , Table S6 ) . For tagG mRNA , RT stops mapped to positions −140 and −250 in both the RN6390 and Δrnc strains and to position −77 in the Δrnc strain . For tagH mRNA , two main RT stops were mapped at −25 and −279 in both strains , while the RT stop at −160 was only observed in the RN6390 WT strain ( Figure 8B; Table 1 and Table S6 ) . To assess a functional importance of the observed RNase processing , we further analyzed the RNase III cleavages on in vitro transcribed tagG/tagH mRNAs containing the long and overlapping 5′UTRs . Using the 5′ end-labeled oligonucleotide 410 complementary to tagH mRNA for primer extension , we observed short RNA fragments that were generated by RNase III hydrolysis only when tagG associated with tagH ( Figure 8C ) . This processing resulted in the formation of a tagH mRNA with a shortened leader whose 5′-end lies several nucleotides upstream of the SD sequence ( Figure 8D ) . Thus , RNase III likely targets the 5′ overlapping regions of divergent mRNAs to generate species with shorter or even leaderless 5′UTRs .
The requirement for the catalytic activity of RNase III was first demonstrated for the maturation of ribosomal RNA precursors in E . coli [34] and B . subtilis [22] . Under optimal growth conditions , the synthesis of ribosomes consumes a major fraction of available energy in cells . Thus , maturation of rRNA has to be efficient and accurate for fitness . As in E . coli and B . subtilis [22] , [34] , S . aureus rRNAs are synthesized as long 30S precursor transcripts containing the three rRNAs genes ( 16S , 23S , and 5S ) interspersed by tRNA genes ( Figure 2 ) . Using specific probes that hybridized to the spacer regions of rRNA operons , precursor transcripts were detected in the Δrnc mutant strains ( Figure 2B ) . The identification of RNase III-dependent cleavage in the processing stalk of S . aureus 16S rRNA precursors together with the conservation of the precursor structure strongly suggest that the initial processing of rRNAs is carried out by RNase III . Alternative pathways seem to substitute for 16S rRNA maturation in the absence of RNase III ( Figure 2C ) , but the responsible enzymes are not yet known in S . aureus . In B . subtilis , the final maturation steps of 23S , 16S , and 5S rRNAs involve the mini-III enzyme , the 5′-3′ exoribonuclease RNase J1 , and the double-strand-specific RNase M5 , respectively [54]–[56] . Because these enzymes are present in S . aureus , we may speculate that these maturation pathways are generally conserved in Gram-positive bacteria . Notably , analysis of tRNA/rRNA precursors in the Δrnc strain ( Figure 2B ) strongly suggested that the maturation of tRNAs is initiated by RNase III cleavage of the large rRNA precursor stalk . The present study also shows a role of RNase III in gene regulation in S . aureus . The enzyme autoregulates its own synthesis by a feedback mechanism similar to that identified in E . coli [16] , [57] and Streptomyces coelicolor [58] . Autoregulation helps to adjust the intracellular amount of the protein to that of the RNA substrates and prevents a potential detrimental over-accumulation of RNase III [59] , [60] . We show here that point mutations in the catalytic site of S . aureus RNase III cause a two to three-fold increase in the level of the mutant protein compared to the WT enzyme ( Figure 1D ) , which argues that autoregulation depends on the catalytic activity of RNase III . Furthermore , S . aureus rnc mRNA is efficiently cleaved by RNase III both in vitro and in vivo at a specific position in a stem-loop structure located in the CDS ( Figure 3D ) , which is conserved among Staphylococci . The ability of RNase III to cleave only one side of the helix is most likely due to the presence of bulged residues that interrupt the helix [27] , [61] . We propose that cleavage at this site is responsible for rnc mRNA destabilization under conditions when RNase III is in excess over its other RNA substrates . Although the feedback mechanism is preserved in distantly related bacteria , the regulatory site varies . In E . coli , RNase III targets a 5′ terminal stem-loop of its own mRNA [57] , while the S . aureus and Streptomyces [58] enzymes regulate themselves via the CDS of their respective gene . Such a structure within the rnc coding sequence might locally alter the speed of translation elongation thereby facilitating the access of RNase III . Our results show for the first time that the abundance and translation efficiency of cspA mRNA , which encodes the major cold-shock protein CspA , is modulated by RNase III-cleavages within the 5′ leader ( Figure 4 and Figure 5 ) . This RNase III processing event generates a more stable mRNA with a shorter 5′ terminal hairpin , which results in strongly enhanced synthesis of the major cold-shock protein . CspA was also found to be involved in the susceptibility of S . aureus to an antimicrobial peptide of human cathepsin G thus linking a stress response system to host-pathogen interaction [62] . Interestingly , the 5′UTR of cspA is highly conserved in Staphylococcus species and Macrococcus species , and a similar long hairpin may form upstream of the SD sequence of cspB mRNA of Listeria monocytogenes ( data not shown ) , suggesting that RNase III-dependent activation may be a conserved mechanism . The fact that RNase J1 , a major 5′-3′ exo- and endoribonuclease in Gram-positive bacteria , is inhibited by a 5′ terminal hairpin [53] , [63] may explain why the shorter stem-loop structure at the 5′ end stabilizes cspA mRNA . In addition , the RNase III-dependent processing of cspA mRNA promotes ribosome recruitment , most likely by resolving the inhibitory structure at the RBS . There are other examples wherein perturbation at the 5′ end impacts the stability and translation of bacterial mRNAs . Binding of deacylated tRNAThr to the 5′ leader region of B . subtilis thrS mRNA induces transcriptional read-through and mRNA cleavage , causing mRNA stabilization due to the formation of a 5′ transcription attenuator hairpin structure [64] . More recently , Streptococcus pyogenes ska mRNA is stabilized by the regulatory RNA FasX through the formation of a 9 bp helix at the 5′ end [65] . Similarly , Clostridium perfringens collagenase mRNA is stabilized by VR-RNA-dependent cleavage in the 5′ UTR , which renders the SD sequence more accessible for ribosome binding [66] . In contrast with these examples wherein trans-acting RNAs are required , we have identified a new mechanism through which RNase III-processing alone confers mRNA stabilization and enhances translation ( Figure 5C ) . We detected 58 ncRNAs that co-immunoprecipitated with RNase III ( Table 1 , Table S3 ) . Most of these RNAs have been identified previously , and many of them carry hairpin motifs that could be specifically cleaved by RNase III ( reviewed in [46] ) . For instance , RNase III-dependent cleavages were detected in the 5′ hairpin motif of RsaA in vitro , and the stability of this RNA was enhanced in the Δrnc strain ( Figure 6 ) . Similar to the quorum-sensing-dependent RNAIII , many of these ncRNAs presumably regulate gene expression by antisense mechanisms [27] and it is likely that they would be co-immunoprecipitated with their respective target mRNAs . For instance , RNAIII and two of its major target mRNAs , encoding Rot and protein A , were detected [24] , [26] , [67] . Likewise , we recovered the 5′UTRs of the sucC and folD mRNAs , which are known to base-pair with RsaE [43] . Thus , our coIP data sets should be useful to improve the prediction of ncRNA-mRNA interactions . Of note , E . coli and Salmonella RNase III were also found to affect the steady-state levels of several ncRNAs [19] , [68]–[70] , suggesting that a significant portion of the E . coli transcriptome was directly or indirectly affected by changes in the abundance of the ncRNAs . Thus , RNase III may play a more general role for trans-acting ncRNAs than it was previously appreciated . A significant number of reads representing putative asRNAs complementary to all types of RNA species were found , namely ncRNAs , sORF , and mRNAs ( Tables S2 , S5 ) . This antisense transcription was directed against 44% of the protein-coding genes . Most asRNAs were present at a low level , suggesting that they might arise from transcriptional noise ( e . g . , asRNAs against cspA and hu mRNAs; Figure 5D and Figure 7C ) . A recent study demonstrated that RNase III might rapidly remove low levels of asRNAs generated by pervasive transcription in S . aureus and other Gram-positive bacteria [28] . Interestingly , we observed that hu mRNA was more rapidly degraded in the WT strain than in the Δrnc strain ( Figure 7B ) . Because hu mRNA was not efficiently cleaved by RNase III ( Figure 7C ) , its rapid degradation might be mediated through asRNA regulation . It is tempting to propose that this RNA quality control mechanism may also contribute to fine-tune the final levels of mRNA in the cell . It is also conceivable that asRNA transcription is transiently enhanced until its concentration reaches a threshold that suffices to regulate the expression of the sense transcript . Indeed , the expression of several asRNAs was recently shown to be SigmaB-dependent , and their decreased expression levels in a ΔsigB mutant strain correlated with increasing expression of the sense transcripts [28] . Our data support the view that RNase III-dependent processing indeed contributes to regulate the level of sense mRNA . Our study further reveals RNase III targets that are derived from long 5′UTRs of divergently transcribed genes . Two of the overlapping 5′UTRs ( tagG/tagH and pdf1/SA0943 ) are processed by an unknown enzyme to generate mRNAs with shorter 5′ ends , while the processed 5′UTRs are rapidly degraded by RNase III ( Figure 8 ) . Shortening of the 5′ end of mRNAs could affect translation and mRNA stability , as illustrated for cspA mRNA ( Figure 5 ) . A coordinated regulation of TagG and TagH enzymes through overlapping 5′UTRs may be particularly important for the efficient synthesis of teichoic acids in S . aureus . Teichoic acids contribute to the structural integrity and shape of the bacteria by regulating the peptidoglycan cross-linking and metabolism during cell division . They are also required for virulence and biofilm formation ( reviewed in [71] ) . Overlapping transcripts from divergently transcribed protein-coding genes with long and overlapping 5′ or 3′UTRs have also been described in Listeria [72] . This indicates a mechanism to regulate and coordinate gene expression between neighboring genes . In conclusion , this study unveiled the sophistication and complexity of post-transcriptional regulation mediated by RNase III in S . aureus . The use of catalytically inactive but binding-competent RNase III mutants allowed the identification of a large set of structured RNase III substrates in vivo . For instance , we demonstrated the involvement of the enzyme in rRNA and mRNA processing , in RNA turnover , in the activation of translation through cis- and trans-acting factors , as well as in antisense RNA-mediated regulation . All of these functions are mediated through the catalytic activity of RNase III . However , we predict that the enzyme may also regulate gene expression through its binding activity , as was shown for the cIII gene of bacteriophage lambda . In this system , RNase III stabilized a conformation of the mRNA that rendered the ribosome binding site accessible to the ribosome [73] . Combining our methodology with comparative proteomics and transcriptomics will help to address more comprehensively the roles of this universally conserved enzyme in gene regulation in response to stress and during host infection .
Mutations E135A and D63A were introduced into the S . aureus RNase III enzyme following the Quickchange XL Site-directed mutagenesis procedure ( Stratagene ) . Experimental details for the preparation of the biological materials and other detailed protocols on Northern blot analysis , RNA structure probing , and toeprinting are given in Text S1 . The strains and plasmids used in this study are listed in Table S7 . Wild-type ( WT ) strain RN6390 or the isogenic Δrnc mutant strains alone or transformed with plasmids expressing either E135A , D63A or WT enzymes were grown in BHI medium at early exponential phase ( OD 600 nm 0 . 2–0 . 3 ) . Then , 10 µM of CdCl2 was added to the cultures , and after 2 h and 4 h of induction , the cells were pelleted and snap-frozen in liquid nitrogen . The bacterial cell pellet was suspended in lysis buffer ( TBS , 1% Triton X-100 and protease inhibitor cocktail ) , transferred onto glass beads ( provided by FastRNA Pro Blue Kit , Qbiogene ) and processed in the FastPrep instrument ( 3×45 s at a setting of 6 . 0 ) . Samples were centrifuged at 13 , 000 rpm for 5 min . The supernatants were mixed with mouse IgG-agarose ( Sigma , A0919 ) to remove non-specifically binding proteins and incubated at 4°C for 50 min . The beads were spun down ( 1 , 500 g , 5 min ) and the pre-cleared supernatants ( 3 ml ) were kept separately . A fraction of the volume ( 0 . 2 ml ) was removed for total RNA isolation and the rest of the sample was mixed with 40 µl ( packed gel volume ) of Anti-Flag M2 Affinity Gel ( Sigma , A2220 ) . Immunoprecipitation was performed according to the manufacturer's instructions . Briefly , the cleared lysates were incubated with the Anti-Flag M2 Affinity Gel for 2 h at 4°C , then the beads were washed three times with TBS . Elution was made with 0 . 2 ml of Flag Peptide ( Sigma , F3290 ) prepared at the concentration recommended by the supplier . The sample was extracted with acidic phenol and then by chloroform: isoamylic alcohol . RNA was precipitated with ethanol , treated with DNase I , extracted with phenol and precipitated . The final RNA samples were dissolved in 50 µl of sterile water and lyophilized . cDNA library construction , pyrosequencing and data analysis were done as previously described [33] , [52] . In brief , cDNA-seq libraries were constructed with RNA samples from coIP experiments under exponential and late-exponential phase growth of the Flag-tagged wild-type and mutant enzymes expressed from the inducible plasmid . The resulting cDNA libraries were sequenced on a Roche 454 sequencer using FLX and Titanium chemistry . From the resulting cDNA reads , 5′-linker sequences and polyA-tails were clipped from the sequenced cDNA reads . Only reads of ≥18 nt were aligned to the reference genome , which was retrieved from the NCBI server ( accession number of the chromosome: NC_002745 . 2; accession number of the plasmid: NC_003140 . 1 ) , using the program segemehl [74] . Based on the resulting mapping data , read coverage files were generated in the GR format representing the number of mapped reads per nucleotide . The GR files were visualized in combination with FASTA and GFF files of the genome using the Integrated Genome Browser ( IGB ) [75] . Additionally , overlaps of mapped reads and gene annotation positions were identified and counted . The overlap between mapped read and a gene annotation had to be at least 10 nucleotides long to be taken into account . Each single overlap counting was normalized by the number of positions to which the overlapping read was mapped and the number of annotations that overlap with the read . For instance , if reads map to multiple regions with exactly the same score ( e . g . this is the case for reads that map to the different multiple copies of the rRNA genes ) , only a relative fraction of one read is counted instead of a count of one read . For example , if a read maps twice , each location gets a score of 0 . 5 reads . Moreover , if a read overlaps two annotations , each annotation gets a score of 0 . 5 reads ( Table S1 ) . Text S1 provided experimental details for all the experiments performed in this study .
|
Control of mRNA stability is crucial for bacteria to survive and rapidly adapt to environmental changes and stress conditions . The molecular players and the degradation pathways involved in these adaptive processes are poorly understood in Staphylococcus aureus . The universally conserved double-strand-specific endoribonuclease III ( RNase III ) in S . aureus is known to repress the synthesis of several virulence factors and was recently implicated in genome-wide mRNA processing mediated by antisense transcripts . We present here the first global map of direct RNase III targets in S . aureus . Deep sequencing was used to identify RNAs associated with epitope-tagged wild-type RNase III and two catalytically impaired but binding-competent mutant proteins in vivo . Experimental validation revealed an unexpected variety of structured RNA transcripts as novel RNase III substrates . In addition to rRNA operon maturation , autoregulation , degradation of structured RNAs , and antisense regulation , we propose novel mechanisms by which RNase III increases mRNA translation . Overall , this study shows that RNase III has a broad function in gene regulation of S . aureus . We can now address more specifically the roles of this universally conserved enzyme in gene regulation in response to stress and during host infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"prokaryotic",
"models",
"model",
"organisms",
"rna",
"processing",
"gene",
"expression",
"genetics",
"bacillus",
"subtilis",
"biology",
"microbiology",
"microbial",
"growth",
"and",
"development",
"genetics",
"and",
"genomics",
"rna",
"stability"
] |
2012
|
Global Regulatory Functions of the Staphylococcus aureus Endoribonuclease III in Gene Expression
|
Endoplasmic reticulum ( ER ) stress generated unfolded stress response ( UPR ) is a basic survival mechanism which protects cell under unfavourable conditions . Leishmania parasite modulates host macrophages in various ways to ensure its survival . Modulation of PI3K-Akt pathway in delayed apoptotic induction of host; enables parasite to stabilize the infection for further propagation . Infected RAW macrophages were exposed to campothecin or thagsigargin and phosphorylation status of PERK , Akt , BAD and Cyt-C was determined through western blotting using phospho specific antibody . Expression at transcriptional level for cIAP1 &2 , ATF4 , CHOP , ATF3 , HO-1 and sXBP1 was determined using real time PCR . For inhibition studies , RAW macrophages were pre-treated with PERK inhibitor GSK2606414 before infection . Our studies in RAW macrophages showed that induction of host UPR against L . donovani infection activates Akt mediated pathway which delays apoptotic induction of the host . Moreover , Leishmania infection results in phosphorylation and activation of host PERK enzyme and increased transcription of genes of inhibitor of apoptosis gene family ( cIAP ) mRNA . In our inhibition studies , we found that inhibition of infection induced PERK phosphorylation under apoptotic inducers reduces the Akt phosphorylation and fails to activate further downstream molecules involved in protection against apoptosis . Also , inhibition of PERK phosphorylation under oxidative exposure leads to increased Nitric Oxide production . Simultaneously , decreased transcription of cIAP mRNA upon PERK phosphorylation fates the host cell towards apoptosis hence decreased infection rate . Overall the findings from the study suggests that Leishmania modulated host UPR and PERK phosphorylation delays apoptotic induction in host macrophage , hence supports parasite invasion at early stages of infection .
Leishmania ability to defy the host immune response is a major cause for persistence of Leishmaniasis . Parasite modulates host in various aspects and hampers the activation of adaptive immune responses against infection [1] . Expression of LPG on parasite surface and TLR 2 modulates the host immune response [2 , 3] by providing resistance to complement , attachment and entry into macrophages , protection against proteolytic damage within acidic vacuoles [4] or inhibition of phagosomal maturation [5] . Leishmania parasite has to counter the oxidative and nitrosative pressure generated in host macrophage against invasion . Parasite modulates host NO and IL-12 production [6 , 7 , 8 , 9] and induces host HO-1which suppress the production of superoxide and increases parasitic burden [10] . Unfolded protein response ( UPR ) is an evolutionary conserved mechanism that restores cellular homeostasis and ensures cell survival during Endoplasmic Reticulum stress . UPR consists of 3 signalling pathways: activation of transcription factor ( ATF ) -6 , inositol-requiring enzyme ( IRE ) -1 , and PKR-like endoplasmic reticulum kinase ( PERK ) [11] . As a downstream consequence , activated IRE1 increases cytosolic concentration of spliced XBP1 ( sXBP1 ) ; an active transcription factor which in turn activates an array of genes to counter ER stress for cell survival [12 , 13 , 14] . L . amazonensis induces the IRE1a/XBP1 pathway to escape cellular defence; particularly oxidative stress by sXBP1 induced expression of antioxidant molecules such as SOD-1 and catalase , and increases expression of IFN1-β , an important cytokine that favour’s L . amazonensis infection [15] . In another aspect of modulation , Leishmania parasite induces a mild ER stress in host macrophages and activates the PI3K-Akt pathway [16] . However , a very few studies has been conducted in correlation with induced UPR and leishmaniasis , but the PI3K-Akt phosphorylation is of great importance in parasite biology . As a part of host modulation , Leishmania promastigote activates host PI3K-Akt pathway , phosphorylates BAD , checks cytochrome-C leakage from mitochondria , inhibits caspase activation; hence delays apoptotic induction in host macrophage [17] . PERK being an important component of integrated endoplasmic reticulum stress response ( IERSR ) [18]; its activation influences cellular adaptation through multiple mechanisms [19 , 20] . Induction of mammalian inhibitor of apoptosis proteins ( IAP proteins ) is one of the PERK-dependent survival mechanisms which protects cell from ER stress- induced apoptosis [21 , 22] . Also , activation of the PI3K–Akt pathway by ER stress is dependent on PERK [23] . Also , PERK phosphorylation under oxidative stress is crucial for protection against oxidative stress and maintaining redox balance for cell survival [24 , 25] . So , in this study we first confirms the previous findings that Leishmania infection delays host cell apoptosis under exposure to apoptotic inducers by phosphorylating Akt , BAD protein which inhibits the caspase-3 activity , resulting delayed apoptosis . Further we investigated the effect of L . donovani infection in imparting endoplasmic reticulum ( ER ) stress in host and activation of unfolded protein response ( UPR ) . Here , we first time reported that Leishmania infection induced host UPR relays on PERK activation . Our findings suggest that inhibition of PERK phosphorylation render RAW cells more prone to apoptosis ultimately leading to decreased infection rate . Moreover , activation of IAP ( inhibitor of apoptosis proteins ) class of proteins which protects cell from various apoptotic stimulus; depends upon PERK phosphorylation . Overall we conclude that UPR activated against L . donovani infection is a PERK dependent process and have an important role in delayed onset of apoptosis in host macrophages; hence contributes to Leishmania infectivity .
Leishmania donovani promastigotes used in all experiments were of clones AG83 ( MHOM/IN/ 1983/AG83 ) . The promastigotes were grown in fresh M199 media ( Gibco ) in 25 cm2 flasks ( Nunc ) at 25°C supplemented with 10% FBS ( Gibco ) . Cultures were allowed to reach stationary phase ( 5–6 days post inoculation ) , as determined by growth curve analysis ( S1 Fig ) , prior to inoculation into fresh medium . Syrian golden hamsters were used to maintain the infectivity of the clone regularly [26] . The RAW 267 . 4 cell line ( obtained from NCCS , Pune ) was maintained in a 25-cm2 flask in RPMI-1640 medium ( Gibco ) supplemented with 10% FBS and antibiotics ( streptomycin and penicillin ) in a humidified 5% CO2 incubator ( Sanyo , Japan ) at 37°C . The growth and condition of cultures were routinely checked under an inverted microscope ( Nikon , USA ) . The cells were sub cultured in every 72 hrs . RAW267 . 4 macrophages were allowed to infect with L . donovani promastigotes for 4 h , washed to remove unattached promastigotes and further infection was carried out for mentioned time periods ( 4 h /12 h/ 36 h ) before campothecin ( 2mM/ 6 h ) or thapsigargin ( 1μM/ 1hr ) treatment . For inhibition studies , RAW267 . 4 macrophages were pre-treated with GSK2606414 for 2 hours before allowing infection . Reverse transcription was performed using 0 . 2 mg total RNA using an anchored oligo ( dT ) ( H-dT11M , where M represents A , C , or G; Gen- Hunter ) . Real-time PCR , was performed in the Light Cycler 480 ( Roche ) using SYBR green ( Roche ) chemistry . The cycling conditions were; 1 cycle at 95°C for 3 min and 40 cycles of 95°C for 15s ( denaturation ) , 58°C for 30s ( annealing ) , and 72°C for 30s ( extension ) . The fluorescence signal was captured at the end of each cycle using the SYBR channel ( excitation wavelength 490-nm and emission wavelength 525-nm ) . Results expressed as target/reference ratios of each sample , normalized by the target/reference ratio of the calibrator . Here , the target/reference value of untreated/normal RAW macrophages was used as the calibrator and the GAPDH gene was used internal control/ reference to normalize the qRT-PCR in the experiments . The primers used for RT-PCR are mentioned in tabular form . ( Table 1 ) . RAW 264 . 7 cells were plated in 6-well polystyrene plates 1 d before infection and were either pre-treated with GSK2606414 or untreated . Total protein extracts were prepared from infected cells 4 h , 12 h or 36 h after infection . For preparation of total cellular extracts , macrophages were washed 3 times with PBS and then lysed in 100 ml lyses buffer [50mMTris- HCl ( pH 7 . 5 ) , 5 mM EDTA , 10 mM EGTA , 50 mM NaF , 20 mM b-glycerophosphate , 250mM NaCl , 0 . 1%Triton X-100 , and 1mg/ml bovine serum albumin ) supplemented with 1:100 dilution of protease inhibitor cocktail ( Sigma-Aldrich ) . Total protein ( 50μg ) was separated by SDS-PAGE . The proteins were transferred to PVDF membranes ( Bio-Rad , Hercules , CA , USA ) and blotted with antibodies against phosphos Akt ( Santa Cruz Biotechnology , Dallas , TX , USA ) , phosphos PERK ( Cell Signaling ) , phosphos BAD ( Santa Cruz Biotechnology ) , GAPDH ( Santa Cruz Biotechnology ) , anti-PERK ( Cell Signaling ) overnight . Membranes were incubated with horseradish peroxidase–conjugated IgG ( 1:5000 ) for 1 h in room temperature and washed extensively with Tris-buffered saline- Tween ( TBST ) . Antibody antigen complexes were detected by enhanced chemiluminescence kit ( Thermo Fischer Scientific , USA ) was used for detection . The intensity of bands was analyzed using Quantity One software ( Bio-Rad ) . RAW 267 . 4 macrophages either untreated or pre-treated with GSK2606414 were infected with L . donovani promastigotes . After desired time periods , the cells were induced with campothecin ( 2μM/ 6 h ) in six well plates ( 1x106 cells ) . The cells were then scraped and harvested , lysed in cell lysis buffer supplied with the caspase-3 Fluorometric Assay Kit ( BioVision ) and the concentration of protein in lysates was determined . Equal amount of protein from each set was used to determine caspase-3 activity . The detection was based on the cleavage of AFC ( 7-amino-4-trifluoromethyl coumarin ) from DEVD-AFC by caspase-3 and the relative fluorescence was measured in the PerkinElmer LS-55 spectrometer ( Perkin Elmer , USA ) at excitations of 390 to 400 nm and emissions of 510 to 550 nm . Raw cells ( 1×106 ) were allowed to adhere into 6- well ( Corning , USA ) and incubated at 37°C in 5% CO2 incubator ( Sanyo , Japan ) . Once macrophages were adhered , pre-treatment with GSK2606414 was given and washed . Further infection was carried out by adding L . donovani promastigotes to each wells and maintained at 37°C in 5% CO2 overnight . Non- internalized promastigotes were removed by gentle washing twice with PBS . Infected macrophages were treated campothecin ( 6 h/ 2μM ) or thapsigargin ( 1 . 0μM / 1hr ) in 5% CO2 at 37°C . Untreated infected macrophages were taken as a control . Total genomic DNA was isolated using an apoptotic DNA laddering kit ( Roche , USA ) as described earlier [27] GSK2606414 ( Calbiochem ) , a PERK inhibitor , was added at a 30nM final concentration to RAW 267 . 4 macrophages , 2 h prior to the infection , washed and the macrophages were allowed to infect with L . donovani promastigotes ( MOI 10:1 ) . Further treatment of campothecin ( 2μM/ 6 h ) [17] or thapsigargin ( 1 . 0μM / 1hr ) [15] was given . For infection studies , RAW 264 . 7 cells were plated in 6-well polystyrene plates 1 day before infection . The infection index was calculated by multiplying the percentage of infected macrophages by the average number of parasites per macrophage on Giemsa-stained slides [15] . Untreated or GSK2606414 pre-treated RAW 267 . 4 macrophages were plated in six well tissue culture plates overnight and then infected with late stationary phase Leishmania promastigotes . After 4 h , free parasites were washed off with PBS and the treatment of either 2mM campothecin for 6 h or thapsigargin ( 1μM/ 1hr ) was given . Cells were trypsinized and resuspended in phosphate-buffered saline . Propidium iodide was added to a concentration of 2 . 5 mg/ml [20] . Propidium iodide fluorescence was measured with a Becton Dickinson FACS ARIA . Nitric oxide ( NO ) concentration was determined by analysing nitrite content in cell supernatant with the Griess reaction . L . donovani infected RAW cells either GSK 2606414 pre-treated or untreated were exposed to H2O2 ( 200μM/ 4hr ) and Culture supernatant from each set ( 50μL ) was mixed with 50μL of a solution containing sulphanilamide ( 10 mg/mL ) , N-[naphthyl] ethylenediamine dihydrochloride ( NEED; 1 mg/mL ) and 5% phosphoric acid . The absorbance was measured by spectrophotometry at 540 nm [28] . Graph Pad Prism ( Version 6 . 0 , GraphPad Software , USA ) was used to analyse the data statistically . Each experiment was repeated three times in separate sets and the results were expressed as Mean ± SD . Statistical differences were determined using Mann–Whitney U test for comparing two groups or Kruskal–Wallis with Dunn’s multiple comparison test for comparing three or more groups . A P- value of <0 . 05 was considered significant .
Leishmania infection is known to elicit ER stress and activates UPR response in host . [16] . To ensure the same , expression level of ER stress marker genes viz; CHOP , ATF3 , ATF4 and sXBP1 was monitored by qPCR at 4hr , 12hr and 36hr post infection . A significant increase in the expression of ER stress marker gene was observed after infection . At 4hr and 12hr post infection , the increase in level of marker genes was more prominent i . e . 1 . 5–2 . 0 folds higher ( Fig 1A ) . At later stage of infection ( 36hr ) , a general decrease in all ER stress marker genes was observed however the value was higher compared to uninfected RAW cells . As a positive control , RAW cells treated with an ER stress inducer ( thapsigargin 1 . 0 μM/ 1hr ) were considered . The selected ER stress marker genes were found to be significantly upregulated with higher magnitude of induction ( Fig 1B ) . GAPDH was chosen as reference gene since its expression did not change significantly following infection . As Leishmania infection induces ER stress in host macrophages , we further investigated the effect of generated ER stress on infection rate . RAW macrophages treated with thapsigargin ( Thg ) or tunicamycin ( Tu ) for 1 hour were infected with L . donovani . After 36 hours , the infection index showed significantly higher parasite burden in thapsigargin ( Thg ) or tunicamycin ( Tu ) treated cells compared with that in non-treated cells ( UNS ) ( Fig 1C ) . The result suggests that ER stress induced RAW macrophages have enhanced parasite infection . Phosphorylation and simultaneous activation of PERK under stress is a major UPR which influences cellular adaptation and survival under stress through multiple mechanisms [18] . As the L . donovani infection induces ER stress in host macrophages , we further investigated the status and involvement of the host PERK in mounting UPR . Western blotting using phospho anti-PERK antibody revealed that in infected RAW cells , host PERK was phosphorylated ( Fig 2A ) . At early stages of infection ( 4hr & 12hr ) the degree of phosphorylation was higher compared to 36 hours post infection as shown in densitometry ( Fig 2B ) . In case of thapsigargin treatment the phosphorylated status of PERK was confirmed and was taken as positive control for the experiment . No any phosphorylation was detected in untreated RAW macrophages . Leishmania parasites modulate the host cell by activating PI3K/Akt signalling and inhibiting caspase-3 activity which delays host cell apoptotic process . To confer this delayed apoptosis , Akt phosphorylation status , caspase3 activity and percentage of cells undergoing apoptosis was determine in RAW cells infected with L . donovani . A significant rise in Akt phosphorylation was observed in infected RAW cells compared to uninfected controls . The figure ( Fig 2C ) shows phosphorylation of Akt at early stages of infection ( 4hr &12hr ) and was well maintained at 36 hours post infection . The infection had no effect on native Akt levels as represented in densitometric analysis of western blot bands ( Fig 2D ) . Further , caspase-3 activity was accessed in uninfected or infected RAW cells or infected RAW macrophages treated with campothecin ( 2mM ) or thapsigargin as inducers of apoptosis . The level of caspase-3 activity was found significantly high in uninfected macrophages subjected to campothecin ( RAW+Cmpt ) or thapsigargin ( RAW+Thg ) exposure than untreated control . In contrary , there was lower/limited activity of caspase-3 in macrophages that were first infected for 4hr , 12hr or 36hr with L . donovani promastigotes before exposure to campothecin or thapsigargin ( Fig 3A ) . However , at 36hr post infection , the level of caspase-3 activity was higher than early stages of infection . Simultaneously , percentage of macrophages undergoing apoptosis was also determined in case of infected RAW macrophages compared to uninfected ones under exposure to apoptotic inducers . In uninfected RAW macrophages , the percentage of cells undergoing apoptosis was recorded significantly high compared to infected RAW cells ( Fig 3B ) . Hence , the results were in agreement with previous findings that Leishmania infected macrophages are more tolerant to apoptotic induction [17] . The mammalian inhibitor of apoptosis ( IAP ) gene family , particularly cellular IAPs ( cIAP1 and cIAP2 ) , provides cell protection during infection and against a variety of apoptotic stimuli [29 , 30 , 31 , 32 , 33] . It has been demonstrated that ER stress also induces the expression of cIAP1 and cIAP2 [34 , 35] in a PERK-dependent manner [23] . As L . donovani infection delays host apoptosis induction , we accessed the level of cIAP1 and cIAP2 in infected RAW macrophages . Real time data revealed that the expression of cIAP1 and cIAP2 was induced in case of infected RAW cell , whereas no such induction was detected in uninfected macrophages . Induction of both cIAP1 and cIAP2 was recorded transiently highest at 4 hours post infection which gradually declined at latter stages of infection . Similar up regulation of cIAP1 and cIAP2 was also observed at mRNA level in RAW macrophages subjected to thapsigargin ( 1h/ 1μM ) or campothecin ( 6h/ 2mM ) treated RAW cell and were taken as control ( Fig 4A and 4B ) . The above results reveal that L . donovani infection induces ER stress , phosphorylates host Akt molecule and delays host macrophage apoptosis which facilitates infection . Also , induced ER stress phosphorylates host PERK and leads to accumulation of cIAP mRNA . So to further explore the connection between L . donovani infections induced ER stress and host PERK phosphorylation mediated UPR in modulating host cell apoptosis , inhibition studies was carried out . For inhibiting PERK mediated UPR response , RAW 267 . 4 macrophages were pre-treated with GSK2606414 , a specific inhibitor of PERK phosphorylation [36 , 37 , 38 , 39 , 40]; followed by washing before any infection or exposure as it is also effective against Leishmania PERK [26] . However , GSK2606414 pre-treatment did not affect the parasite entrance as initially at 4 hours , no any significant difference in terms of parasite entrance was observed in the GSK2606414 pre-treated RAW macrophages than untreated/normal one ( S2 Fig ) . Finally , the percentage of infected RAW macrophages undergoing apoptosis after campothecin or thapsigargin or H2O2 exposure was determined in presence or absence of GSK2606414 . The results revealed that GSK2606414 inhibition significantly increases the percentage of cells undergoing apoptosis ( Fig 9A ) . Under H2O2 exposure , inhibition of PERK phosphorylation leads to significant increase in apoptotic cell percentage . L . donovani infected RAW macrophages in presence of Wortamanian also showed increased apoptotic rate; however the change was less significant compared to inhibitory effect of GSK2606414 , suggesting that the effect of PERK inhibition is more prominent . As the ER stress favours L . donovani infection in RAW macrophages , infection rate after inhibition of host PERK was investigated . RAW macrophages pre-treated with GSK2606414 were infected with L . donovani promastigotes . After 36 hours , infection status was determined in terms of infection index . In case of PERK inhibition , a drastic decrease in parasite burden was observed compared to untreated normal RAW macrophages subjected to infection as represented in infection index ( Fig 9B ) . The similar pattern was also reported in thapsigargin induced RAW cells . This finding suggests that PERK mediated UPR is required for successful infection .
Various intracellular parasites deploy different strategies for their survival inside host; however one thing common among them is delayed/resistance against induced apoptosis via wide range of agents [46 , 47] . Leishmaniasis disease is mainly contributed to the parasite ability to modulate host machinery to thrive the harsh phagolysosomal environment . Leishmania parasite deploy different strategies like; increases expression of host redox homeostasis enzymes like HO-1 , SOD-1or catalase to sustain oxidative burst [10 , 15] , delays host cell apoptosis via PI3K-Akt pathway [17] , induces ER stress to escape cellular defence , increases expression of IFN1-b etc . In accordance with previous findings , we observed that L . donovani infection induces ER stress in RAW macrophages and consecutively activates UPR of the host as the level of marker genes like ATF3 , ATF4 , sXBP1and CHOP were found upregulated ( Fig 1A ) . A similar trend of gene expression was observed in thapsigargin treated RAW macrophages ( Fig 1B ) , which further validated the observation . The magnitude of UPR induction during infection was comparatively lower than thapsigargin treatment which gradually decreases at later time points; might be a requirement for spread and progression of infection after successful establishment . Moreover , a general down regulation of gene expression at later stages of infection has been reported in literature [48 , 49] . Akt mediated signalling pathway has its unique importance in apoptosis induction . Leishmania parasite , as a survival strategy delays the apoptotic induction in host macrophages via phosphorylation and activation of Akt molecule [17] . ER stress also is capable of inducing Akt phosphorylation [23] which establishes a connection between UPR and PI3K-Akt pathway . In our studies , we observed Akt phosphorylation under apoptotic inducers exposure and L . donovani infection ( Fig 2C ) . Simultaneously , with Akt phosphorylation , PERK was also found phosphorylated ( Fig 2A ) in infected macrophages . So the result bridges a connection between Leishmania infections , PERK mediated UPR activation and Akt pathway activation . Leishmania infection confers host cell resistance to apoptosis by modulating PI3K/Akt signalling pathway [17] which we also confirmed in our studies using campothecin as apoptotic inducer . As compared to uninfected RAW macrophages , Leishmania infection brings about phosphorylation of host Akt with low caspase-3 activity in infected RAW cells . The level of caspase-3 activity was found higher in uninfected RAW cells which indicates early onset of apoptosis ( Fig 3A ) . The mRNA level of cIAP1 and cIAP2 during Leishmania infection was comparatively higher than under apoptotic inducers ( Fig 4 ) however; the induction was transient which decreases at later stages of infection . It has also been reported that at translational level both cIAP1 and cIAP2 protein continues to accumulate even at later stages and have key role in delayed apoptosis [23] . To further investigate the importance of PERK mediated UPR in L . donovani infected macrophages in connection with delayed apoptosis of host macrophages; inhibition studies using GSK2606414 was performed . RAW macrophages pre-treated with GSK2606414 were infected with L . donovani promastigotes . After 4 hours post infection , infected RAW macrophages were exposed to campothecin or thapsigargin for apoptotic induction and the molecular mechanism involved in apoptosis was investigated . The level of Akt phosphorylation was found to be decreased in case of GSK treated macrophages compared to untreated one under thapsigargin exposure signifies that UPR-dependent Akt activation requires PERK phosphorylation ( Fig 5A ) . Similar results were obtained in case of campothecin exposure; hence it can be concluded that induction of Akt phosphorylation under apoptotic stimulus is mediated via PERK phosphorylation . Akt activation is implicated in the regulation of IAP gene transcription [34] and PERK regulates IAP expression in an Akt-dependent manner [23] . Similarly in GSK treated RAW macrophages , the level of cIAP1 and cIAP2 mRNAs was relatively in lower amount signifies that inhibition of host PERK hampers cIAP1 and cIAP2 accumulation ( Fig 5C ) . Following the consequence , Akt phosphorylation brings about deactivation of downstream BAD protein . BAD being a proapoptotic molecule forwards the apoptosis process . Akt phosphorylation further phosphorylates BAD protein which checks its activity by facilitating its binding with 14 . 3 . 3 protein . In our findings , we observed decreased phosphorylation of BAD in GSK2606414 pre-treated infected macrophages than infected RAW macrophages without GSK2606414 ( Fig 6C ) . Hence , PERK inhibition decreases infection induced BAD phosphorylation . Phosphorylation of BAD render it inactive whereas , unphosphorylated BAD alters mitochondrial membrane permeability leading to cytochrome C leakage from inner mitochondrial space to cytosol which in turn activates proteases i . e . caspase like protein . Campothecin being an apoptotic activator enhances the release of cytochrome C in RAW macrophages . In Leishmania infected RAW macrophages , a low level of released cytochrome C suggests that infected macrophages are more tolerant to apoptotic inducers however , PERK inhibition leads to increased cytochrome C release even in infected RAW cell ( Fig 7A ) . For further determination of apoptotic progression , caspase-3 activity in lysates of infected RAW macrophages was accessed after campothecin or thapsigargin treatment . In our studies , we confirm the delayed onset of apoptosis in L . donovani infected RAW macrophages by interfering caspase-3 activity . However GSK2606414 treatment abolishes the effect of infection as higher caspase-3 activity was observed . In GSK pre-treated RAW macrophages , PERK inhibition alters the Akt mediated anti-apoptotic mechanism resulting higher caspase-3 activity ( Fig 7C ) . Increased caspase-3 activity finally damage cell DNA to ensure apoptotic death of cell . DNA fragmentation analysis following campothecin treatment suggests that PERK inhibition leads to increased DNA damage even in infected macrophages ( Fig 7D ) . Apart from activation of caspase cascade , UPR induced Akt phosphorylation also regulates the level of cIAP mRNA in PERK dependent manner . In infected RAW cells , inhibition of PERK phosphorylation interferes with mRNA accumulation . A decreased cIAP1 and cIAP2 mRNA level due to initial inhibition of PERK; reflects the same controlling mechanism . Contrary to this , in L . donovani infected RAW cells , an upregulated level of cIAP1 and cIAP2 mRNA was observed ( Fig 5C ) . Hence it can be extracted that parasite induces expression of IAP to delay apoptotic induction of host macrophages . Increased level of released Cyt-C from mitochondria as a result of altered mitochondrial membrane potential; propagates the apoptosis [50] . ROS is one of the important reasons which alter mitochondrial membrane potential and leads to increased leakage of Cyt-C [51] responsible for apoptotic induction [52 , 53] . Since PERK mediated signaling pathway has a distinct protective role against ROS; inhibition of the same resulted in increased NO level ( Fig 8C ) . Also , inhibition of oxidative stress induced PERK phosphorylation leads to higher apoptotic cell percentage ( Fig 9A ) with less chances of survival of infected cells; hence decreased infection rate ( Fig 9B ) . So overall it can be concluded that inhibition of host PERK during L . donovani infection , hampers the parasite ability to induce ER stress mediated host Akt phosphorylation and consecutively the phosphorylation of host BAD protein; which in turn alters mitochondrial membrane , increases cytochrome C release and finally DNA fragmentation . Simultaneously , decrease in IAP production in RAW macrophages due to PERK inhibition leads to an early onset of apoptotic process which ultimately results in decreased the parasitic burden ( Fig 10 ) . Hence , findings from this study will help in better understanding of host-parasite interaction in relation with UPR activated PERK mediated mechanism in providing host cell resistance against apoptosis . It will open a new therapeutic option for drug target .
|
Visceral Leishmaniasis or Kala-azar is one of the severe tropical neglected parasitic diseases caused by Leishmania donovani in Indian subcontinent . Modulation of host in terms of delayed apoptotic induction is one of the aspects which favours disease establishment; however the mechanism is not clearly understood yet . In the present study , we tried to explore the connection between L . donovani infection induced UPR in host with delayed onset of apoptosis . We found that L . donovani infection phosphorylates the PERK and Akt molecule in host along with delayed apoptosis . Simultaneously , the levels of cellular IAP ( cIAP1 & 2 ) genes were also up-regulated in infected macrophages . To assess the involvement of PERK in delayed apoptosis of host , we inhibited the phosphorylation of PERK under the exposure to apoptotic inducers . We found that PERK inhibition decreased the Akt phosphorylation and fails to activate other associated downstream molecules involved in delayed apoptosis of host . Also , a significant reduction in cIAP levels was observed . Under oxidative exposure , inhibition of PERK phosphorylation debilitates infected RAW cell’s ability to maintain redox homeostasis leading to higher nitric oxide production . Altogether , L . donovani infection modulates host apoptosis in a PERK dependent manner and favours infection .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"phosphorylation",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immunology",
"cell",
"processes",
"microbiology",
"messenger",
"rna",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoan",
"life",
"cycles",
"developmental",
"biology",
"protozoans",
"leishmania",
"promastigotes",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"life",
"cycles",
"leishmania",
"donovani",
"biochemistry",
"rna",
"eukaryota",
"cell",
"biology",
"post-translational",
"modification",
"nucleic",
"acids",
"apoptosis",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"protozoology",
"macrophages",
"organisms"
] |
2018
|
Leishmania donovani induced Unfolded Protein Response delays host cell apoptosis in PERK dependent manner
|
DNA repair mechanisms are essential for preservation of genome integrity . However , it is not clear how DNA are selected and processed at broken ends by exonucleases during repair pathways . Here we show that the DnaQ-like exonuclease RNase T is critical for Escherichia coli resistance to various DNA-damaging agents and UV radiation . RNase T specifically trims the 3′ end of structured DNA , including bulge , bubble , and Y-structured DNA , and it can work with Endonuclease V to restore the deaminated base in an inosine-containing heteroduplex DNA . Crystal structure analyses further reveal how RNase T recognizes the bulge DNA by inserting a phenylalanine into the bulge , and as a result the 3′ end of blunt-end bulge DNA can be digested by RNase T . In contrast , the homodimeric RNase T interacts with the Y-structured DNA by a different binding mode via a single protomer so that the 3′ overhang of the Y-structured DNA can be trimmed closely to the duplex region . Our data suggest that RNase T likely processes bulge and bubble DNA in the Endonuclease V–dependent DNA repair , whereas it processes Y-structured DNA in UV-induced and various other DNA repair pathways . This study thus provides mechanistic insights for RNase T and thousands of DnaQ-like exonucleases in DNA 3′-end processing .
It is well known that DNA repair mechanisms maintain genomic integrity and are essential for cell survival . Damaged DNA can be restored by a variety of DNA repair processes , such as direct reversal , base excision , nucleotide excision , mismatch , and recombination repair pathways [1] . Although diverse proteins play different roles in these pathways , DNA repair is generally accomplished by a coordinated effort via several types of DNA enzymes , including endonucleases that nick DNA near the damaged site , exonucleases that trim DNA from the broken end , helicases that unwind duplex DNA , polymerases that make new strand DNA with correct sequences , and ligases that seal the restored DNA strands . Among all these DNA enzymes , the molecular functions of exonucleases , which bind at the 3′ or 5′ end of DNA and cleave one nucleotide at a time , are least understood . How they select , rather than randomly bind to , a broken end of DNA and process it up to the site for the next-step processing remains to be investigated . Here we use the bacterial exonuclease RNase T as a model system to study the processing of DNA in various DNA repair pathways . RNase T is a member of the DnaQ-like 3′–5′ exonucleases with a DEDDh domain that contains four acidic DEDD residues ( D23 , E25 , D125 , and D186 ) for binding of two magnesium ions , and one histidine residue ( H181 ) for functioning as the general base in the active site for the hydrolysis of the 3′-terminal phosphodiester bond of a nucleic acid chain [2] . The family of DnaQ-like exonucleases constitutes thousands of members , all with exonuclease activity either processing RNA during RNA maturation , interference , and turnover , or processing DNA during DNA replication , degradation , repair , and recombination . A number of the DnaQ-like exonucleases have been shown to play a role in DNA repair . Usually the DEDDh domain can be linked to a DNA polymerase domain for proofreading during DNA replication , such as the DnaQ domain of the ε subunit of E . coli pol III holoenzyme and the exonuclease domain of human pol δ , ε , and γ [3] . Mutations in or deletion of the proofreading 3′ exonuclease domain for these polymerases are either lethal or induce high mutation rates and high incidence of cancers [4] . The DEDDh domain can also be linked to a helicase domain and functions in processing of broken DNA strands during DNA repair and recombination , such as that of human WRN [5] . Mutations in the DEDDh exonuclease domain of WRN are associated with Werner syndrome that results in premature aging and increased risk of cancer [6] . However , most of the DEDDh domain functions as an autonomous protein and is not linked to a polymerase or a helicase domain . Some of these exonucleases participate in DNA 3′-end processing in DNA repair , such as ExoI and ExoX from E . coli and TREX1 and TREX2 from human [7]–[9] . ExoI and ExoX are monomeric enzymes that digest single-stranded DNA in mismatch and DNA recombination repair pathways [10]–[12] , whereas the human TREX1 and TREX2 are dimeric enzymes , likely processing single-stranded DNA in mammalian cells [13] , [14] . Mutations in TREX1 are linked to the autoimmune diseases Aicardi-Goutieres syndrome and systemic lupus erythematosus , probably due to the accumulation of nonprocessed intermediate DNA during replication and repair pathways [15]–[17] . The crystal structures of ExoI [18] , TREX1 [19] , and TREX2 [20] reveal that they all bear a classical α/β fold of the DEDDh domain; nevertheless , their precise functions in DNA processing remain uncertain . RNase T has also been implicated in the UV-repair pathways based on the observations that the cells lacking RNase T are less resistant to UV radiation and overexpression of RNase T can rescue the UV sensitivity of the ExoI knockout E . coli strain [21] , [22] . Yet RNase T was originally recognized as an RNase based on its indispensable role in tRNA 3′-end processing during tRNA maturation [23] . RNase T also performs the final trimming for various stable RNA , including 5S and 23S rRNA [24] , [25] . RNase T can digest both DNA and RNA and it has a unique specificity that its exonuclease activity is reduced by a single 3′-terminal C or completely abolished by a dinucleotide 3′-terminal CC in digesting either DNA or RNA , referred to as the C effect [26] . Moreover , its exonuclease activity is inhibited by duplex structures , referred to as the double-strand effect; therefore , a 3′ overhang of a duplex DNA or RNA is only digested near the duplex region by RNase T [26] , [27] . Previous crystal structures of RNase T in complex with various single-stranded DNA ( 3′-terminal G versus C ) and double-stranded DNA ( 1 versus 2 nucleotide 3′ overhang ) reveal the structural basis for C effect and double-strand effect [27] , [28] . The binding of an uncleavable substrate , such as a single-stranded DNA with a 3′-terminal C or a duplex DNA with a short 2-nucleotide 3′ overhang , induces an inactive conformational change in the active site and thus inactivates the exonuclease activity . Therefore , in digesting a duplex DNA with a 3′ overhang , RNase T can accurately differentiate its cleavable or noncleavable substrates based on the C effect and double-strand effect , and it produces a precise final product of a duplex with a 1-nucleotide ( if the last base pair in duplex region is AT ) or 2-nucleotide ( if the last base pair in duplex region is GC ) 3′ overhang if the CC dinucleotide is not present within the 3′ tail , or else it stops at the 3′-CC end . RNase T hence is capable to trim various precursor RNA to produce mature RNA with a precise 3′ overhang depending on the structure and sequence of these precursors: 1 nt for 5S rRNA , 2 nt for 23S rRNA , 4 nt for 4 . 5S RNA , and 4 nt for tRNA [27] , [28] . In fact , RNase T is a more efficient DNase than RNase in that it digests DNA with a 10-fold efficiency as compared to RNA ( see figure S2 in [28] ) , supporting its possible cellular role in DNA processing . However , it is not certain if RNase T indeed processes DNA in DNA repair , and if it does , how it selects and processes its DNA substrates . To determine the molecular function of RNase T , we show here by biochemical and structural approaches that it is a structure-specific DNase capable of digesting intermediate structured DNA during DNA repair . We found that RNase T not only digests bubble and bulge DNA in Endonuclease V ( Endo V ) –dependent DNA repair but also digests Y-structured DNA in UV-induced DNA repair pathways . The crystal structures of RNase T in complex with a bulge and a Y-structured DNA further demonstrate how this dimeric enzyme elegantly binds and processes these structured DNA molecules in different ways . Our results reveal , for the first time , the precise molecular role of an exonuclease in the 3′ end DNA processing and may hint at the molecular function for other members of DnaQ-like exonucleases .
To confirm the possible roles of RNase T in DNA repair , we measured the chronic and acute sensitivity of the RNase T knockout E . coli strain ( Δrnt ) ? against various DNA-damaging agents , including hydrogen peroxide ( H2O2 ) , methyl methanesulfonate ( MMS ) , 4-nitroquinoline-1-oxide ( 4NQO ) , mitomycin C ( MMC ) , and UV light . A number of exonucleases that have been shown to play a role in DNA repair , including ExoI ( ΔexoI ) [7] , [10] , ExoX ( ΔexoX ) [7] , PNPase ( Δpnp ) [29] , and RecJ ( Δrecj ) [30] , were tested in parallel for a comparison ( Table 1 and Figures S1 and S2 ) . The wild-type K-12 strain resisted all DNA-damaging agents when present at a chronic dose , whereas RNase T-deficient strain ( Δrnt ) had a slow growth phenotype and was sensitive to the chronic dose of H2O2 , MMS , 4NQO , and UV-C ( Figure 1A ) . The RNase T-deficient strain ( Δrnt ) was also sensitive to the acute dose of H2O2 in various concentrations from 20 to 80 mM ( Figure 1B ) . The sensitivity of Δrnt strain to UV-C was different from those observed in the previous report [21]; therefore , we further confirmed the UV and H2O2 sensitivity by rnt-rescued experiments , which restored the resistance of Δrnt cells against UV-C and H2O2 ( see Figure S1 ) . This result shows that the sensitivity of the RNase T knockout cells to UV-C and H2O2 is indeed due to the deficiency of RNase T . H2O2 produces a wide variety of DNA lesions , including single-strand/double-strand breaks ( DSB ) , oxidation and deamination of bases , and sugar modifications [31] , [32] , that are usually restored by direct repair ( DR ) , base excision repair ( BER ) , and alternative repair ( AR ) [1] , [29] , [30] , [33] . DNA alkylating agent MMS produces methylated DNA bases that can be restored by DR and BER [34] , [35] . MMS also leads to the accumulation of single-strand gaps ( SSGs ) and DSB-related DNA damage [29] , [35] . MMC is a DNA cross-linking agent that can trigger the SOS response and creates damage repaired by NER [29] , [36] , [37] . UV light-mimetic agent 4NQO can produce replication-blocked DNA base adducts , SSGs , and DSB-related DNA damages [29] , [38] that are mainly repaired by NER [38] . UV-C irradiation ( 100–290 nm ) leads to three major base modifications and DSB-related DNA damage that are usually repaired by BER and DNA recombination [31] , [39] , [40] . The sensitivity of the Δrnt strain to H2O2 , MMS , 4NQO , and UV-C suggests that RNase T may play a role in BER , AR , and DSB-related DNA repair pathways . In comparison to the known DNA-repair exonucleases , RNase T had a wider sensitivity to various DNA-damaging agents . The ExoI-deficient cells were only sensitive to H2O2; the ExoX-deficient cells were not sensitive to any DNA-damaging agents; the PNPase-deficient cells were sensitive to H2O2 , MMS , and UV-C; and the RecJ-deficient cells were sensitive to 4NQO and UV-C ( Table 1 and Figure S1 ) . RNase T had a more apparent and wider sensitivity as compared to those of ExoI , ExoX , PNPase , and RecJ , suggesting that RNase T plays more extensive and crucial roles in various DNA repair pathways . To further characterize the role of RNase T in DNA repair pathways , the single-stranded DNA containing a methylated , deaminated , or oxidized base , or an abasic site at the 3′-terminal end ( 5′-GAGTCCTATAX-3′ ) were incubated with RNase T in the DNA digestion experiment . We found that RNase T digested the DNA with a methylated base—O4-methylthymine ( O4-mT ) and O6-methylguanine ( O6-mG ) —and a deaminated base—uracil and hypoxanthine . However , the DNA with a 3′-terminal oxidized base , 8-oxoguanine ( 8-oxo ) , and an abasic site were more resistant to RNase T digestion ( see Figure 1C ) . This result suggests that RNase T can function as an exonuclease in the excision step for methylated and deaminated bases in BER and AR . The next question we tackled was what types of DNA that can be processed by RNase T in DNA repair , besides the single-stranded DNA with a lesion . RNase T is not an appropriate exonuclease for digesting single-stranded DNA since its exonuclease activity is easily blocked by any C within a DNA strand . A variety of intermediate structured DNAs are generated during DNA repair , such as bulge , bubble , and Y-structured DNA . Bulge DNAs are produced in frameshift DNA mutations during DNA replication of repetitive sequences [41] , whereas bubble DNA are generated in mismatch replication or deamination of DNA bases [42] . Y-structured DNAs are generated in various DNA repair pathways , such as mismatched DNA repair and DNA recombination ( see Discussion ) . To test if RNase T processes these intermediate structured DNA , we incubated RNase T with different DNA and found that RNase T can digest Y-structured DNA and blunt-end bubble DNA with an I-T or I-G bubble ( Figure 2A ) . In digesting the Y-structure DNA , the exonuclease activity of RNase T was blocked by the duplex structure—that is , double-strand effect—and RNase T generated a final product of 1-nucleotide 3′ overhang duplex ( Figure 2A ) . In digesting bulge and bubble DNA , the double-strand effect did not occur , and the blunt-end bulge and bubble DNA was cleaved by RNase T ( Figure 2A ) . We further tested the sequence preference of RNase T in digesting the structured DNA . In digesting a classical duplex DNA with a short 3′ overhang , the exonuclease activity of RNase T was blocked by a dinucleotide 3′-end CC sequence ( Figure 2B ) . In contrast to the duplex DNA , the bulge and Y-structured DNA with terminal 3′-CC were processed by RNase T into a final product with a 1-nucleotide 3′ overhang ( Figure 2B ) . Therefore , in digesting bubble and bulge DNA , RNase T has no sequence preference , and it removes the last paired nucleotide of any sequence to generate a 1-nucleotide 3′ overhang . In digesting Y-structured DNA , RNase T also has no sequence preference and processes the 3′ tail of any sequence close to the duplex region to generate a 1-nucleotide 3′ tail as the final product . We were intrigued by how RNase T could bind and process a bubble or bulge in DNA with a blunt end . Previous studies showed that the double-strand effect of RNase T requires a 3′ overhang of a duplex with a length of more than 2 nucleotides for inserting into the active cleft for digestion ( see Movie S1 ) . To reveal how RNase T binds and processes a DNA bulge with a blunt end , we co-crystallized RNase T with two bulge DNA molecules , one with a 3′-end TC and one with a 3′-end CC sequence in acidic conditions , pH 5 . 5 and pH 6 . 0 , respectively ( see Table S2 ) . RNase T only digests nucleic acids in basic conditions because the general base H181 has to be deprotonated to activate a nucleophilic water for hydrolysis . Therefore , due to the low pH , the bulge DNA in the crystal were not cleaved by RNase T . The crystal structure of the two complexes was solved by X-ray diffraction methods at a resolution of 1 . 8 and 2 . 0 Å , respectively ( see Figure 3 ) . In the RNase T–bulge DNA complex structures , the dimeric RNase T bound to two bulge DNAs , with the 3′ end of DNA binding at the active site of each protomer ( Figure 4 ) . The bulge DNA was bound between the two RNase T protomers , in a way similar to that of the classical duplex DNA with a 3′ overhang [28] . However , in contrast to the previous duplex DNA complex , the aromatic side chain of Phe29 was inserted into the bulge and stacked with the two neighboring GC base pairs in both of the bulge DNA complexes ( see Figure 4A and 4B ) . In the previous duplex DNA complex , Phe29 was stacked with the 5′-end base of the opposite nonscissile strand , and the stacking stopped the further cleavage of the scissile strand at the 3′ end , resulting in the double-strand effect ( see the schematic comparison in Figure 4C ) . The crystal structure of the bulge DNA complex revealed how RNase T can overcome the double-strand effect by inserting Phe29 into the bulge so that the 3′-end scissile phosphate was moved accordingly into the active site ( see Movie S2 ) . We found that the active site of the bulge DNA complex indeed had an active conformation with two bound Mg2+ ions , and the general base His181 was located close to the scissile phosphate ( Figure 4B ) . The crystal structure of the bulge DNA complex also revealed how RNase T could overcome the C effect . The 3′-end cytosine was paired with the 5′-end guanine , and this base pairing prevented Glu73 from interacting with the 3′-end C to induce the C effect ( Figure 2B , Figure S2B ) . Therefore , the bulge DNA could be processed by RNase T without any sequence preference . Moreover , the 5′ end of bulge DNA was not hindered by any residue and could further extend ( Figure S2A ) , suggesting that RNase T can cleave bulge DNA with a long single-stranded region at the opposite strand , similar to those DNA in the frameshift DNA mutations ( see Discussion ) [41] . The crystal structure thus reveals at the atomic level how RNase T binds and processes a bulge DNA with a blunt end without a sequence preference . The bubble and bulge DNA can be produced by Endo V , which makes a nick at the 3′ side one base pair away from a damage site with a deaminated base in the alternative DNA repair [42] . Endo V also processes mismatched DNA , hairpin-containing DNA , bulge DNA , and flap DNA [43]–[45] , however the downstream process following Endo V nicking has not been characterized . The bulge DNA in our crystal structures had a conformation similar to the bubble DNA produced by Endo V nicking , suggesting that RNase T might be the downstream exonuclease of Endo V , responsible for removing the last base-paired nucleotide at the 3′ end to release the single-stranded DNA or the damaged DNA bases , such as hypoxanthine , xanthine , and uracil ( Figure 2D ) . To test this possibility , we prepared the hypoxanthine-containing—that is , inosine-containing—heteroduplex DNA for examination of Endo V–dependent inosine excision repair in vitro [46] . The heteroduplex DNA plasmid contained the I-G base pair with an AlwNl cutting site and a potential XhoI cutting site . Once the inosine was restored to cytosine , the plasmid could be cleaved by AlwNl and XhoI into two linear double-stranded DNA molecules of 4 . 1 and 3 . 1 kilobases ( Figure 2D ) . The I-G–containing plasmid was then incubated with Endo V , RNase T , ligase , and the Klenow fragment exo− ( Polymerase I Klenow fragment with a defected 3′–5′ exonuclease activity ) . The inosine in the plasmid was restored to cytosine with a higher repair efficiency ( 86 . 7% ) as compared with those incubated with the wild-type DNA Polymerase I with a proofreading exonuclease domain ( 61 . 4% ) ( Figure 2C ) . The repair efficiency was positively correlated with the RNase T concentration and the time of incubation ( Figure S3 ) . Interestingly , ExoI and ExoX could not work with Endo V to restore the inosine to cytosine ( unpublished data ) . These results show that RNase T can work with Endo V in the Endo V–dependent DNA repair . Beside bubble/bulge DNA , RNase T also processed Y-structured DNA , which can be generated during various DNA repair pathways , such as mismatch repair and DNA recombination . However , it remained unknown how an exonuclease can specifically process the 3′-end tail of the intermediate Y-structured DNA . To reveal how RNase T binds and processes a Y-structured DNA , we co-crystallized RNase T with a Y-structured DNA and solved the complex crystal structure at a resolution of 1 . 9 Å ( Figure 3 and Figure 5 ) . In the crystal structure , the Y-structured DNA was bound to RNase T in a unique way , different from those of the bulge DNA and the duplex DNA that were bound between the two protomers with one strand of DNA bound to one protomer ( see Figure 4C ) . In contrast , both strands of the Y-structured DNA were bound to a single protomer , one Y-structured DNA bound to protomer A and the other DNA molecule bound to protomer B ( Figure 5 ) . This unique binding mode can avoid the hindrance produced by Phe29 , which might stack with the 5′-end base of the opposite nonscissile strand if the Y-structured DNA was bound in a way similar to that of a duplex DNA . Therefore , in this complex , the opposite nonscissile strand of the Y-structured DNA rotated about 180° to interact with the same protomer of RNase T ( see Figure 5B ) . Several residues , including Gln169 , Asp174 , Phe175 , and Ser177 , interacted with the nonscissile strand forming hydrogen bonds with the first and second phosphates in the 5′-overhang region , making it fit snugly onto the molecular surface of RNase T ( Figure S5 ) . The 3′ tail of the Y-structured DNA in the crystal structure had a dinucleotide 3′-CC sequence . However , the 3′-CC did not induce the C effect and inhibit the exonuclease activity of RNase T . A close look at the crystal structure of the Y-structured DNA complex showed that the 3′-end C did not interact with Glu73 as it did in the duplex complexes ( left panel in Figure S4B ) . Moreover , the scissile phosphate of the 3′-end C did not shift away from the active site , and as a result , two Mg2+ ions were bound in the active site in an active conformation ( right panel in Figure S4B ) . Therefore , due to the unique binding mode , the C effect did not occur when RNase T was bound to a Y-structured DNA with a 3′-end CC . In summary , this crystal structure reveals how RNase T binds a Y-structured DNA in a unique way and how it processes the 3′ tail of any sequence close to the duplex region ( see Movie S3 ) . Besides RNase T , two monomeric DnaQ-like exonucleases ExoI and ExoX also process DNA during DNA repair in E . coli . ExoI is suggested to play a role in BER [47] , mismatch repair [7] , [10] , [48] , [49] , UV-related repair [41] , [50] , and DNA replication [51] . ExoX is involved in mismatch repair [11] , [12] and UV-related repair [8] . ExoI binds and cleaves long single-stranded DNA [52] , whereas ExoX digests both single-stranded and double-stranded DNA [8] . The exonuclease activity of RNase T , ExoI , and ExoX probably overlap and are redundant in these pathways or they may target different substrates . To compare the substrate preference of RNase T to those of ExoI and ExoX , we further expressed and purified ExoI and ExoX for DNA digestion assays . The dynamic light scattering confirmed that ExoI and ExoX were monomeric proteins in contrast to RNase T , which existed as dimeric proteins ( Figure S5 ) . We found that RNase T , ExoI , and ExoX digested single-stranded 11-nucleotide DNA with similar efficiencies ( Figure 6A ) . However , in digesting Y-structured DNA with a short 3′ overhang , only RNase T and ExoX could process the 3′ overhang close to the duplex region , whereas ExoI did not digest Y-structured DNA at low concentrations but did digest Y-structured DNA randomly into small nucleotides at high concentrations ( Figure 6B ) . In digesting duplex DNA with a short 3′ overhang , RNase T processed DNA into a specific length close to the duplex region , generating a final duplex product with a 1-nt 3′ overhang at low concentrations ( Figure 6C ) . ExoX also digested duplex substrates but was less specific , generating various end products with 3′ overhangs of different lengths . On the contrary , ExoI could not digest the duplex substrates at the low concentration ( 0 . 02 µM ) ( Figure 6C ) . At the high exonuclease concentrations ( 0 . 1 and 1 µM ) , both ExoI and ExoX digested the duplex DNA substrates in the single-stranded and double-stranded regions into small nucleotides . However , RNase T still retained its specificity , only cleaving in the 3′ overhang but not in the duplex region ( Figure 6C ) . These results suggest that RNase T is a highly specific exonuclease that targets the 3′ overhang of structured DNA and produces a precise final product . On the other hand , ExoX is less specific and generates 3′ overhangs of different lengths in digesting duplex substrates with 3′ overhangs , whereas ExoI is specific for single-stranded DNA . Besides DNA digestion assays , the gel shift assays further showed that RNase T bound with similar affinities to single-stranded DNA , duplex DNA with 4- , 6- , and 10-nucleotide 3′ overhangs ( Figure S6 ) . In contrast , ExoI had lower binding affinity for duplex DNA with short 3′ overhangs , such as 4 and 6 nucleotides , in agreement with its low activity for these substrates . ExoX also preferred to bind to single-stranded DNA , but not duplex DNA with short 3′ overhangs at similar concentrations ( Figure S6 ) . Combining these results of the exonuclease activity and DNA-binding assays , we conclude that RNase T is the ideal exonuclease for trimming the 3′ overhang of structured DNA closely to the duplex region , including Y-structured DNA and duplex DNA , whereas ExoI and ExoX mainly process single-stranded DNA in DNA repair .
Our results suggest that RNase T is likely involved in the Endo V–dependent DNA repair pathway . Endo V is a conserved endonuclease playing critical roles in maintaining genome stability in prokaryotes and eukaryotes [53] . Endo V recognizes bubble DNA with mismatched base pairs and deaminated DNA lesions and initiates the Endo V–dependent DNA repair pathway that is independent of BER and MMR [42] , [44] , [45] , [53] , [54] . Moreover , Endo V nicks frameshift and structured DNA , such as insertion/deletion loops , hairpins , and flap DNA [43] , [55] . Frameshift DNA mutations are mistakenly generated during replication of repetitive sequences [41] , and as a result , the bulge DNA are produced by slipped misalignment of tandem repeats [56] , [57] . Rearrangements between tandem repeated DNA are important factors for genome instability and have been implicated in Friedreich ataxia in humans [58] , [59] . Slipped misalignment of tandem repeat DNA may cause palindrome-stimulated deletion or expansion by two RecA-independent recombination mechanisms—that is , single-strand annealing and replication slipped mispairing [60] , [61] . Single-strand-specific exonucleases , such as ExoI , ExoX , and RecJ , were reported to stabilize tandem repeats and limit RecA-independent recombination [56] , [62] . However , the downstream structure-specific exonuclease of Endo V for the further trimming of the DNA from the broken end has not yet been identified . Our structural and biochemical data of RNase T show that it can bind and digest bulge/bubble and Y-structured DNA . Moreover , RNase T can work with Endo V , DNA Polymerase I ( Klenow fragment exo− ) , and ligase to restore an inosine to cytosine in a heteroduplex DNA molecule in vitro . The crystal structures of RNase T bound with a blunt-end bulge DNA further show how RNase T removes the last base pair at the 3′ end by a special Phe-inserting binding mode . All these results suggest that RNase T may function as the downstream exonuclease of Endo V in alternative DNA repair . Taking together these lines of evidence , we suggest that RNase T likely recognizes these bulge and bubble DNA structures generated by Endo V and trims at the 3′ end of the nicked site to remove the last base pair next to the lesion . The single-stranded DNA or damaged DNA is then released for the next step of processing ( see Figure 7A ) . After removing the 3′-end base-paired nucleotide by RNase T , insertion DNA , hairpin DNA , and deaminated DNA lesions are released as single-stranded DNA . These single-stranded insertion DNA and hairpin DNA are probably further trimmed by the single-strand-specific exonucleases , such as ExoI and/or ExoX , with the help of single strand binding protein ( SSB ) and helicases . RNase T can further digest the 3′-end short overhang close to the duplex region in a way that we observed in the crystal structure of the Y-structured DNA complex . Deaminated DNA lesions are likely also removed by RNase T since we show that RNase T can digest single-stranded DNA containing oxidized bases and deaminated bases ( Figure 1C ) . It has been shown that the dimeric Exo I from Thermus thermophilus shares a sequence homology to RNase T and plays a similar role in digesting damaged DNA with methylated and deaminated bases [30] . It is very likely that Exo I from Thermus thermophilus is a functional homologue of RNase T and both of them play key roles in DNA repair . Therefore , after nicking by Endo V , the single-strand-specific exonucleases and structure-specific RNase T likely work together to further trim DNA from the broken end . After this trimming , polymerases and ligases can complete the DNA repair pathway . RNase T plays crucial roles in various DNA repair pathways , as shown by the sensitivity of the rnt knockout strain to a wide range of DNA-damaging agents . The indispensable role of RNase T might be due to its unique specificity for structured DNA that are generated during various DNA repair pathways . For instance , UV radiation can lead to single/double-strand breaks and base modifications , such as cross-linked pyrimidine dimers , photoproducts , and thymine glycols , and as a result , three different UV-induced DNA repair pathways are initiated [39] , [63] . In the first pathway , the base modification induced by UV may stall replication forks . In such a case , RecFOR and RecA bind to the lagging strand template and the invasion-containing leading strand to promote double-strand formation and repair by NER [64] . During this process , the Y-structured DNA formed on the leading strand requires a structure-specific exonuclease , very likely RNase T , to trim its 3′ overhang ( see Figure 7C ) . In the second pathway , UV radiation can induce single-strand breaks that can be repaired by homologous recombination [65] . During this process , Y-structured DNA is formed as an intermediate during gap-filling recombination ( see Figure 7D ) . ExoI was reported to promote this RecA-dependent 5′-end strand exchange by digesting the 3′ competitor strand [66] , [67] . However , ExoI cannot digest the 3′ overhang close to the duplex region , and thus most likely RNase T is responsible for processing the Y-structured DNA intermediates in the gap-filling recombination pathway . In the third pathway , the double-strand breaks induced by UV radiation are generally repaired by the RecA-dependent homologous recombination in bacteria [68] . This DNA repair pathway is initiated by RecBCD or RecJ to generate 3′ overhangs and is followed by RecA and RecFOR to promote strand invasion . DNA repair synthesis is then primed by PolI and PolIII from the invaded strand of the D-loop structure . RuvC resolvase cleaves the Holliday junctions that are synthesized after branch migration and LigA seals the nick to complete the homologous recombination [1] . In this process , ExoI was reported to affect RecBCD-mediated recombination [69] since the 3′–5′ exonucleases are required to degrade the 3′ tail of the intermediate Y-structured DNA after RecA dissociation [48] , [51] , [70] , [71] . Yet ExoI is not an appropriate exonuclease for digesting the 3′ tail near the duplex region . Based on our results , we suggest that most likely RNase T is involved in digesting the 3′ tail close to the duplex region in the UV-induced DNA homologous recombination ( Figure 7E ) . Moreover , in comparison with FEN1 , which is a flap endonuclease that binds DNA with one 3′-flap nucleotide and cleaves one nucleotide into the double-stranded DNA at the 5′ flap end to produce a ligatable product during DNA replication and repair [72] , RNase T is likely required to produce a DNA with a short 3′ overhang with one or two nucleotides that can be further processed in DNA homologous recombination . Besides UV-induced DNA repair , RNase T may also participate in other DNA repair processes that require a structure-specific 3′–5′ exonuclease , such as MMR . It has been shown that ExoI and ExoX are essential for methyl-directed mismatch repair in E . coli [7] , [10]–[12] , [49] , [50] . These two monomeric exonucleases are responsible for removing the 3′ single-stranded tail in Y-structured DNA during MMR ( see Figure 7B ) . However , they cannot process the 3′ single-stranded tail close to the double-stranded region [1] , [49] . ExoI only processes DNA with a long single-stranded region ( over 13 nucleotides ) in a processive manner , while a SSB stimulates its exonuclease activity [52] , [73] , [74] . ExoX , however , interacts with MutL during MMR and is not specific for processing Y-structured DNA [8] , [12] . On the other hand , the RNase T homolog Thermus thermophilus ExoI is suggested to excise the 3′ overhang of a Y-structured DNA and plays a role in MMR [30] . Therefore , it is very likely that RNase T processes the 3′ tail of the Y-structured DNA in MMR in E . coli . Our structure and biochemical assays show that the C effect does not occur when RNase T digests short 3′ overhang of a Y-structured DNA , and hence RNase T is capable of processing any sequence of the 3′ overhang of a Y-structured DNA during MMR . Therefore , we propose here that the monomeric ExoI and ExoX work with a helicase or SSB to process long 3′ tails , while the dimeric RNase T further trims the short 3′ overhang of Y-structured DNA during MMR . In conclusion , RNase T is a unique structure-specific exonuclease responsible for processing the 3′ ends of structured DNA in various DNA repair pathways . RNase T has an ideal dimeric architecture for binding and processing the 3′ end of various structured DNA in diverse ways , including duplex , bulge/bubble , and Y-structured DNA . Therefore , this intriguing exonuclease has multiple functions not only for processing duplex RNA during RNA maturation , but also processing bubble/bulge and Y-structured DNA during DNA repair . The diverse functions and different specificities of RNase T are closely correlated to its dimerization architecture and various binding modes against different substrates . We provide solid data here showing how the dimeric RNase T processes structured DNA in DNA repair that will serve as a model for understanding the molecular functions of thousands of members of DnaQ-like exonucleases .
Wild-type E . coli K-12 , single gene knockout ( Δrnt , Δsbcb , Δexox , Δpnp , and ΔrecJ ) strains used in the survival studies were from the Keio collection [75] . All E . coli cells were grown to an OD600 of 0 . 5–0 . 6 in LB medium at 37°C . To measure the acute sensitivity to hydrogen peroxide ( H2O2 ) , cells were exposed to 0 , 20 , 40 , and 80 mM H2O2 for 20 min . After removing H2O2 , cells were diluted 100-fold into 10 ml LB medium and further grown on a rotary shaker ( 200 r . p . m . ) at 37°C for the measurement of A600 ( OD ) at 60 min intervals . To measure the chronic sensitivity to H2O2 , MMS , mitomycin ( MMC ) , and 4NQO , serial dilutions of cells were spotted on plates containing indicated concentrations of the DNA-damaging agents and incubated overnight at 37°C . To measure the sensitivity against UV-C , serial dilutions of cells were spotted on plates and exposed to UV-C ( 254 nm ) in 20 J/m2 for 10 s by Hoefer UVC 500-Ultraviolet Crosslinker ( Hoefer Inc . ) . After UV-C irradiation , cells were incubated overnight at 37°C . The full-length rnt , sbcb , and exox genes were amplified by PCR using E . coli genomic DNA from JM109 or K-12 strains and cloned into NdeI/XhoI sites of expression vectors pET-28a ( Novagen ) to generate the N-terminal His-tagged fused recombinant proteins . The expression plasmid was transformed into the E . coli BL21-CodonPlus ( DE3 ) -RIPL strain ( Stratagene ) cultured in LB medium supplemented with 35 µg/ml kanamycin . Cells were grown to an OD600 of 0 . 5–0 . 6 at 37°C and induced by 0 . 8 mM IPTG at 18°C for 18 h . The harvested cells were dissolved in 50 mM Tris-HCl ( pH 7 . 5 ) buffer containing 300 mM NaCl and disrupted by a microfluidizer . Each exonuclease was purified by chromatographic methods using a HiTrap TALON column ( GE Healthcare ) , a HiTrap Heparin column ( GE Healthcare ) , and a gel filtration column ( Superdex 75 , GE Healthcare ) . Purified RNase T , ExoI , and ExoX samples were concentrated to 15–35 mg/ml in 300 mM NaCl and 50 mM Tris-HCl ( pH 7 . 0 ) . DNA oligonucleotides used for nuclease activity assays were synthesized ( BEX Co . , Tokyo , Japan or MDBio , Inc . , Taiwan ) and labeled at the 5′ end with [γ-32P]ATP by T4 polynucleotide kinase and purified on a Microspin G-25 column ( GE Healthcare ) to remove the nonincorporated nucleotides . Purified substrates ( 20 nM; see Table S1 for sequences ) were incubated with RNase T , ExoI , or ExoX at various concentrations in a buffered solution of 120 mM NaCl , 2 mM MgCl2 , and 50 mM Tris-HCl ( pH 7 . 0 ) at room temperature for 20–60 min . The reaction was quenched by addition of the stop solution ( 2× TBE ) and heating at 95°C for 5 min . Reaction samples were then resolved on 20% denaturing polyacrylamide gels and visualized by autoradiography ( Fujifilm , FLA-5000 ) . DNA binding affinities of RNase T , ExoI , and ExoX were measured by gel shift assays . The 5′-end 32P-labeled DNA substrates ( 20 nM ) were incubated with RNase T , ExoI , or ExoX in a solution of 100 mM NaCl , 30 mM EDTA , 10 mM EGTA , and 50 mM Tris-HCl ( pH 7 . 0 ) for 20 min at room temperature . The concentrations of each protein used in the assays were 0 , 5 , and 50 µM . Reaction samples were then resolved on 20% TBE gels ( Invitrogen ) and visualized by autoradiography ( Fujifilm , FLA-5000 ) . The E . coli strain NM522 . RS5033 was used in the assay as described in Fang et al . [76] . DNA polymerase I ( E . coli ) , the Klenow fragment exo− ( DNA polymerase I Klenow fragment without the 3′–5′ exonuclease activity ) , E . coli DNA ligase , T4 polynucleotide kinase , recombinant Endo V , and restriction endonucleases were obtained from New England Biolabs . RecBCD nuclease was purchased from EPICENTRE Biotechnologies . Construction of dI-G heteroduplex DNA substrates was prepared as described in Lee et al . [46] . M13mp18 replicative form DNA was hydrolyzed with HindIII and mixed with a 4-fold molar excess of M13LR1 viral DNA , followed by alkaline denaturation and re-annealing . The excess ssDNA was removed by hydroxyapatite ( Biorad ) chromatography and benzoylated naphthylated DEAE cellulose ( Sigma ) chromatography , and the linear dsDNA was removed by RecBCD nuclease ( EPICENTRE ) treatment . The resulting circular duplex DNA containing 22-nt gaps was further purified by Vivaspin 20 ultrafiltration ( GE Healthcare ) . A 5′-phosphorylated deoxyinosine-containing 22-bp synthetic oligonucleotides , 5′-AGCTCTIGAGGCTGCTGCTGCT-3′ ( Blossom Biotech ) , was then annealed to the gap and sealed by T4 DNA ligase in the presence of ethidium bromide . The covalently closed dG:I heteroduplex DNA was isolated by CsCl-EtBr density gradient centrifugation . The repair conditions were modified from Lee et al . [46] . The dI-G heteroduplex substrates ( 0 . 1 µg ) were incubated with repair enzymes ( 1 . 1 nM Endo V , 0 . 13 µM DNA polymerase I/0 . 13 µM Klenow fragment exo- , and 5 µM RNaseT ) for 30 min at 37°C in 15-µl reactions containing 50 mM NaCl , 10 mM Tris-HCl ( pH 7 . 9 ) , 10 mM MgCl2 , 1 mM dithiothreitol , 50 µg/ml bovine serum albumin , 0 . 3 mM NAD+ , and 125 µM of each dNTP . The reactions were terminated by heat inactivation at 75°C for 20 min . The DNA was then analyzed by restriction endonuclease hydrolysis and agarose gel electrophoresis . The gel images were captured by a gel documentation CCD camera ( UVP Ltd . ) using Viewfinder 3 . 0 , and band intensities were then measured by NIH Image J 1 . 45s software . Wild-type RNase T ( 25–35 mg/ml ) in 300 mM NaCl and 50 mM Tris-HCl ( pH 7 . 0 ) were mixed with different stem-loop DNA substrates in the molar ratio of 1∶1 . 2 . Detailed information for DNA sequences and crystallization conditions of the three structures is given in the Table S2 . All crystals were cryo-protected by Paraton-N ( Hampton Research , USA ) for the data collection at 100 K . X-ray diffraction data were collected using synchrotron radiations at SPXF beamline BL13B1 at NSRRC , Taiwan , or at the BL44XU beamline at SPring-8 , Japan . All diffraction data were processed by HKL2000 , and the diffraction statistics are listed in Table 1 . Structures were solved by the molecular replacement method using the crystal structure of E . coli RNase T ( PDB ID code 3NGY ) as the search model by program MOLREP of CCP4 . The models were built by Coot and refined by Phenix . Structural coordinates and diffraction structure factors have been deposited in the RCSB Protein Data Bank with the PDB ID codes of 4KB0 and 4KB1 for RNase T-bulge DNA complexes and 4KAZ for the RNase T-Y structured-DNA complex .
|
DNA repair relies on various enzymes , including exonucleases that bind and trim DNA at broken ends . However , we know little about how an exonuclease precisely selects and trims a DNA broken end in specific repair pathways . In this study , the enzyme RNase T , previously known for its involvement in processing RNA substrates , is shown to also possess DNase activity . RNase T is a DnaQ-like exonuclease and is characterized in this work as the exoDNase responsible for trimming the 3′ ends of structured DNA in various DNA repair pathways . Based on the high-resolution crystal structures of RNase T-DNA complexes , an insightful working model is provided showing how RNase T processes bulge , bubble , and Y- structured DNA in various DNA repair pathways . RNase T thus represents a unique structure-specific exonuclease with multiple functions not only in processing 3′ overhangs of duplex RNA during RNA maturation , but also in processing 3′ ends of bubble , bulge , and Y-structured DNA during DNA repair . These findings advance our understanding of the precise function of an exonuclease in DNA repair and suggest possible roles for thousands of members of DnaQ superfamily exonucleases in DNA repair and replication .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biophysics",
"biochemistry",
"biology"
] |
2014
|
Structural Insights Into DNA Repair by RNase T—An Exonuclease Processing 3′ End of Structured DNA in Repair Pathways
|
Genome sequence comparisons have highlighted many novel gene families that are conserved across animal phyla but whose biological function is unknown . Here , we functionally characterize a member of one such family , the macoilins . Macoilins are characterized by several highly conserved predicted transmembrane domains towards the N-terminus and by coiled-coil regions C-terminally . They are found throughout Eumetazoa but not in other organisms . Mutants for the single Caenorhabditis elegans macoilin , maco-1 , exhibit a constellation of behavioral phenotypes , including defects in aggregation , O2 responses , and swimming . MACO-1 protein is expressed broadly and specifically in the nervous system and localizes to the rough endoplasmic reticulum; it is excluded from dendrites and axons . Apart from subtle synapse defects , nervous system development appears wild-type in maco-1 mutants . However , maco-1 animals are resistant to the cholinesterase inhibitor aldicarb and sensitive to levamisole , suggesting pre-synaptic defects . Using in vivo imaging , we show that macoilin is required to evoke Ca2+ transients , at least in some neurons: in maco-1 mutants the O2-sensing neuron PQR is unable to generate a Ca2+ response to a rise in O2 . By genetically disrupting neurotransmission , we show that pre-synaptic input is not necessary for PQR to respond to O2 , indicating that the response is mediated by cell-intrinsic sensory transduction and amplification . Disrupting the sodium leak channels NCA-1/NCA-2 , or the N- , P/Q , R-type voltage-gated Ca2+ channels , also fails to disrupt Ca2+ responses in the PQR cell body to O2 stimuli . By contrast , mutations in egl-19 , which encodes the only Caenorhabditis elegans L-type voltage-gated Ca2+ channel α1 subunit , recapitulate the Ca2+ response defect we see in maco-1 mutants , although we do not see defects in localization of EGL-19 . Together , our data suggest that macoilin acts in the ER to regulate assembly or traffic of ion channels or ion channel regulators .
One of the most striking innovations in Metazoa is a nervous system with specialized nerve cells , pre- and post-synaptic structures , and associated signaling molecules . Neuronal signaling depends on complexes of multipass transmembrane proteins such as ion channels and G-protein-coupled receptors . Over the past few years several studies have identified specialized molecular machines in the endoplasmic reticulum by which particular complexes are assembled with appropriate stoichiometries and trafficked to their destination [1] . The emerging picture is that neurons have a highly specialized endoplasmic reticulum ( ER ) , allowing channels to undergo quality control prior to export . However the identity of such maturation complexes remains unclear except for a handful of channels . The extensive intracellular membrane system that makes up the ER varies , depending on cell type , but two domains , the rough and smooth ER ( RER and SER ) , can usually be distinguished . The RER is studded with ribosomes and mediates translocation of secretory proteins across the membrane and insertion of membrane proteins . The SER is implicated in lipid synthesis and regulation of Ca2+ storage and signalling . Whereas most ER proteins are found in both RER and SER , a subset of proteins involved in translocation of newly synthesized proteins across the ER membrane is highly concentrated in the RER [2] , [3] . In C . elegans neurons , RER proteins are concentrated in the cell body and excluded from dendrites and axons , whereas general ER proteins are found in both cell body and neurites [4] . Electron microscopy confirms that ribosomes and RER are abundant in the cell body of C . elegans neurons but rare in neurites , whereas smooth ER-like structures can be seen in axons and dendrites as well as the cell body . Over the last decade , genome sequencing projects have provided gene catalogs for animals representing a spectrum of metazoan phyla , including Placozoa [5] , Cnidaria [6] , Echinodermata [7] , Annelida ( http://genome . jgi-psf . org/Capca1/Capca1 . home . html ) and Chordata [8] . Genome-wide comparisons have identified human genes that are conserved across these animal phyla , and highlighted their shared structural features . However the biological function of many of these conserved gene families remains mysterious . Genetic studies in model organisms such as Drosophila and C . elegans have provided a powerful way to functionally characterize novel conserved genes . This is exemplified by discovery of ion channel families , e . g . TRP [9] , axon guidance pathways ( UNC-6/netrin; UNC-40/DCC; ROBO [10] ) and molecules involved in synaptic release ( e . g . UNC-13 [11] and UNC-18 [12] ) . In all these cases genetic studies in flies or worms were recapitulated in mammals and catalyzed subsequent vertebrate work . Here , we functionally characterize , for the first time , a member of a conserved family of proteins called macoilins . We find a macoilin gene in all available genome sequences of animals , from placozoa to man , but not in yeast or Dictyostelium . C . elegans macoilin , like mouse macoilin [13] , is expressed throughout the nervous system . In C . elegans expression begins embryonically at the time neurons are born , and persists to adulthood . Using antibodies and compartment specific markers we show that C . elegans macoilin is localized to the rough endoplasmic reticulum and is excluded from neurites . We identify multiple C . elegans mutants of macoilin and show that these have altered behavior , but normal development of the nervous system . Using Ca2+ imaging , we show that macoilin is required for cell-intrinsic neuronal excitability in the O2-sensing neuron PQR . This phenotype is mirrored in animals defective in EGL-19 , the sole C . elegans L-type voltage gated ion channel ( L-VGCC ) alpha1 subunit . Our data suggest that macoilin is involved in assembly or traffic of ion channels or ion channel regulators .
N2 , the laboratory wild-type strain of C . elegans , feeds in isolation; however most wild-collected strains of this species feed in groups [14]–[15] . N2 animals fail to aggregate because of a gain-of-function mutation in the neuropeptide receptor npr-1: if this receptor is knocked out they aggregate strongly [14][16] . To define genes that promote aggregation we mutagenized npr-1 ( null ) animals and sought non-aggregating mutants . One complementation group we identified comprised three recessive alleles , db1 , db9 and db129 , and defined a gene we called maco-1 ( for macoilin-1 see below ) . All three maco-1 mutations strongly suppressed aggregation and bordering behaviors ( Figure 1A ) and disrupted the ability of npr-1 animals to switch between roaming widely and dwelling locally according to ambient O2 levels ( Figure 1B ) [17] . maco-1; npr-1 mutants were healthy and displayed strong attraction to diacetyl and benzaldehyde ( Figure S1A , S1B ) , odorants that are detected by the AWA and AWC olfactory neurons respectively [18] . They also strongly avoided high osmotic potential , a response mediated by ASH nociceptive neurons ( Figure S1C ) [19] . Like N2 animals , maco-1 single mutants did not aggregate and only showed weak bordering behavior ( Figure S1D ) . Closer examination , however , revealed additional behavioral phenotypes associated with maco-1 mutations . We quantified these phenotypes in the presence of the N2 allele of npr-1 , since this is the standard genetic background . First , harsh touch to the head elicited significantly longer reversals in maco-1 mutants compared to N2 controls ( Figure 1C ) . Second , maco-1 mutants exhibited swimming defects [20] characterized by a decrease in the frequency of body bends ( Figure 1D ) , an increase in the amplitude of body bends ( data not shown ) , and coiling ( Figure 1E ) . maco-1 mutants also showed increased coiling when on an agar substrate ( data not shown ) . Interestingly , the maco-1 locomotory defects increased with age ( Figure S1E , S1F ) . Third , whereas N2 animals suppress egg-laying in the absence of food [21] , this inhibition was partly relieved in db9 and db129 mutant animals ( Figure 1F ) . These subtle but pleiotropic defects of maco-1 mutants suggest deficits in multiple neural circuits . We mapped maco-1 to a 30 kb interval on Chromosome I , close to D2092 . 5 , a previously uncharacterized gene . DNA sequencing revealed that all three maco-1 alleles disrupted D2092 . 5 ( Figure 2A ) . The db1 allele modified the splice donor site of intron 9 ( G6024A ) ; db9 changed the arginine codon at position 534 to a stop codon and is predicted to truncate the MACO-1 protein ( Figure 2D ) ; the db129 allele was associated with a mutation in the splice acceptor site of intron 2 ( G595A ) . Together , these data suggest that maco-1 corresponds to D2092 . 5 . This was confirmed by transgenic rescue of the maco-1 mutant phenotypes with a wild-type D2092 . 5 transgene ( Figure S2A–S2D ) . None of our maco-1 alleles were unambiguously null mutants . However the premature stop codon associated with the db9 allele would be expected to cause nonsense mediated degradation of maco-1 mRNA , as well as truncating half the MACO-1 protein , and therefore to be a strong loss-of-function mutation . Consistent with this , the phenotypes of maco-1 ( db9 ) /maco-1 ( db9 ) and maco-1 ( db9 ) /qDf16 were similar ( Figure S1G ) ; qDf16 is a large deletion that spans the maco-1 interval ( see Materials and Methods ) . cDNA analyses indicated that the D2092 . 5 gene can encode two splice isoforms by alternate splicing at exon 5 ( Figure 2A ) . The proteins encoded by the resulting mRNAs were 908 ( MACO-1a ) and 897 ( MACO-1b ) amino acids long . Blast searches identified these proteins as homologues of vertebrate macoilins . Reciprocally , searching the C . elegans genome with vertebrate macoilins identified only one homologue , D2092 . 5 ( Figure 2D ) . At least one macoilin can be found in every animal genome sequenced so far ( Figure 2B , 2C ) . In some fish lineages but not in other vertebrates , a second MACO homologue can be found ( MACO-2 ) ; MACO-2 probably arose during the genome duplication event that is thought to have occurred in teleost fish . Despite this ubiquity , little is known about this protein family . The only macoilin previously investigated is the mouse homologue , highlighted because it is expressed differentially between wild-type mice and reeler mutants [22] , [13] . In situ hybridization indicates that mouse macoilin mRNA is highly expressed in all neuronal differentiation fields from embryonic stage E12 . 5 to birth [13] . After birth ( PG10 ) , expression decreases but remains associated with some neurons such as cerebellar granule cells , olfactory mitral and granule cells , pyramidal neurons in the hippocampus , and granule cells in dentate gyrus [13] . No significant expression of mouse macoilin has been reported outside the nervous system . Comparison of macoilin sequences across phyla highlighted several common features ( Figure 2D ) . All MACO-1 homologs were predicted to have at least three transmembrane domains towards the N-terminus and at least two coiled-coil domains towards the C-terminus . The exact number of transmembrane and coiled-coil domains varied depending on the prediction algorithm and species ( data not shown ) . At the sequence level the transmembrane and coiled-coil domains were the most conserved parts of the protein ( Figure 2D ) . To determine where maco-1 was expressed , we created transgenic C . elegans that co-expressed maco-1 and gfp coding sequences as a polycistronic message from the maco-1 promoter . These animals first showed GFP fluorescence 6 hours after the first cell division ( Figure S3A ) . By late embryogenesis and in the L1 larva , fluorescence was visible throughout the nervous system ( Figure S3A ) . In adults , the GFP signal remained pan-neuronal , with very occasional expression in other tissues ( Figure S3B ) . Thus , C . elegans MACO-1 , like its mouse homologue , is expressed widely and almost exclusively in the nervous system . To study endogenous MACO-1 , we raised polyclonal antibodies against the C-terminus of both its protein isoforms . N2 worms stained with the antibody showed bright expression in most or all C . elegans neurons ( Figure 3A–3C and Figure S4A ) . In contrast , maco-1 ( db9 ) mutants which have a nonsense mutation that truncates MACO-1 upstream of the epitope sequence showed no signal in the nervous system ( Figure 3D , 3E , and Figure S4B , S4C ) . Interestingly , MACO-1 antibody staining was restricted to the neuronal cell body: no staining was observed in dendrites , in the synapse-rich axon bundles that make up the nerve ring , or in the axons comprising the ventral and dorsal cords ( Figure 3A , 3B ) . To confirm that MACO-1 was absent from synapses , we co-stained worms expressing the synaptic marker SNB-1-GFP in GABAergic neurons ( juIs1 ) with anti-MACO-1 and anti-GFP antibodies: the two markers did not show co-localization ( Figure 4A–4D ) . Moreover , worms co-expressing juIs1 and psnb-1::maco-1-mcherry showed that transgenic MACO-1-cherry was also restricted to the cell body of neurons , overlapping only with the SNB-1-GFP signal in the cell body of GABAergic neurons ( Figure 4N ) . Both transgenic and endogenous MACO-1 was also found restricted to neuronal cell bodies at embryonic developmental stages ( Figure S5 ) . The striking localization of MACO-1 to the cell body raised the possibility that it resides in a specific membrane compartment . To investigate this we created different transgenic lines that expressed organelle-specific markers in a subset of neurons , using the glr-1 promoter ( glr = glutamate receptor ) ( Figure 4E–4M; Figure S6 ) . The markers used ( a kind gift of M . M . Rolls , Pennsylvania State University ) were phosphatidylinositol synthase ( PIS ) for the general endoplasmic reticulum ( ER ) , translocating chain-associating membrane protein ( TRAM ) for rough ER , Emerin for the nuclear envelope , and Mannosidase ( MANS ) for the Golgi . All markers were tagged at the N-terminus with YFP . Co-immunostaining of MACO-1 and YFP revealed partial co-localization between MACO-1 and the ER general marker , YFP-PIS ( Figure 4E–4I ) . Within the ER , MACO-1 was further co-localized with a marker restricted to the rough ER , YFP-TRAM , and nuclear envelope , YFP-Emerin ( Figure 4J-4M and Figure S6A–S6D ) . Similar ER localization was seen in embryonic stages ( Figure S6 ) . No significant co-localization was observed with the Golgi-specific marker , YFP-MANS ( Figure S6E–S6H; however the cell bodies of C . elegans neurons are small ( 2 microns ) , making it difficult to exclude the possibility that there is a low amount of MACO-1 in Golgi . These co-localization experiments suggest that MACO-1 predominantly resides in the ER of neurons , in particular in rough ER and nuclear envelope . We observed no gross morphological defects in maco-1 worms in any of the sub-cellular compartments expressing the YFP tagged constructs described above ( data not shown ) . The broad neuronal expression patterns of C . elegans and mouse macoilins suggest that this protein family has a general role in the development or function of the nervous system . To elucidate this role , we first examined the anatomy of the nervous system in maco-1 ( db9 ) mutants using neuron-specific GFP reporters . We examined mechanosensory neurons , chemosensory neurons , and GABAergic motor neurons . We detected no overt abnormalities in the cell bodies , axons , dendrites or cilia of any neuron we examined ( Figure S7A–S7C and data not shown ) . We next asked if maco-1 regulates neuronal polarity ( i . e . the placement of synapses ) or axonal trafficking of synaptic vesicles . To test this we visualized synaptic vesicles in live animals using a fluorescently-tagged synaptic vesicle marker , synaptobrevin-GFP ( SNB-1::GFP ) . We saw no defects either in the GABAergic DD motor neurons or in the URX O2-sensing neurons ( Figure S7D–S7G and data not shown ) . These results suggest that MACO-1 is not required for correct establishment or maintenance of neuron polarity . Precursor vesicles containing synaptic proteins are generated at the cell body and transported to synapses by microtubule-based motor proteins [23] . In C . elegans , this transport requires the KIF1A kinesin homologue unc-104 [23] . In unc-104 mutants tagged synaptobrevin expressed from the punc-25::snb-1::gfp transgene is retained in the cell bodies of the DD and VD motor neurons [24] . Macoilins have been proposed to function in axonal traffic [13]; however , in adult maco-1 ( db9 ) worms expressing the punc-25::SNB-1::GFP marker , the tagged synaptobrevin was still localized along the ventral and dorsal nerve cords ( Figure 6A and 6C ) . This suggests that MACO-1 is not essential for transport of synaptic vesicles . C . elegans synaptic function can be assayed by studying responses to the acetylcholinesterase inhibitor aldicarb and the acetylcholine receptor agonist levamisole [20] , [25] . Aldicarb inhibits acetylcholinesterase ( AChE ) , leading to accumulation of acetylcholine at the neuromuscular junction ( NMJ ) , overstimulation of acetylcholine receptors , and paralysis of wild-type animals . Mutants defective in synaptic release have reduced acetylcholine accumulation and are therefore resistant to aldicarb [20] . However these mutants retain sensitivity to levamisole , which directly activates post-synaptic acetylcholine receptors . In contrast , mutants defective in postsynaptic responses to ACh are resistant to both aldicarb and levamisole [26] . maco-1 ( db9 ) mutants were resistant to aldicarb but sensitive to levamisole ( Figure 5A , 5B ) , suggesting they have presynaptic defects in neurotransmission . The aldicarb resistance of maco-1 mutants prompted us to examine synapse structure in these animals more closely . GABAergic type D motor neurons form neuromuscular junctions ( NMJs ) with ventral and dorsal body wall muscles [27] . We visualized the presynaptic terminals of these neurons using the punc-25::snb-1::gfp transgene juls1 [28] . Wild-type animals bearing juIs1 have SNB-1::GFP puncta of uniform shape and size distributed evenly along the ventral and dorsal nerve cord . These puncta correspond to the presynaptic termini of the 13 VD and 6 DD neurons , respectively . We measured puncta size and number along a 100 µm section of the ventral nerve cord ( Figure 6 ) . In wild-type animals , average puncta area in the ventral nerve cord was 1 . 58±0 . 13 µm2 , with an average of 23 . 28±0 . 79 puncta per 100 µm ( n = 22 animals ) . maco-1 mutants had fewer puncta that tended to be larger: the average size in the ventral nerve cord was 2 . 62±0 . 54 µm2 with an average of 19 . 87±1 . 18 µm puncta per 100 µm ( n = 23 ) ( Figure 6A–6D ) . These data suggest that maco-1 influences pre-synaptic structure . We next investigated whether maco-1 mutants exhibit active zone defects , using SYD-2::GFP [29] and UNC-10::GFP [30] as markers ( Figure 6E–6L ) . Both fusion proteins were expressed in the GABAergic VD and DD motorneurones from the unc-25 promoter . Wild-type animals carrying the punc-25::syd-2::gfp transgene hpIs3 have regularly-sized and spaced puncta along the dorsal and ventral nerve cords , with an average punctal area of 0 . 32±0 . 017 µm2 and on average 36 . 78±1 . 96 ( n = 17 ) puncta per 100 µm . In maco-1 ( db9 ) adult animals the number of SYD-2::GFP puncta along the dorsal nerve cord was increased but their size was similar to wild-type ( Figure 6I–6L ) . The average area of puncta in maco-1 mutants was 0 . 31±0 . 014 µm2 , with an average of 45 . 44±1 . 51 ( n = 23 ) puncta per 100 µm . Wild-type animals carrying the punc-25::unc-10::gfp transgene hpIs61 have uniformly shaped and evenly distributed puncta along the ventral and dorsal cord . In maco-1 ( db9 ) adult animals , there was a reduced number of puncta but they were larger than in wild-type animals in the ventral nerve cord . The average punctum area in the ventral nerve cord for wild-type animals was 0 . 55±0 . 035 µm2 , with an average of 35 . 16±0 . 99 ( n = 13 ) puncta per 100 µm . Average puncta size in maco-1 mutants was 0 . 85±0 . 12 µm2 , and the average number of puncta was 28 . 12±2 . 29 ( n = 12 ) per 100 µm ( Figure 6E–6H ) . Together these data suggest that loss of maco-1 alters the structure of the synaptic active zone , but the effects are subtle . We also examined the periactive zone , using the marker RPM-1::GFP [31] . This region just surrounds active zones and has been proposed to regulate synapse growth [32] . We found a slight increase in the size and number of puncta in maco-1 mutants ( 0 . 75±0 . 068 µm2 and 37 . 24±3 . 376 , n = 10 ) , compared to wild-type animals ( 0 . 63±0 . 036 µm2 and 33 . 09±2 . 7 , n = 16 ) . However , these differences were not significant ( p = 0 . 1 and p = 0 . 34 , respectively ) , suggesting that the periactive zone was not disorganized in maco-1 mutants . Synapse development can be influenced by neural activity [33] . This prompted us to explore if the subtle synaptic defects in maco-1 reflected altered neuronal excitability . We focused our analyses on the O2-sensing neuron PQR , since a subset of the phenotypes of maco-1 mutants resembled those associated with loss of O2 sensing neurons [34] . To image Ca2+ transients we used the cameleon reporter YC3 . 60 [35] expressed from the gcy-32 promoter [15] . Baseline Ca2+ levels in 7% O2 were similar in wild type and maco-1 mutants , suggesting that PQR neurons were not chronically depolarized in maco-1 mutants . However the Ca2+ rise seen in wild type when O2 is raised to 21% was absent in most maco-1 mutant animals ( Figure 7A ) . These data suggest that maco-1 is required for efficient activation of PQR neurons in response to a rise in O2 . To explore this further we first asked if pre-synaptic input was required for PQR neurons to respond to O2 stimuli . Null mutations in unc-13 or unc-31 CAPS , which disrupt release of synaptic vesicles and dense core vesicles respectively , did not significantly alter Ca2+ transients in PQR to a 7 – 21% O2 upstep or a 21 to 7% downstep ( data not shown ) . This suggests that Ca2+ fluxes in PQR reflect cell-intrinsic responses to the O2 stimuli , and that loss of maco-1 disrupts either primary sensory transduction of ambient O2 or amplification of the sensory potential . O2-stimulated Ca2+ influx in PQR requires the atypical soluble guanylate cyclases GCY-35 and GCY-36 . These soluble guanylate cyclases are themselves O2 sensors and activate a cGMP-gated ion channel [36]-[37] . Consistent with this , in a separate study we have shown that a rise in O2 stimulates a rise in cGMP in PQR neurons ( A . C . and M . dB , in preparation ) . Mutations in maco-1 did not alter PQR cGMP responses to an O2 stimulus , suggesting that O2 sensing by GCY-35/36 was unaffected ( A . C . and M . dB , in preparation ) . Previous work has shown that tax-4 , which encodes a cGMP-gated cation channel alpha subunit is required for the O2-evoked Ca2+ transients in PQR [15] . To explore how depolarization evoked by the cGMP channel is amplified and leads to Ca2+ influx in the cell body we imaged PQR responses to O2 stimuli in animals defective in various ion channels . The C . elegans genome does not appear to encode voltage-gated sodium channels . Instead , electrical signals are thought to propagate via voltage-gated Ca2+ channels and cation leak channels [38] , [39] . C . elegans encodes 3 voltage gated Ca2+ channel α1 subunits: egl-19 ( CaV1 , L-type ) , unc-2 ( CaV2 , N- , P/Q , R-type ) and cca-1 ( CaV3 , T-type ) [40] . It also encodes 2 homologs of the vertebrate cation leak channel NALCN that regulates neuronal excitability [38] . Animals mutant for the UNC-2 P/Q-like voltage-gated Ca2+ channel ( VGCC ) [41] , the T-type channel CCA-1 [42] , or double mutant for the NCA-1and NCA-2 NALCN-like leak channels [38] showed overtly wild-type Ca2+ transients in the cell body of PQR in response to a 7 to 21% O2 shift ( Figure 7C and data not shown ) . This is consistent with previous results in other neurons that suggest these channels contribute to Ca2+ signals at synapses and axons , but are not essential for Ca2+ changes in the cell body [38] , [39] . In contrast , animals with partial loss-of-function mutations in the EGL-19 L-type VGCC showed frequent failure of cell body Ca2+ transients ( Figure 7B ) . L-type VGCCs have previously been shown to contribute to dendritic Ca2+ currents both in C . elegans [39] and vertebrates [43] . Consistent with these imaging results , egl-19 ( ad1006 ) ; npr-1 ( ad609 ) double mutants animals failed to aggregate . Together , our Ca2+ imaging results suggest that MACO-1 acts in the endoplasmic reticulum to promote assembly and/or traffic of either a cGMP-gated cation channel that contains the TAX-4 alpha subunit , or of an L-type Ca2+ channel containing the EGL-19 α1 subunit , or of another as yet unknown regulator that modulates O2-evoked Ca2+ entry into PQR . To investigate the first possibility we made transgenic animals that expressed a functional GFP-tagged TAX-4 protein in the AQR , PQR and URX neurons , and compared the localization of this channel subunit in npr-1 and maco-1; npr-1 mutant animals . We saw enrichment of TAX-4-GFP in the sensory endings of the O2-sensing neurons , as expected for a sensory transduction channel ( Figure S8 ) . We also observed TAX-4-GFP in the cell body and on axons and dendrites . However we found no effect of loss-of-function mutations in maco-1 on this distribution pattern ( Figure S8 and data not shown ) . These data suggest maco-1 is not required for TAX-4 to be exported from the ER , although they do not rule out a potential role in the function of a TAX-4-containing channel . Next , we transgenically expressed EGL-19 protein that is N-terminally tagged with GFP from its endogenous promoter , and examined its localization in wild type and maco-1 mutant animals . As expected , GFP-EGL-19 was expressed very broadly , and both in muscles and neurons ( Figure S9 ) [44] . In neurons GFP-EGL-19 was enriched in sensory endings and in cell bodies . However we did not see any striking defects in the EGL-19 localization pattern in maco-1 mutants ( Figure S9 ) . This does not rule out that MACO-1 modulates the function of an EGL-19-containing channel , but it does suggest that if it has a role it involves only a subset of EGL-19-containing channels; alternatively maco-1 regulates function of other , as yet unknown , ion channels .
Macoilins are a conserved family of multipass transmembrane proteins whose function has been mysterious . Members of the family can be found in eukaryotes that have a recognizable nervous system , from placozoa to humans , but not in yeast or Dictyostelium . Macoilins are expressed broadly but specifically in the nervous system . C . elegans macoilin is absent from neurites and is localized to the RER suggesting that it is involved in folding , assembly , or traffic of secreted or transmembrane proteins . The structure of macoilin contains two conserved regions: an N-terminal part that includes multiple transmembrane domains , and a C-terminal region that has coiled coil domains; the transmembrane domains are the most highly conserved parts of the protein . This combination of structural motifs is reminiscent of that of RIC-3 and its orthologues , which are implicated in assembly and traffic of nicotinic acetylcholine receptors in C . elegans [45] and of nicotinic acetylcholine receptors and 5-HT3 receptors in vertebrates [46]–[47] . Like macoilin , RIC-3 is expressed broadly in the nervous system and is an ER membrane protein with a coiled-coil region towards the C-terminus . Macoilin mutants exhibit defects in cell intrinsic neuronal excitability , not only in PQR ( this study ) but also in other neurons ( see associated paper by Miyara et al . ) . Previous work has reported that neural activity levels regulate the morphology of certain synaptic connections in C . elegans [33]; the synaptic morphology defects of maco-1 mutants could therefore reflect loss of neuronal excitability . A simple hypothesis is that macoilin acts in the endoplasmic reticulum of neurons to promote the folding , assembly , or traffic of ion channels or ion channel regulators that control excitability of neurons . What might these targets be ? Since baseline Ca2+ levels are normal in maco-1 mutants we do not think loss of macoilin disrupts function of ion pumps that keep neurons hyperpolarized . Instead our data point towards compromised signal transduction or signal amplification downstream of the GCY-35/GCY-36 O2-sensing soluble guanylate cyclases . As far as we can tell the cGMP-gated ion channel that transduces the O2-evoked cGMP rise in PQR , and which includes the TAX-4 α subunit , is appropriately expressed and localized in maco-1 mutants , although we cannot exclude the possibility that its function is somehow compromised . cGMP channels are expressed in only a small subset of C . elegans neurons [48] and some of these are clearly functional in maco-1 mutants ( e . g . AWC ) ; by contrast MACO-1 is expressed throughout the nervous system , not only in C . elegans but also in mouse . The L-type VGCC α1 subunit EGL-19 is also required for O2-evoked responses in PQR , and is expressed widely both in the nervous system and in muscle ( this work; [44] . Loss-of-function mutants of egl-19 have much more severe phenotypes than maco-1 mutants: egl-19 ( null ) mutants die as embryos . This discrepancy in phenotype makes it unlikely that MACO-1 is critical for function of all EGL-19 containing channels . Consistent with this , mutations in maco-1 do not appear to disrupt localization of GFP- EGL-19 either in muscle or in neurons . However it remains possible that MACO-1 regulates assembly of particular subtypes of EGL-19–containing channels . L-type VGCC are composed of multiple subunits and it is the precise combination of subunits that determines the channel's regulatory properties; additionally egl-19 mRNA itself is alternatively spliced close to its C-terminus , in a region implicated in Ca2+ feedback regulation [49] . An alternative scenario is that MACO-1 regulates an as yet undiscovered pathway that helps amplify the depolarization initiated by cGMP-gated ion channel activation . Identifying proteins that interact with macoilin or mutants that recapitulate the maco-1 phenotypes will allow these hypotheses to be tested to help further unravel the function of this novel family of nervous system proteins .
Strains used were maintained as described previously [50] and are listed in Text S1 ) . db1 and db9 were isolated as suppressors of aggregation from a screen of 20 , 000 haploid genomes; details of the screen will be described elsewhere . db129 was isolated in a non-complementation screen using the db1 allele . The db1 mutation was mapped to a 30 kb interval at the centre of Chromosome I between the SNP markers in cosmids F48A9 and D2092 using a combination of three-factor mapping and SNP genotyping [51] . PSI-Blast was used to search for Macoilin protein sequences , using human Macoilin as probe , at the NCBI ( www . ncbi . nlm . nih . gov ) , Joint Genome Institute ( www . jgi . doe . gov ) , ENSEMBL ( www . ensembl . org ) , and the Sanger Institute ( www . sanger . ac . uk ) . From approximately 100 sequences retrieved ( e-value > e-10 ) , a subset was obtained after removing splice variants , and redundant sequences . The amino acid sequences were aligned using various programs run under the umbrella of the M-Coffee server [52] . The multiple alignment was visually inspected and curated using BioEdit [53] ( Figure S10 ) . Un-rooted phylogenetic trees were generated using a Neighbour-Joining method [54]; the robustness of the nodes was verified with 10000 bootstrap replicates using the program Phylo-Win [55] . The cDNA sequence of maco-1 was obtained by sequencing clone yk1296a05 from the Kohara collection and by using RT-PCR . Briefly , N2 mixed stage animals were extracted with Trizol , and 1–5 µg of purified total RNA reverse transcribed using an oligo-d ( T ) primer and SUPER RTase at 42 °C for 1 h . Primers specific for maco-1 exons and the SL1 spliced leader sequence were used to amplify maco-1 cDNA , and the PCR products sequenced . The maco-1 expression construct was generated using the Gateway system ( Invitrogen ) [56] . The Destination vector included 4 kb of the sequence upstream of the maco-1 start site but omitted 318 bp between the trans-splice site and the initiation codon . The Entry vector places maco-1 cDNA plus 9 bp of the sequence upstream in an artificial operon with gfp [34] . This construct was sequenced and injected at 50 ng/µl with lin-15 ( + ) as the co-injection marker . Transgenic rescue: The fosmid WRM0640bE08 , containing D2092 . 5 , was injected into the strain AX129 , maco-1 ( db9 ) ;npr-1 ( ad609 ) at a concentration of 2 ng/µl , with 50 ng/µl of punc-122::gfp as a co-injection marker . Further transgenic rescue experiments were carried out using a PCR amplified genomic DNA fragment containing the D2092 . 5 gene , including 4 kb upstream of the initiation codon and 1 kb after the stop codon . This PCR product was injected into the AX59 , maco-1 ( db9 ) ; npr-1 ( ad609 ) strain at a concentration of 2 ng/µl , using pmyo-2::gfp as a co-injection marker ( 4 ng/µl ) and 1 kb-ladder ( 96 ng/µl ) as carrier . Sub-cellular markers ( a kind gift of Melissa M . Rolls , Penn State University ) were used as described [4] . The plasmids used were C24F3 . 1a ( pglr-1::yfp-TRAM ) , Y46G5a . 5 ( pglr-1::yfp-PIS ) , F558H1 . 1 ( pglr-1::yfp-MANS ) and M01D7 . 6 ( pglr-1::yfp-EMERIN ) . These plasmids were individually injected at 4 ng/µl into AX206 , lin-15 ( n765ts ) animals with a lin-15 ( + ) co-injection marker ( 40 ng/µl ) . All primer sequences used are available upon request . The pgcy-32::tax-4-gfp transgene was made using Gateway; tax-4 cDNA was tagged at the 3'end with gfp and injected ( at 10 ng/ul ) with a lin-15 ( + ) co-injection marker ( 40 ng/ul ) into npr-1 ( ad609 ) lin-15 ( n765ts ) animals . A fosmid containing the full-length egl-19 gene was modified by recombineering so as to express N-terminally GFP tagged EGL-19 from its endogenous control sequences . The recombineered fosmid was co-injected at 5 ng /ul with a pgcy-32::mcherry co-injection marker ( 25 ng/ul ) and carrier DNA ( DNA ladder at 70 ng/ul ) . Mix-staged worms were stained following the modified Ruvkun and Finney method [57] . Primary antibodies were used at dilutions of 1/50 and 1/500 for mouse monoclonal anti-GFP antibody ( clones 7 . 1 and 13 . 1; Roche , Germany ) and rabbit polyclonal anti-MCL , respectively . Antibodies were incubated at 4 °C for 16 hrs with gentle mixing . The secondary antibodies Alexa Fluor 546 nm goat anti-rat IgG ( H+L ) ( Invitrogen , UK ) and AlexaFluor 488 nm goat anti-mouse ( Invitrogen , UK ) were used at a final dilution of 1/500 ( 4 mg/ml ) and 1/250 ( 10 mg/ml ) , respectively; DAPI was added to a final concentration of 5 mM . After an incubation of 2 hrs at room temperature , worms were thoroughly washed with AbB Buffer ( the details of buffer composition can be found in the Text S1 ) , mounted in agarose and imaged . Live animals were anaesthetized with 10 mM sodium azide , mounted on 2 % agarose pads , and examined under epifluorescence using a Zeiss Axioskop fluorescent microscope . Confocal images were taken using a Radiance Plus Confocal Scanning System ( Bio-Rad ) . The images were processed and analyzed with LaserSharp2000 software ( Bio-Rad ) . The different GFP markers were visualized in the different backgrounds as described previously [29] , [58] . Measurements of GFP puncta were performed on confocal images . Briefly , confocal images were projected into a single plane using the maximum projection method and exported as a tiff file with a scale bar . Fluorescence intensity , number of puncta , total fluorescence and punctum area were measured in ImageJ . These numbers were exported to Microsoft Excel for statistical analyses using Student's two-tailed t-test . Aggregation assays were done as previously described [14] . Egg-laying assays followed [59] with the following modifications . Worms were synchronized and young adults picked to unseeded plates to remove adhering food before transfer to plates seeded with 50 µl of E . coli OP50 or to un-seeded plates . Rings of 100 µl of 4M D-Fructose were painted on no-food plates to trap animals . Worms were left for one hour , then removed and eggs counted . Plates in which worms could not be found were discarded ( around 10% in plates without food ) . Harsh touch was assayed by poking animals with a platinum wire pick . To analyze swimming defects , single worms were transferred to M9 media and left to equilibrate for a minute; head swings were counted during 10 second intervals; for coiling we counted the number of times the worm's nose touched the mid body in one minute . The results of the behavioral assays were analyzed using a two-tailed t-test . Aldicarb and levamisole assays were done as described [25] . Briefly , sensitivity to 1 mM aldicarb ( Chem Services ) or 0 . 4 mM levamisole was determined by assaying the time course of the onset of paralysis following acute exposure of a population of animals to these drugs . In each experiment , 25 worms were placed on drug plates and prodded every 10 min over a 2 h period to determine if they retained the ability to move . Worms that failed to respond at all to the harsh touch were classified as paralyzed . Each experiment was repeated five times . Ca2+ responses of PQR neurons to O2 stimuli were imaged as described previously ( Persson et al 2009 ) on an inverted microscope ( Axiovert , Zeiss ) , using a 40× C Apochromat lens and Metamorph acquisition software ( Molecular Devices ) . To measure Ca2+ we used the ratiometric FRET sensor YC3 . 60 [35] . Briefly , worms were glued to agarose pads using Nexaband glue ( WPI Inc ) and placed under the stem of a Y-chamber microfluidic device . Photobleaching was limited by using a 2 . 0 optical-density filter and a shutter to limit exposure time to 100 ms per frame . An excitation filter ( Chroma ) restricted illumination to the cyan channel . A beam splitter ( Optical Insights ) was used to seperate the cyan and yellow emission light . The ratio of the background-subtracted fluorescence in the CFP and YFP channels was calculated with Jmalyze [60] . Fluorescence ratio ( YFP/CFP ) plots and measurements of mean baseline ratios and mean peak ratios were made in Matlab ( The MathWorks ) . Movies were captured at 2 frames per second . Average Ca2+ traces were compiled from at least six recordings per condition made across two or more days . Whenever the data fitted a normal distribution ( p<0 . 05 , Kolmogorov-Smirnov ) a two-sample ( unranked ) t-test was used . For non-normally distributed data , the non-parametric Kolmogorov-Smirnov ( K-S ) -test was used .
|
The human genome project has given us a catalog of the genes that make a human; however , the function of about 40% of these genes remains elusive . Many of these mysterious genes have relatives in simpler organisms like worms and flies , where their function can be studied much more easily than in a mammal . Here , we investigate one such family of genes , called macoilins , using the worm C . elegans . We show that worm macoilin , like mouse macoilin , is expressed widely but specifically in nerve cells . We create worms in which the macoilin gene is defective and show that , although they retain a nervous system that looks normal , they have behavioral defects . We show that these behavioral defects reflect an inability of nerves to signal efficiently . Nerve signalling relies on calcium channels and the defect of macoilin mutants resembles that of animals defective in a particular calcium channel component . We find that in nerve cells the macoilin protein resides specifically in the “factory” that assembles nerve signalling molecules , including calcium channels . Our results suggest that macoilin either directly helps assemble an ion channel or is needed to make a channel regulator . Our work in worms provides a blueprint to investigate the function of macoilins in mammals .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/gene",
"function",
"neuroscience/neuronal",
"signaling",
"mechanisms"
] |
2011
|
Macoilin, a Conserved Nervous System–Specific ER Membrane Protein That Regulates Neuronal Excitability
|
The molecular determinants that render specific populations of normal cells susceptible to oncogenic reprogramming into self-renewing cancer stem cells are poorly understood . Here , we exploit T-cell acute lymphoblastic leukemia ( T-ALL ) as a model to define the critical initiating events in this disease . First , thymocytes that are reprogrammed by the SCL and LMO1 oncogenic transcription factors into self-renewing pre-leukemic stem cells ( pre-LSCs ) remain non-malignant , as evidenced by their capacities to generate functional T cells . Second , we provide strong genetic evidence that SCL directly interacts with LMO1 to activate the transcription of a self-renewal program coordinated by LYL1 . Moreover , LYL1 can substitute for SCL to reprogram thymocytes in concert with LMO1 . In contrast , inhibition of E2A was not sufficient to substitute for SCL , indicating that thymocyte reprogramming requires transcription activation by SCL-LMO1 . Third , only a specific subset of normal thymic cells , known as DN3 thymocytes , is susceptible to reprogramming . This is because physiological NOTCH1 signals are highest in DN3 cells compared to other thymocyte subsets . Consistent with this , overexpression of a ligand-independent hyperactive NOTCH1 allele in all immature thymocytes is sufficient to sensitize them to SCL-LMO1 , thereby increasing the pool of self-renewing cells . Surprisingly , hyperactive NOTCH1 cannot reprogram thymocytes on its own , despite the fact that NOTCH1 is activated by gain of function mutations in more than 55% of T-ALL cases . Rather , elevating NOTCH1 triggers a parallel pathway involving Hes1 and Myc that dramatically enhances the activity of SCL-LMO1 We conclude that the acquisition of self-renewal and the genesis of pre-LSCs from thymocytes with a finite lifespan represent a critical first event in T-ALL . Finally , LYL1 and LMO1 or LMO2 are co-expressed in most human T-ALL samples , except the cortical T subtype . We therefore anticipate that the self-renewal network described here may be relevant to a majority of human T-ALL .
An important attribute of stem cell populations is the capacity to self-renew indefinitely both in normal development and during the process of cell transformation . Cancer stem cells , initially identified in acute myeloblastic leukemias [1] , can self-renew indefinitely to propagate and maintain the disease [2] . This led to the experimental definition of leukemia initiating cell ( LIC ) characterized by their capacities to initiate the disease in transplanted host mice [1] , [3] . Important questions remain to be resolved with regards to the nature of the cell of origin of cancer , that is the normal cells from which cancer originates [4]–[6] and the mechanisms that drive the transition to an initiated state [7] . It was initially thought that the capacity for self-renewal of LICs , also referred to as leukemic stem cells ( LSCs ) , is conferred by the cell of origin of cancer , that is , primitive hematopoietic stem/progenitor cells ( HSPCs ) , even though the leukemic phenotype is manifest in differentiating myeloblasts [3] . Alternatively , oncogenes acting on committed progenitors can induce a stem cell gene signature [8] , leading to the reprogramming of non-self-renewing progenitors into pre-leukemic stem cells ( pre-LSCs ) [9] , [10] . Nonetheless , only subsets of progenitors are susceptible to oncogenic reprogramming , raising questions on the molecular events that determine the susceptibility of target cells to oncogenes . Normal thymic progenitors have limited if any self-renewal capacity [11] , [12] . Bone marrow-derived progenitors settle in the thymus and gradually acquire T cell characteristics while losing “stemness” [13] . The NOTCH1 pathway is a master regulator of thymopoiesis acting at several steps , in particular at the DN3 stage where NOTCH1 together with the pre-TCR drives irreversible T-lineage commitment [14] . NOTCH1 gain-of-function mutations were found in more than half of human T-ALL [15] and in most mouse models [16] , [17] . The significance of Notch1 for oncogenic transformation has been well established whereas the role of Notch1 in hematopoietic stem cell ( HSC ) self-renewal has been controversial ( reviewed in [18] ) . NOTCH activity is highly context-dependent [19] . Hence , a hyperactive Notch1 allele ( NICD; hereafter referred to as the Notch1 oncogene ) is shown to cause an exhaustion of HSCs at the expanse of T-LSCs [20] . Once transformed , LICs in Notch1-induced T-ALL depend on continued Notch1 signals for maintenance [15] , [21]–[23] and on several downstream effectors that include Hes1 [24]–[28] and Myc [27] , [29] . These LICs were found in the immature single positive ( ISP8 ) population , raising the question whether or not ISP8 are the cell of origin of T-ALL . Moreover , Notch1 is a weak tumor initiator [30] . Finally , the importance of Notch1 in pre-LSCs remains to be clarified . Self-renewal in normal HSCs is controlled by a network of transcription factors [31] . This network includes the basic helix-loop-helix ( bHLH ) transcription factors SCL/TAL-1 [32] , [33] and the highly homologous LYL1 [34] . Both SCL [35] and LYL1 form DNA binding heterodimers with E-proteins ( e . g . E2A and HEB ) that are also bHLH factors and directly interact with nuclear co-factors LIM-only ( LMO ) proteins to form transcription complexes that drive lineage-specific gene expression in hematopoietic cells [36] , [37] . SCL is partly redundant with LYL-1 in HSCs [34] . SCL , LYL1 and LMO1/2 expression decreases drastically at early stages of T-cell differentiation [13] . Their ectopic expression in the thymus , commonly driven by chromosomal rearrangements , is associated with T-ALL [38] . Overexpression of LMO1 or LMO2 in the thymus induces leukemia in mice with low penetrance and long latency [39] . This results from the emergence of pre-LSCs with altered gene expression [9] . Strikingly , T-ALL onset is accelerated by genetic collaboration with SCL [40] , [41] . How SCL induces T-ALL remains to be clarified . Indeed , two mechanisms have been proposed for SCL-mediated leukemogenesis . SCL heterodimerizes with and inhibits the activity of E-proteins [42]–[44] [45]–[47] , in particular of E2A and HEB that are nodal regulators in the T lineage ( reviewed in [48] , [49] ) . Accordingly , SCL inhibitory activity is sufficient to cause differentiation arrest in both B- [50] and T lineages [51] . Inhibition of E protein and differentiation blockade were , however , insufficient for leukemogenesis since most SCL transgenic lines did not develop T-ALL [40] , [51] , [52] , with the exception of one transgenic model [53] , [54] . In parallel , inhibitor of DNA-binding ID1 that sequesters E2A/HEB away from DNA was found to induce T-ALL in transgenic mice [55] . This led to the current view that bHLH oncogenic transcription factors that include SCL ( or TAL1 ) , TAL2 and LYL1 form inactive transcriptional complexes that induce T-ALL via inhibition of E proteins ( reviewed in [49] , [56] ) . With the predicament that cancer development is a Darwinian evolutionary process , the natural selection for genetic variants in which E proteins are inhibited should involve a variety of mechanisms , upregulation of bHLH transcription factors , of ID1-4 proteins that sequester E proteins away from DNA and/or inactivation of E protein encoding genes . The absence of the two latter categories so far in human T-ALL samples argues in favor of the second hypothesis , that transcription activation by oncogenic bHLH factors is an important leukemogenic driver . In support of this hypothesis , there is evidence for target gene activation in leukemic T cells [9] , [57]–[60] . Nonetheless , how the SCL-LMO1/2 collaboration establishes a pre-leukemic state to initiate T-ALL remains ill-defined . Recently , Lyl1 gene invalidation is shown to abrogate LMO2 self-renewal activity in pre-LSCs , suggesting that Lyl1 is an important downstream target of LMO2 [61] . However , overexpressing LYL1 on its own is clearly insufficient for thymocyte reprogramming [9] , indicating that the molecular context for cell transformation and/or thymocyte reprogramming by LYL1 remains to be uncovered . The inability of SCL or LYL1 to induce T-ALL on their own and the long latency required for LMO1/2-induced leukemogenesis strongly support the view that oncogene cooperativity drives synergistic modulation of gene expression , associated with major change in cellular reorganization [62] . Understanding the process of oncogene cooperativity in leukemia initiation can reveal mechanisms that control the growth of leukemic stem cells [63] . Recent genome-wide studies of leukemic samples at diagnosis have been highly informative on the mutational process and potential driver mutations in acute leukemias [64] , [65] . These powerful approaches did not allow for a clear distinction between initiating events in leukemogenesis and collaborating events that contribute to disease progression , which were revealed through two distinct approaches , the study of rare monochorionic twins [10] or of mouse models . Major questions remain nonetheless to be investigated . For example , it is not clear what determines the nature of the target cells of oncogenic reprogramming [5] . We used converging genome-wide approaches together with molecular and genetic approaches to provide novel evidence how the necessary collaboration between SCL , LMO1 and Notch1 determines the target cells of transformation in T-ALL and to identify novel mechanisms by which these oncogenes cooperate to activate stem cell genes and to convert normal thymocytes into self-renewing pre-LSCs . In particular , transcription activation posits a requirement for direct SCL-LMO1 interaction to assemble transcription activation complexes at target loci . In the present study , we generated transgenic mice expressing a mutant SCL that is unable to associate with LMO1/2 but retains its capacity to inhibit E2A/HEB , to provide genetic evidence for the importance of transcription activation in thymocyte reprogramming and in leukemogenesis .
The capacity for sustained self-renewal is best observed in serial transplantation assays . While normal thymocytes did not engraft in transplanted hosts , SCLtgLMO1tg thymocytes afforded thymic reconstitution which was sustained through three serial transplantations ( Fig . 1A–C ) . Thymocyte differentiation in the thymus progresses from the double negative stages ( DN1-4 ) to the CD4+CD8+ double positive ( DP ) stage and finally mature single positive CD4+ ( SP4 ) or CD8+ ( SP8 ) cells ( S1A , B Fig . ) . In primary and secondary transplantation , donor-derived cells retained a capacity to give rise to DP cells . However , after the tertiary transplantation , the proportion of donor-derived DN3 thymocytes increased markedly ( Fig . 1B ) , resulting in a cumulative 75-fold amplification ( Fig . 1C ) . In contrast , the other thymocyte subsets decreased during the same time-frame . We transplanted purified ETP , DN1-4 and DP populations from pre-leukemic SCLtgLMO1tg mice ( Fig . 1D ) . Only purified DN3 cells efficiently engrafted the thymus of recipient mice , ( left panel ) . A fraction of mice transplanted with DN1 and DN2 cells exhibited less than 1% engrafment and were “negative” by definition , although this was different from the absence of engrafment from DP cells . Furthermore , purified DN3 thymocytes retained the capacity to differentiate in vivo into DP and SP cells and , at the same time , to expand and self-renew ( right panel ) . Interestingly , donor-derived SP4 or SP8 thymocytes recovered from transplanted mice were activated by TCR stimulation to the same extent as normal host thymocytes by upregulating the CD69 activation marker ( Fig . 1E ) . This indicates that engrafted SCLtgLMO1tg thymocytes were non-leukemic . Accordingly , transplanted mice remained aleukemic , with small thymi and normal spleen size , despite the elevated expansion of DN3 thymocytes ( S1C Fig . ) . Together , our results indicate that the SCL and LMO1 oncogenes reprogram DN3 thymocytes into pre-LSCs that have acquired de novo self-renewing activity and retained their capacity to differentiate into functional T cells . The DN3 stage in the thymus marks T-lineage commitment driven by NOTCH1 acting in concert with the pre-TCR . We therefore addressed the question whether these two pathways contribute to DN3 cell reprogramming by SCL-LMO1 . We first addressed the functional importance of NOTCH1 in this process by lowering or increasing NOTCH activity . The expansion of pre-leukemic SCL-LMO1 DN3 cells was recapitulated in vitro by co-culture on OP9 stromal cells expressing the NOTCH ligand Delta-like-1 ( OP9-DL1 ) [66] ( Fig . 2A ) . DAPT , an inhibitor of the -secretase , abrogated this expansion ( Fig . 2A ) without affecting the viability of the OP9-DL1 stromal cells ( S2 Fig . ) . Strikingly , DAPT-treated DN3 cells were no longer able to engraft compared to control cells exposed to the vehicle alone when transplanted at equal numbers , suggesting that physiologic Notch1 signaling is required for SCL-LMO1 activity . We then addressed the consequences of supraphysiologic Notch1 signaling on thymocyte reprogramming . Oncogenic Notch1 has well established functions in leukemia induction and leukemia maintenance ( reviewed in [18] ) . Nonetheless , the role of Notch1 during this initial transition stage from a cell with finite life span to an aberrantly self-renewing pre-LSC remains to be addressed . Surprisingly , pre-leukemic Notch1tg thymocytes did not repopulate the thymus of recipient mice ( Fig . 2B ) . Rather , Notch1 significantly enhanced the engraftment of SCLtgLMO1tg thymocytes ( Fig . 2B ) . These cells also became independent of the thymic microenvironment ( S3A–B Fig . ) . Therefore , Notch1 acts as a strong enhancer of SCL-LMO1 self-renewal activity but lacks intrinsic reprogramming activity in the absence of other oncogeneic transcription factors . To determine whether the Notch1 oncogene modifies the frequency of SCL-LMO1 pre-LSCs and/or their expansion at the clonal level , we performed limiting dilution assays using DN3 pre-leukemic thymocytes ( Fig . 2C ) . A hyperactive Notch1 allele increased by 60-fold the frequency of SCL-LMO1-induced pre-LSCs ( Fig . 2C ) . In contrast , Notch1 did not significantly modify the expansion potential of individual pre-LSC when transplanted at ∼1 competitive repopulating unit , ( S3C Fig . ) . Therefore , Notch1 expands the pool of SCLtgLMO1tg pre-LSCs in vivo . We took advantage of the Tcrβ gene rearrangement as a clonal mark to assess the diversity of pre-LSCs in transplantation assays ( S3D Fig . ) . Pre-leukemic thymocytes were polyclonal before transplantation . Engrafted SCLtgLMO1tg thymocytes exhibited an oligoclonal signature whereas Notch1tgSCLtgLMO1tg thymocytes remained polyclonal after transplantation . Furthermore , we ruled out the possibility that SCLtgLMO1tg thymocytes had acquired Notch1 mutations ( S1 Table ) . These results indicated that a limited number of SCL-LMO1 expressing clones were able to self-renew in the absence of Notch1 while multiple clones were able to self-renew in the presence of Notch1 . We next addressed the role of the pre-TCR in the self-renewal activity induced by SCL-LMO1 and Notch1 . We exploited the Cd3ε-/- genetic mouse model in which thymocyte differentiation is blocked at the DN3a stage because of a non-functional pre-TCR/TCR ( S1 Fig . and S4A Fig . ) . We observed that pre-TCR signalling did not modify the frequency of SCL-LMO1-induced pre-LSCs nor the genetic collaboration between Notch1 and SCL-LMO1 in thymocyte reprogramming ( Fig . 2C ) . Moreover , the transplantation of pre-leukemic Cd3ε-/-SCLtgLMO1tg thymocytes resulted in thymic reconstitution in primary , secondary and tertiary recipient mice ( S4B Fig . , left panel ) , associated with DN3 cell expansion over serial transplantations ( S4B Fig . , right panel ) , as observed with SCLtgLMO1tg thymocytes ( Fig . 1C ) . We therefore took advantage of Cd3ε-/- mice to specifically assess the effects of the Notch1 transgene . We transplanted SCLtgLMO1tg thymocytes in competition with Notch1tgSCLtgLMO1tg thymocytes . The formers were marked with GFP to distinguish between the two cell types . Strikingly , the hyperactive Notch1 allele conferred a marked competitive advantage to SCLtgLMO1tg pre-leukemic thymocytes when transplanted at equal concentrations both at the limiting ( 1×103 ) and higher ( 1×106 ) cell doses ( Fig . 2D and S4C Fig . ) . SCLtgLMO1tgGfptg thymocytes became competitive only when transplanted at 20-fold excess . These results indicate that oncogenic Notch1 confers a competitive advantage to SCLtgLMO1tg pre-LSCs . In addition , the capacity of Notch1tgSCLtgLMO1tg thymocytes to engraft was no longer confined to DN3 but was found in all DN subsets ( DN1-DN4 ) and immature single-positive CD8 ( ISP8 ) cells but not in DP thymocytes ( S4D Fig . ) . Strikingly , these purified DN-ISP8 thymocytes preferentially gave rise to the same populations in transplantation , indicative of self-renewal activity ( Fig . 2E ) . Therefore , elevating Notch1 activity was sufficient to convert all immature thymocytes ( DN1 to ISP8 ) into cellular targets of SCL-LMO1 reprogramming activity . This expansion of cellular targets concur with the limiting dilution assay indicating that Notch1 increased the frequency of pre-LSCs . We conclude that NOTCH1 levels determine the expressivity of SCL-LMO1 in thymocyte reprogramming . Our findings indicate that SCL-LMO1 self-renewal activity is confined to the DN3 stage ( Fig . 1D ) , is GSI-responsive and is sensitive to NOTCH1 levels ( Fig . 2A–B ) . Interestingly , DN3 thymocytes are normally more sensitive to decreased Notch1 gene dosage compared to earlier thymocyte progenitors [67] . We therefore capitalized on the comprehensive gene expression data from the Immunological genome project ( Immgen ) together with NOTCH1 ChIP-Seq data [68] and HSC self-renewal resources to inform about candidate genes in pre-LSC self-renewal . First , we investigated the upregulation pattern of NOTCH1-bound genes that are GSI-responsive during early thymocyte differentiation . Considering genes that increased by more than 1 . 3-fold at each transitional stage , the analysis revealed that the percentage of up-regulated NOTCH1-bound genes steadily increased from the ETP to the DN3a stage and decreased thereafter ( Fig . 3A and S2 Table ) as expected ( reviewed in [69] ) . The general trend was also observed for the total transcriptome but the magnitude of the effect was stronger for the NOTCH1-bound genes ( Fig . 3A ) . Furthermore , NOTCH1-bound genes sharply decreased at the DP stage when the total transcriptome increased . Finally , DN3 cells in WT and SCL-LMO1 mice exhibit the highest levels of Notch1 and Notch3 genes and of the NOTCH reporter activity in Transgenic Notch Reporter ( TNRtg ) mice ( S5A–B Fig . ) as reported [23] . Therefore , NOTCH activity was highest in DN3 thymocytes , coinciding with the self-renewal activity of SCL-LMO1 . MYC has been implicated downstream of NOTCH1 in T-ALL [27] . Interestingly , we found that the increase in MYC target genes coincided with that of NOTCH1 and peaked at the DN2-DN3a transition ( Fig . 3A ) . Candidate genes operating with SCL-LMO1 at the DN3 stage should also be GSI-responsive , as engraftment by SCL-LMO1 DN3 thymocytes was DAPT-sensitive . These genes should operate prior to pre-TCR signalling , i . e . at the DN3a stage , since SCL-LMO1-induced self-renewal activity was fully efficient in Cd3ε-deficient DN3 thymocytes ( Fig . 2C and S4B Fig . ) . Based on the list of GSI-responsive NOTCH1-bound genes published by Wang et al [68] , 25 were found to increase at the DN2 to DN3a transition ( S2 Table ) . We next intersected this short list with HSC self-renewal resources [70] , [71] and found 3 genes Hes1 , Myc and Bcl6 ( Fig . 3B ) . We ruled out Bcl6 because of high expression in DP cells that are resistant to cellular reprogramming while both Myc and Hes1 decreased at this stage ( Fig . 3C ) . We noticed that Notch1 target genes correlate well with Notch1 mRNA levels during thymocyte differentiation , except Myc . Despite this , the increase in MYC-bound genes at the DN2-DN3a transition correlates with that of NOTCH1-bound genes ( Fig . 3A ) . These observations suggest MYC activity is subject to additional levels of regulation . Myc is a well known target of NOTCH1 in T-ALL [27] , [72] . Moreover , Hes1 overexpression expanded HSCs in culture [70] , [73] whereas Hes1 invalidation decreased LSCs in Notch1-induced T-ALL [24] . We therefore determined whether Hes1 or Myc may be important for this new activity of Notch1 at enhancing SCL-LMO1 reprogramming activity . To determine whether Hes1 or Myc can substitute for Notch1 as an enhancer of SCL-LMO1 , we overexpressed these genes in HSCs from SCL-LMO1 transgenic mice using the MSCV retroviral vector ( Fig . 3D ) . Both Hes1 and Myc caused an expansion of the DN3 population in transplanted mice , which was twenty to forty fold higher than that observed with the control vector ( GFP ) ( Fig . 3E ) . Furthermore , all DN populations were expanded by Myc whereas the activity of Hes1 was more specific to the DN3 population ( Fig . 3E and S6A Fig . ) . Thymocytes overexpressing Hes1 or Myc were recovered and transplanted into secondary mice at the limiting dose of ∼1 CRU per mouse . Consistent with this limiting dose , the proportion of engrafted mice remained at 30% in the Gfp and Hes1 groups ( Fig . 3F ) , suggesting that the frequency of pre-LSC was not modified by Hes1 . Nonetheless , the total number of DN3 thymocytes recovered from these mice were modestly higher with Hes1 . In contrast , Myc overexpression expanded the population of DN3 and 5 of 6 mice were reconstituted , indicative of increased pre-LSC frequency ( Fig . 3F ) . Therefore , Myc expanded DN3 thymocytes and increased their self-renewal activities , thus recapitulating the activity of the Notch1 transgene . In comparison , Hes1 activity was mostly in DN3 expansion . Accordingly , thymocytes from Notch1tgSCLtgLMO1tg mice in which Hes1 levels were decreased by a Hes1-directed shRNA ( S6B Fig . , upper panel ) exhibited two-fold decreased regenerative capacities compared to control cells expressing the empty vector ( S6B Fig . , lower panel ) . Moreover , the self-renewing DN and ISP8 populations were similarly decreased while DP cells that lacked self-renewal activity were unaffected ( S6B Fig . , lower panel ) . Therefore , Hes1 is required downstream of Notch1 as an expansion factor , whereas Myc controls both self-renewal activity and cell expansion . In summary , our results indicate that Notch1 signal controls both Hes1 and Myc and determines the capacity of DN3 thymocytes to be reprogrammed by SCL-LMO1 . To identify candidate genes that confer self-renewal capability to pre-leukemic DN3 thymocytes , we made use of the Cd3ε-deficient mouse model in which oncogene-induced self-renewal activity was unaltered ( Fig . 4A ) . We compared gene expression profiles of thymocytes from SCL-LMO1 transgenic and age-matched non transgenic Cd3ε-/- mice , taken three weeks after birth . At this time point , the transcriptome analysis identified only 53 up-regulated and 33 down-regulated genes in SCL-LMO1 expressing thymocytes ( S3 Table ) , indicating that the gene expression programs in the two cell types were comparable . We compared this list of differentially expressed genes with the genome binding profiles of SCL and LMO2 in several hematopoietic cell lines identified from a compendium of ChIP-seq datasets [74] . Within the down-regulated genes , only three had SCL peaks ( Cdc6 , Cdkn1a and Slc4a1 ) and none are presumed LMO2 target . In contrast , 9 of the up-regulated genes are presumed direct SCL and LMO2 targets ( S7A Fig . and S3 Table ) . These observations concur with the view that SCL together with LMO2 preferentially enhances transcription . We overlapped the SCL-LMO1 up-regulated gene set with a compendium of molecular signatures ( http://discovery . hsci . harvard . edu/ ) . We found a subset of genes that are frequent in stem cell and cancer signatures ( S7B Fig . and S4 Table ) , that includes Hhex , Nfe2 and Lyl1 . In particular , Lyl1 is associated with HSC and cancer cell signature ( S7B Fig . ) and controls HSC survival [34] ( S7B Fig . ; www . bonemarrowhsc . com ) . Next , we applied gene set enrichment analysis ( GSEA ) to uncover transcription factor signatures enriched in SCL-LMO1 thymocytes , using a compendium of 55 ChIP-seq datasets representing 31 hematopoietic transcription factors from the HemoChIP project and others ( see Materials and Methods ) . Surprisingly , the LYL1 signature was the most up-regulated in SCL-LMO1-expressing DN3 thymocytes ( Fig . 4B ) . Significantly , GSEA analysis also detected an up-regulated signature of SCL transcriptional partners , GATA2 , LMO2 , LDB1 , ETO2 and SCL , together with LYL1 and RUNX1-bound genes [75] . On the other hand , NOTCH1 signature was not significantly enriched in this gene set , concurring with the view that SCL-LMO1 and Notch1 operate in parallel pathways . Furthermore , all LYL1-bound genes are comprised within the SCL-LMO1-bound gene set ( S7A Fig . , right panel ) . Overall , our transcriptome analysis predicted a hierarchy downstream of SCL-LMO1 in which Lyl1 could coordinate a Notch1-independent self-renewal network ( Fig . 4C ) . By ChIP analysis , we found that SCL occupancy of the Lyl1 locus in SCL-expressing DN cells ( + SCL ) induced a 2- to 4-fold higher LMO1 binding to the Lyl1 promoter compared to control cells ( -SCL ) ( Fig . 4D ) . Finally , we observed by qRT-PCR that Lyl1 expression was significantly up-regulated by SCL-LMO1 ( Fig . 4E ) , concurring with our microarray results . In contrast , the Notch1 oncogene did not modify Lyl1 expression in DN3 thymocytes expressing or not SCL-LMO1 ( Fig . 4E ) . We conclude that SCL and LMO1 induce aberrant stem cell gene expression in DN3 thymocytes and reprogram these cells to acquire stem cell-like properties . SCL activates or represses gene expression , depending on its protein partners ( reviewed in [76] ) . Transcription activation critically depends on direct SCL-LMO1 or -LMO2 interaction to assemble a transcription complex on DNA [36] , [37] . This interaction is dispensable for transcription inhibition of E protein targets , which is directly attributed to SCL interaction with E2A or HEB . In particular , GSEA analysis indicated that E2A-presumed targets were not enriched within the list of differentially expressed genes ( S8A Fig . ) , suggesting that inhibition of E2A activity by SCL-LMO1 in DN3 thymocytes was not a major perturbation at the molecular level . We designed the SCLm13 that is specifically defective in LMO1/2 binding while heterodimerization with E2A/HEB was unaffected [37] ( S8B Fig . ) . Compared to wild type SCL , SCLm13 failed to activate the transcription of Lyl1 in transient assays whereas inhibition of E protein activity remained intact ( S8C Fig . ) . We previously identified Ptcra as a direct target of HEB/E2A that is inhibited by SCL [46] . We therefore stably introduced SCL and SCLm13 in the DN cell line AD10 and found that both genes inhibited the expression of Ptcra to the same extent , indicating that direct SCL-LMO1/2 interaction was dispensable for inhibition of E proteins ( S8D Fig . ) . E proteins are major cell fate deteminants in the thymus [77] , [78] , leading to the current view that T-ALL induction by SCL-LMO1/2 is due to E protein titration and inhibition [47] . To directly address the question whether the inhibition of E2A by SCL-LMO1 was sufficient for leukemogenesis , we generated transgenic mice expressing wild type SCL or the SCLm13 mutant at comparable levels ( Fig . 5A–B ) . We observed that SCLm13 fully retained its capacity to inhibit the expression of E protein target genes in DN3 thymocytes ( Fig . 5C ) . Significantly , while SCLm13 still inhibited E proteins ( S8D Fig . ) , there was a striking difference between the survival curves of SCLtgLMO1tg and SCLm13tgLMO1tg transgenic lines ( Fig . 5D ) . LMO1tg mice develop T-ALL with 20% penetrance and delayed onset at 400 days , as reported [39] . In contrast , the disease was fully penetrant in SCLtgLMO1tg mice , with an accelerated onset of 170 days [17] , [40] . In SCLm13tgLMO1tg mice however , leukemia onset was delayed to 380 days and the penetrance reduced to 65% ( Fig . 5D ) , underscoring the importance of SCL-LMO1 interaction in leukemogenesis . To further address the question whether the genetic collaboration between SCL and LMO1 in leukemogenesis was due to inhibition of E proteins , we generated E2a+/-LMO1tg mice . Loss of one E2a allele significantly decreased expression levels of E2A target genes in DN thymocytes ( S8E Fig . ) but did not mirror the collaboration of the SCL transgene with LMO1 to induce T-ALL . Together , our results indicate that inhibition of E2A is insufficient for leukemogenesis and that direct SCL-LMO1 interaction is an important determinant of leukemia onset and disease penetrance . We next addressed the question whether direct SCL-LMO1 interaction is required for self-renewal activity in DN3 thymocytes . The m13 mutation severely impaired the activation of self-renewal genes including Lyl1 ( Fig . 5E ) and drastically decreased the capacity of total thymocytes ( Fig . 5F ) or purified DN3 thymocytes ( S9A Fig . ) to reconstitute the thymus of transplanted hosts . Thymic engraftment of SCLm13tgLMO1tg thymocytes were reproducibly decreased to levels observed with LMO1tg only . Nonetheless , SCLm13 retained the same capacity as SCL to block the DN to DP transition compared to LMO1 alone ( Fig . 5G and S9B Fig . ) , a transition stage controlled by E2a and Heb gene dosage [79] , [80] . Together , our results indicate that inhibition of E protein and thymocyte differentiation blockade are distinct from the acquisition of self-renewal activity , which requires direct SCL-LMO1 interaction and transcription activation of a self-renewal program . Network analysis point to the importance of Lyl1 downstream of SCL-LMO2 ( Fig . 4C ) , consistent with published results [61] . Yet , ectopic expression of Lyl1 on its own did not recapitulate LMO2-induced aberrant self-renewal in thymocytes [9] . We reasoned that LYL1 activity most likely requires interaction with LMO1/2 for the following reasons: ( i ) the SCL interaction interface with LMO1/2 is conserved in LYL1 [36]; ( ii ) LYL1 is in complex with SCL and LMO2 [81]; ( iii ) LYL1 binding to DNA often overlaps with SCL and LMO2 binding [75]; and ( iv ) Lyl1 is redundant with Scl in controlling HSC self-renewal [34] . We therefore generated LYL1tgLMO1tg mice to address the question whether LYL1 enhances LMO1 self-renewal activity . LYL1 enhanced by 3-fold the activity of LMO1 on thymocyte engraftment ( compare Fig . 6A , left panel and Fig . 5F ) , whereas LYL1 alone did not reprogram thymocytes as expected . Similar to SCL-LMO1 , LYL1-LMO1 expanded DN3 cells only after transplantation ( Fig . 6A , right panel and S10 Fig . ) and this expansion was in the same order of magnitude compared to the inactive SCLm13-LMO1 ( Fig . 6B ) . The virtual convergence of SCL-LMO2 and LYL1-LMO2 target genes ( Fig . 4C ) may explain the capacity of LYL1-LMO1 to mimic SCL-LMO1 in DN3 thymocytes ( Fig . 6A–B ) . By RNA-Seq of 12 T-ALL patient samples , we found that LYL1 and HHEX mRNA levels are highly correlated with LMO2 levels ( r = 0 . 8 , Fig . 6C ) , concurring with the view that LYL1 and HHEX are downstream targets of LMO2 in T-ALL . Interestingly , LYL1 expression in the absence of TAL1 was found in 4 of 12 samples but TAL1 expression was never found in the absence of LYL1 ( Fig . 6C ) . These observations concur with the view that TAL1 is upstream of LYL1 ( Fig . 4C–E ) and with the essential role of Lyl1 in pre-thymic progenitors as well as in ETP-DN2 [82] . Moreover , the absence of correlation between TAL1 and LMO2 mRNA levels are consistent with the observations that LYL1 , but not TAL1 , is essential for LMO2-induced T-ALL [61] . We observed higher LYL1 , LMO2 , HHEX and MEF2C levels in ETP and pro-T ALL in adult ( Fig . 6C ) and pediatric ( S11 Fig . [38] ) T-ALL , consistent with this gene triad being direct targets of activation by MEF2C [83] . Nonetheless , LYL1 and LMO1/2 expression was detected in a majority of T-ALL samples independently of MEF2C or of phenotypic classification and included TLX1/3- and HOXA9-expressing leukemias ( Fig . 6C and S11 Fig . ) . These observations suggest that the molecular pathways controlling self-renewal described here is not limited to T-ALL samples harboring TAL1 or LMO1/2 translocations but may be relevant to other oncogenic subtypes of T-ALL .
Self-renewal is a mandatory trait of cancer stem cells as drivers of clonal expansion and evolution through layers of selective pressure [7] . This self-renewal activity is essential for long-term propagation . We now provide evidence that self-renewal is an initiating event triggered by the reactivation of stem cell genes in thymocytes ( Fig . 7A ) , as exemplified by chromosomal translocations driving ectopic SCL , LYL1 or LMO1/2 expression in thymocytes . Our data indicate that LYL1 coordinates a self-renewal network downstream of SCL-LMO1 to reprogram thymocytes with a finite life span into self-renewing pre-LSCs . Importantly , these self-renewal genes require the high levels of physiological NOTCH1 in DN3 thymocytes for expressivity . Their activities are therefore modulated by the thymic mircroenvironment . Furthermore , the Notch1 oncogene is devoid of intrinsic self-renewal activity but dramatically enhances SCL-LMO1 activity by conferring a proliferative advantage to SCL-LMO1-primed pre-LSCs and by recruiting all immature thymocytes into division to expand the pool of pre-LSCs ( Fig . 7B ) . Consequently , the hyperactive NOTCH1 allele acts as a strong enhancer of SCL-LMO1 by conferring additional fitness traits to SCL-LMO-initiated pre-LSCs , and allows for escape from envrironmental signals . LMO2 interaction with SCL has several consequences . First , interaction with SCL protects LMO1/2 from proteasomal degradation [37] . Second , SCL brings LMO2 to DNA , with two possible outputs: transcription activation or transcription inhibition . E proteins are major drivers of thymocyte development by activating gene expression programs that control cell survival , cell cycle and T-cell differentiation . In particular , SCL-LMO1 inhibit E protein activity and thymocyte differentiation [46] , [47] , leading to the current view that SCL-LMO1/2 induced T-ALL is due to E protein inhibition [47] . We bring several lines of evidence to indicate that the inhibition of E proteins is not the major cause of T-ALL . First , within the differentially expressed gene set in SCL-LMO1 DN3 thymocytes , we found a significant enrichment for binding by all SCL transcriptional partners , whereas E2A binding was not enriched . Second , removal of one E2a allele did not collaborate with LMO1 to induce T-ALL even though E2a was haploinsufficient for target gene expression . Third , we show that inhibition of E protein activity by the SCLm13 mutant did not enhance LMO1 self-renewal activity , resulting a dramatically decreased leukemogenic activity compared to wild type SCL , as assessed by decreased penetrance and increased latency . The modest enhancement of LMO1 in T-ALL induction by SCLm13 remains compatible with a tumor suppressor function for E proteins [47] . Therefore , the interaction of LMO1 with SCL , which is required to assemble a transcriptionally active complex on DNA [37] , is an important determinant of T-ALL development due to the reactivation of stem cell genes in DN3a thymocytes , during the pre-leukemic stage . By network analysis of the SCL-LMO1 transcriptome in DN3a thymocytes , we identified a hierarchy downstream of SCL-LMO1 which is controlled by Lyl1 . Previous work indicated that Lyl1 is critical for the oncogenic functions of LMO2 , consistent with a non-redundant function for Lyl1 in lymphoid progenitors and ETP [82] . This finding not mirrored by ectopic Lyl1 expression in the thymus [9] whereas Hhex deficiency [84] is mirrored by Hhex overexpression [9] . Considering that LYL1 and LMO2 chromosomal rearrangements were found simultaneously in a rare case of human T-ALL [85] , we now report that LYL1 collaborates with LMO1 to reprogram DN3 thymocytes . In summary , we provide genetic evidence that transcription activation by SCL and LMO1 is a major determinant of self-renewal in pre-LSCs and of the aggressiveness of T-ALL . NOTCH signaling is essential for T-cell commitment and specification . In particular , NOTCH1 cooperates with the pre-TCR to control cell survival and proliferation at the DN to DP transition [14] , at a critical checkpoint in the thymus . We previously showed that pre-TCR activity at the DN3 stage is required for the acquisition of Notch1 mutations in SCLtgLMO1tg thymocytes [17] . Once mutated , these hyperactive Notch1 alleles are sufficient to drive progression to T-ALL in concert with SCL-LMO1 . Therefore , the pre-TCR is a strong determinant of leukemia onset and of disease penetrance . Strikingly , we show here that the initiating event of reprogramming DN3 thymocytes into self-renewing pre-LSCs by SCL-LMO1 is independent of the pre-TCR but requires NOTCH1 signal . Taken together , our observations indicate that the pre-TCR is a collaborating event in disease progression but dispensable for the initial transition from DN3 cells to pre-LSCs . In contrast , we show that high levels of physiologic Notch signals in DN3 cells were required for SCL-LMO1 reprogramming activity . Functional studies of the NOTCH1 oncogene at time of overt leukemia in both human [21] , [22] and murine LSCs [23] , [86] , [87] showed that NOTCH1 controls leukemia initiating cell activity . In contrast , the role NOTCH1 in HSC self-renewal was controversial [30] , [88] , [89] . Using the mouse model as a unique opportunity to specifically understand initiating events in T-ALL , we unexpectedly found that a hyperactive NOTCH1 allele is devoid of intrinsic reprogramming activity in thymocytes , suggesting that weaker leukemia-associated Notch1 alleles [30] also lack this activity , similar to Notch3 [9] . Instead , high levels of NOTCH1 activity sensitize target cells to the reprogramming activity of SCL and LMO1 . Indeed , supraphysiologic NOTCH signaling was required past the DN3 stage , when physiologic NOTCH activity fell sharply . Therefore , our work provides a distinct conceptual framework to grasp the significance of the frequent co-occurrence of NOTCH1 gain of function mutations with major classes of oncogenic transcription factors in T-ALL . Multiple genetic interactions have been described for Notch1 [26] , [27] , [90] ( reviewed in [91] ) . Similar to Notch1 , Hes1 also drives T-cell development and inhibits alternate fates [92] . Interestingly , the conditional invalidation of Hes1 in adult hematopoietic cells led to T-cell defects and disrupted T-ALL maintenance [24] . Whether Hes1 contributes to oncogenic reprogramming of thymocytes at the initiation of the disease remained to be addressed . Here , we show that the hyperactive Notch1 allele upregulates Hes1 by 4-fold in DN3 thymocytes , which was insufficient for thymocyte self-renewal in vivo and required co-expression of the SCL-LMO1 oncogenes . Myc is required for the correct balance between self-renewal and differentiation of normal HSCs . Indeed , enforced Myc expression leads to HSC exhaustion whereas Myc deficiency results in increased HSC pool and self-renewal [93]-[95] . Our analysis of the Immgen data set indicates that MYC target genes but not Myc mRNA levels correlate with NOTCH1 activity during normal differentiation . This indicates additional levels of regulation; in particular MYC proteins are regulated by the ubiquitin ligase FBW7 in HSCs [96] , which is frequently mutated in T-ALL patients [97] . These observations point to the critical importance of regulating MYC levels in thymocytes . MYC is a well-documented target of NOTCH1 in leukemogenesis [27] . Furthermore , Myc promotes fibroblast reprogramming into induced pluripotent stem cells [98] . We now show that ectopic Myc expression in thymocytes recapitulates the activity of the Notch1 transgene to enhance thymocyte reprogramming by SCL-LMO1 . Our observations on the role of Notch1-Myc as an enhancer of SCL-LMO1 shed light on the pathway through which the BET-bromodomain inhibitors ( JQ1 ) that inhibited Myc could decrease the growth of primary leukemic cells , i . e . most likely due to interference with the NOTCH1 pathway [87] , [99] , [100] . Finally , our observations on the primordial role of Myc over Hes1 in substituting for NOTCH1 signals is consistent with the model of feed-forward-loop activated by NOTCH1 and MYC that promotes leukemic cell growth [29] . Therefore , our work clarifies the important role of Notch1-Hes1/Myc in the thymus as enhancers of self-renewal , but not as oncogenes with reprogramming activity . The Scl and Lmo2 genes [46] are silenced in DN3 thymocytes by a repressive histone mark [13] . We therefore surmised that chromosomal translocations or retroviral integration upstream of the LMO2 locus observed in pediatric T-ALL overcome these repressive marks to cause ectopic expression of oncogenes such as LYL1 , SCL and LMO2 which , in the context of DN3 thymocytes , collaborate with NOTCH1-HES1/MYC to confer aberrant self-renewal to these cells . We therefore propose a model in which SCL-LMO1-Lyl1 and Notch1-Hes1 are complementary in thymocyte reprogramming ( Fig . 7B ) . Phenotypic plasticity or lineage infidelity is often observed in cancer [101] . A recent report indicates that phenotypic plasticity predisposes reprogrammed fibroblasts to express stem cell characteristics and to induce tumors in nude mice [102] . In contrast , we show here that pre-leukemic stem cells conserve their DN3 phenotype through three rounds of transplantation and that the acquisition of self-renewal as an essential stem cell characteristic can occur in the absence of phenotypic plasticity . Therefore , our data indicate that phenotype plasticity is not an essential premise for oncogenic reprogramming whereas self-renewal is a mandatory trait [7] . The cell of origin of T-ALL was inferred from the phenotype of the leukemic cells [38] , or of LICs which was closer to the phenotype of a T-cell progenitor [22] , [103] . Nonetheless , LICs . have evolved through several selective constraints and acquired additional complexity and are defined as cells that produce an overt invasive leukemia . Here we define the cell of origin of T-ALL and the mechanisms by which oncogenes reprogram normal thymocytes . We bring evidence that the activation of a self-renewal program requires collaboration between several genes and incoming environmental signals , which is likely to determine the nature of the cell of origin of leukemia . Thus , SCL , LYL1 or Notch1 are not endowed with intrinsic reprogramming activity . Both SCL and LYL1 strongly enhanced LMO1 self-renewal activity in DN3 thymocytes due to higher endogenous NOTCH1 . Furthermore , the combination of three oncogenes , Notch1 , SCL and LMO1 had the strongest effect on self-renewal . Therefore , our data provide new mechanistic insights into the original two-hit model of cell transformation . Instead of each oncogene acting independently as a master switch in leukemia initiation , our work argues for the coincidence detection model in which biological outputs depend on the simultaneous occurrence of multiple signals within a network . Such cooperativity governs the process of self-renewal in pre-LSCs , which is an initiating event in T-ALL . High LYL1 and LMO2 expression in T-ALL was previously associated with immature or ETP-ALL [38] , [83] , [104] . While TAL1 expression in T-ALL was linked with a late cortical stage of T cell differentiation on the basis of cell surface markers [38] or whole transcriptome [104] , we provide cellular and genetic evidence that the initiating events occur in earlier stages in which NOTCH1 signals are highest , i . e . at the DN2b to DN3a transition , and that the Cd3ε gene is dispensable . These observations prompted us to examine the transcriptome of human adult T-ALL ( Leucegene-IRIC ) [105] and pediatric T-ALL [38] . This analysis also revealed that LYL1 and LMO2 were high in ETP and pro-T ALL but were detectable in almost all samples , suggesting that the molecular network defined in our study might operate in most T-ALL [106] . The importance of the SCL-LMO1 interaction described here for pre-LSC self-renewal activity , combined with the molecular view of this interaction interface suggests that targeting SCL-LMO interaction might represent a novel and promising therapeutic avenue . Such approach will be applicable to LYL1-LMO2 since the residues interacting with LMO2 are conserved between SCL and LYL1 .
All animals were maintained in pathogen-free conditions according to institutional animal care and guidelines set by the Canadian Council on Animal Care . Our protocol entitled “T-cell acute lymphoblastic leukemia induced by the SCL oncogene” was approved by the Ethics Committee of experimentation on animals of the University of Montreal , CDEA ( Comité de d ? ontologie de l'expérimentation sur les animaux ) . Transgenic mice were previously described: pSil-TSCL ( SCLtg ) [40] , Lck-LMO1 ( LMO1tg ) and Lck-NotchIC9 ( Notch1tg ) ( NIAID/Taconic Repository Bethesda ) , E2a+/- [107] , Lck-LYL1 ( LYL1tg ) ( International Mouse Strain Resource ) , Transgenic Notch Reporter ( TNRtg ) ( Tg ( Cp-EGFP ) 25Gaia , The Jackson Laboratory , Maine , United States ) and Cd3ε-/- [108] . Mice cohorts were generated by cross-breeding . Their genotypes were verified by PCR . The gene encoding the short isoform ( p22 ) of wild-type and m13 mutant [37] SCL protein was amplified by PCR using the following primers: 5′-GCGCGAATTCATGGAGATTACTGATGGT-3′ and 5′-TATACCCGGGTCACCGAGGGCCG-GCTCC-3′ . These fragments were digested with EcoRI and SmaI and subcloned in Cd2-VA minigene construct ( gift from Dr Dimitris Kioussis , National Institute for Medical Research , London , UK ) [109] , [110] . DNA was microinjected into the pronucleus of C57BL6 mice by IRIC Transgenesis Core Facility , University of Montreal . Transgenic mice were backcrossed into the C57BL6 background for more than 10 generations . Pre-leukemic thymocytes from donor mice ( CD45 . 2+ ) are transplanted intravenously into sub-lethally irradiated ( 600cGy ) recipient mice ( CD45 . 1+ ) . Thymic chimerism in the T-lineage ( Thy1 . 2+ ) was analysed by flow cytometry ( FACS ) and illustrated by the percentage of donor-derived cells ( % CD45 . 2+ ) found in the recipient thymus . Pre-leukemic thymocytes from SCLtgLMO1tg and Notch1tgSCLtgLMO1tg mice were transplanted into sub-lethally ( 600 cGy ) irradiated hosts ( CD45 . 1+ ) at various cell doses ( 107 , 106 , 105 , 104 , 103 , and 102 ) per recipient mouse ( n = 7 mice for each dose ) . Mice were scored positive when T-cell lineage reconstitution was more than 1% . Pre-leukemic stem cell ( pre-LSC ) frequency ( Range pre-LSC ± Confidence Interval ) and Competitive Re-populating Unit ( CRU ) frequency for the indicated genotypes were calculated by applying Poisson statistics using the Limiting Dilution Analysis software ( Stem Cell Technologies ) . The same strategy was used to compare the pre-LSC frequencies of DN3 SCLtgLMO1tg and Cd3ε-/-SCLtgLMO1tg thymocytes . The mean activity of pre-leukemic stem cells ( MAS ) is calculated according to the Harrison formula [111] , [112] . MAS represent the pre-LSC potential of approximately 1 CRU: MAS = [RU]/[CRU] where RU represents the re-populating activity of pre-LSC and CRU was determined by limiting dilution analysis as above . RU was calculated as previously described [33] . Since the number of competitor cells corresponds to the number of cells in the thymus of sub-lethally irradiated recipient mice , the formula was applied as follows: RU = [number of donor-derived cells]/[number of competitor host cells in recipient mouse thymus] . Pre-leukemic Cd3ε-/-Notch1tgSCLtgLMO1tg thymocytes ( CD45 . 2+ GFP- ) from one-week-old mice were mixed with Cd3ε-/-GfptgSCLtgLMO1tg competitor thymocytes ( CD45 . 2+ GFP+ ) in two ratios ( 1∶1 and 1∶20 ) at the indicated cell doses in Fig . 2D and S4C Fig . Mixed cells were then transplanted in irradiated hosts ( CD45 . 1+ ) . Thymic reconstitution by transplanted cells was assessed by FACS analysis 3 weeks post-transplantation . Single-cell suspensions were prepared from thymi of mice of the indicated ages and genotypes . Immunostaining was done as previously described [46] . All antibodies used for flow cytometry analysis were from Pharmingen ( BD Biosciences , Mississauga , Ontario , Canada ) : CD44 ( IM7 ) , CD25 ( PC61 . 5 ) , CD4 ( RM4-4 ) , CD8 ( 53-6 . 7 ) , Thy1 . 2 ( 30-H12 ) and CD24 ( 30-F1 ) . Dead cells were excluded by propidium iodide staining . FACS , cell cycle and cell division analysis were performed on a LSRII cytometer ( BD Biosciences ) using DIVA ( BD Biosciences ) and ModFit LT ( Verity Software House , Topsham , Maine , United States ) software . For nuclear SCL labeling , thymocytes were fixed and permeabilized with Fixation/Permeabilization Solution Kit and washed 3 times with Perm/Wash buffer ( BD Cytofix/Cytoperm , 554714; BD Biosciences , Mississauga , Ontario , Canada ) . The cells were then labeled with the monoclonal anti-human SCL BTL73 [113] at 1∶10 dilution , washed extensively with PBS , followed by a goat anti-mouse antibody coupled to FITC . The antibody was a generous gift from Danièle Mathieu-Mahul ( Institut de Génétique Moléculaire , Montpellier , France ) . Pre-leukemic cells were purified by FACS from transgenic mice and co-cultured on ( GFP-positive ) OP9 and OP9-DL1 stromal cell lines , as described previously [66] . Briefly , pre-leukemic cells were co-cultured on OP9 and OP9-DL1 cells in reconstituting a-MEM medium ( 12561 , Gibco , Life Technologies , Burlington , Ontario , Canada ) supplemented with 10% FBS ( 12318 , Gibco ) , HEPES 10 mM ( 15630-080 , Gibco ) , sodium pyruvate 1 mM ( 11360-070 , Gibco ) , b-mercaptoethanol 55 µM ( 21985-023 , Gibco ) , glutamax 2 mM ( 15750-060 , Gibco ) , penicillin/Streptomycin ( 15140-122 , Gibco ) , 5 ng/mL FLT-3 Ligand ( 308-FK , R&D system ) and 5 ng/mL IL-7 ( 407-ML , R&D system ) . Medium was half changed twice per week and the cells were counted and phenotyped by FACS after co-culture . T-cell stimulation was assessed using anti-CD3/CD28 beads as previously described [114] . Briefly , engrafted SCLtgLMO1tg pre-leukemic T cells ( donor thymocytes ) and host thymocytes were purified by FACS and co-cultured on a OP9-DL1 stromal cell line over 3 days with anti-CD3/CD28 beads ( Dynabeads Mouse T-Activator CD3/CD28 , 114 . 52D , Invitrogen , Life Technologies , Burlington , Ontario , Canada ) . The expression of the activation marker CD69 ( H1 . 2F3 , eBioscience , San Diego , California , United States ) was then analyzed by flow cytometry at the surface of SP4 and SP8 cells . Host B cells purified from the spleen were used as a negative control . DN3 thymocytes from WT , SCLtgLMO1tg , Notch1tg and Notch1tgSCLtgLMO1tg mice were purified by FACS and co-cultured on OP9-DL1 stromal cell line during 3 days . Derived-thymocytes were immunostained with T cell markers and then fixed and permeabilized ( CytofixCytoperm Plus , BD Bioscience ) during 30 minutes before the staining with the Ki67-FITC antibody . The DAPI was added at the end of the staining as a marker of DNA content . Cycle cycle analysis of DN3 thymocytes was finally analysed by FACS . RNAs collected from Cd3ε-/- and Cd3ε-/-SCLtgLMO1tg thymocytes were amplified and hybridized onto Affymetrix Mouse Genome 430A 2 . 0 arrays ( Ottawa Genome Centre , Ottawa , Ontario , Canada ) . Raw data pre-processing and differential expression analysis was carried out using Bioconductor packages in the R environment , according to the following pipeline: ( i ) probesets were summarized and normalized using the RMA procedure implemented in the Affy package [115]; ( ii ) absent/present probesets were detected using the MAS5 implementation of the Affy package , and probesets deemed absent in both conditions ( Cd3ε-/- and Cd3ε-/-SCLtgLMO1tg ) were removed from downstream analysis; and ( iii ) detection of differentially expressed genes was carried out using the Rank Products package [116] . We collected genome-wide chromatin occupancy data for 31 hematopoietic transcription factors ( 51 ChIP-seq experiments in total ) from Wang et al [68] and the HemoChIP project [74] . NOTCH1-binding peaks in G4A2 and T6E murine cell lines were computed using the Galaxy tool , according to the following steps: ( i ) sequence reads were mapped to the mouse genome mm9 using Bowtie with default parameters ( maximum 2 mismatches ) ; and ( ii ) peak coordinates were determined by the MACS tool , using the Pvalue cutoff <10–9 . Peak coordinates for the HemoChIP dataset mapped to the mouse genome mm9 were downloaded from http://hscl . cimr . cam . ac . uk/ChIP-Seq_Compendium/ChIP-Seq_Compendium2 . html . Finally , all peaks were associated to their closest transcription start sites in the mouse genome using PeakAnalyzer v . 1 . 4 tool [117] . Gene lists bound by transcription factors used in downstream analyses ( Figs . 4A , 5D , 6A ) included only those genes containing at least one binding site for the given regulator within the proximal promoter ( 2 kb region around the transcription start site ) . Total RNAs were prepared from 50 , 000 purified cell population cells from 1-week-old mice using RNeasy extraction kit ( Qiagen , Mississauga , Ontario , Canada ) . First strand cDNA syntheses were performed by reverse transcription as described [46] . Primer sequences are listed in S5 Table . Real-time quantitative PCR was done with SYBR Green Master Mix ( Applied Biosystems , Foster City , California , United States ) on Stratagene Mx3000 apparatus ( Stratagene , La Jolla , California , United States ) . ΔΔCt values were calculated using Ct values from β-actin gene as reference . The DN thymoma cell line AD10 . 1 [46] was cultured in IMDM ( Invitrogen , Burlington , Ontario , Canada ) containing 10% inactivated foetal calf serum ( FSC ) and 50 µM β-mercaptoethanol . The parental cell line was retrovirally transduced with MSCV empty vector or MSCV-SCL-expressing vector , and stable transfectants were kept under neomycin selection ( 1 mg/mL ) . Chromatin immunoprecipitation were performed as described previously [118] using the following antibodies: anti-SCL mouse monoclonal antibodies BTL73 ( generously provided by Dr . D . Mathieu-Mahul , Institut de Génétique Moléculaire , Montpellier , France ) , rabbit anti-LMO1 ( Bethyl Laboratories , A300-314A; Cedarlane Laboratories , Burlington , Ontario , Canada ) , and anti-rabbit IgG ( Sigma , St-Louis , Missouri , United States ) . Oligonucleotide sequences used for promoter amplification are shown in S5 Table . Gene transfer into bone marrow cells from 1-week-old pre-leukemic Notch1tgSCLtgLMO1tg mice was performed essentially as previously described [119] . Bone marrow cells were depleted of lineage positive cells through immunomagnetic bead cell separation ( LIN- ) and plated in suspension culture in IMDM with 15% FCS , 100 ng/mL murine Steel Factor ( SF ) , 10 ng/mL human IL-6 , 100 ng/ml human IL-11 and 5 ng/mL murine IL-3 , at a concentration of 1×106 cells/mL . All cytokines were produced as COS cell supernatants and were calibrated against recombinant standards . For the over-expression of Hes1 and c-Myc , LIN- cells were overlaid on irradiated ( 1500 cGy ) virus producing GP+E-86 cells contenaining the MSCV-GFP or –Hes1 or c-Myc in the presence of 0 . 8 µg/mL of polybrene ( Sigma Aldrich ) for 48h . For the down-regulation of Hes1 , LIN- cells were infected using lentiviral vectors containing either non-targeting shCTL or shHes1 ( Sigma , TRCN0000028854; St . Louis , Missouri , United States ) for 48 h . Following infection , cells were selected for 2 d with puromycin ( 1 . 5 µg/ml ) and transplanted into irradiated CD45 . 1 hosts . 11 T-ALL samples were collected by the Quebec Leukemia Cell Bank with informed consent . The project was approved by the Research Ethics Board of the Maisonneuve-Rosemont Hospital and Université de Montréal . These samples include the complete array of phenotypic T-ALL , ranging from ETP ( 1 sample ) to cortical T ( 3 samples ) , as previously published [105] . Transcriptome libraries were generated from 4 µg total RNA . Sequence data obtained by paired-end sequencing ( 2×100 bp , Illumina HiSeq2000 ) were mapped to the mouse reference genome and analyzed as reported . RNA-seq yielded 15 Gb of mapped reads per sample , with an average of 15 . 2 reads per kilobase per million ( RPKM ) . Data were log2 transformed and normalized between samples . RPKM values are taken as measures of the relative molar RNA concentration for each set of transcript . Correlation coefficients calculated for LMO2 are shown in Fig . 6C . Additional details for clonality analysis , co-immunoprecipitation , Luciferase assays and Notch1 sequencing are provided in S1 Protocol .
|
Deciphering the initiating events in lymphoid leukemia is important for the development of new therapeutic strategies . In this manuscript , we define oncogenic reprogramming as the process through which non-self-renewing progenitors are converted into pre-leukemic stem cells with sustained self-renewal capacities . We provide strong genetic evidence that this step is rate-limiting in leukemogenesis and requires the activation of a self-renewal program by oncogenic transcription factors , as exemplified by SCL and LMO1 . Furthermore , NOTCH1 is a pathway that drives cell fate in the thymus . We demonstrate that homeostatic NOTCH1 levels that are highest in specific thymocyte subsets determine their susceptibilities to oncogenic reprogramming by SCL and LMO1 . Our data provide novel insight into the acquisition of self-renewal as a critical first step in lymphoid cell transformation , requiring the synergistic interaction of oncogenic transcription factors with a cellular context controlled by high physiological NOTCH1 .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"hematologic",
"cancers",
"and",
"related",
"disorders",
"developmental",
"biology",
"leukemias",
"medicine",
"and",
"health",
"sciences",
"cancer",
"genetics",
"hematopoietic",
"stem",
"cells",
"lymphoblastic",
"leukemia",
"stem",
"cells",
"animal",
"cells",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"molecular",
"cell",
"biology",
"hematology",
"acute",
"lymphoblastic",
"leukemia",
"adult",
"stem",
"cells"
] |
2014
|
SCL, LMO1 and Notch1 Reprogram Thymocytes into Self-Renewing Cells
|
Double-stranded ( ds ) RNA fungal viruses are currently assigned to six different families . Those from the family Totiviridae are characterized by nonsegmented genomes and single-layer capsids , 300–450 Å in diameter . Helminthosporium victoriae virus 190S ( HvV190S ) , prototype of recently recognized genus Victorivirus , infects the filamentous fungus Helminthosporium victoriae ( telomorph: Cochliobolus victoriae ) , which is the causal agent of Victoria blight of oats . The HvV190S genome is 5179 bp long and encompasses two large , slightly overlapping open reading frames that encode the coat protein ( CP , 772 aa ) and the RNA-dependent RNA polymerase ( RdRp , 835 aa ) . To our present knowledge , victoriviruses uniquely express their RdRps via a coupled termination–reinitiation mechanism that differs from the well-characterized Saccharomyces cerevisiae virus L-A ( ScV-L-A , prototype of genus Totivirus ) , in which the RdRp is expressed as a CP/RdRp fusion protein due to ribosomal frameshifting . Here , we used transmission electron cryomicroscopy and three-dimensional image reconstruction to determine the structures of HvV190S virions and two types of virus-like particles ( capsids lacking dsRNA and capsids lacking both dsRNA and RdRp ) at estimated resolutions of 7 . 1 , 7 . 5 , and 7 . 6 Å , respectively . The HvV190S capsid is thin and smooth , and contains 120 copies of CP arranged in a “T = 2” icosahedral lattice characteristic of ScV-L-A and other dsRNA viruses . For aid in our interpretations , we developed and used an iterative segmentation procedure to define the boundaries of the two , chemically identical CP subunits in each asymmetric unit . Both subunits have a similar fold , but one that differs from ScV-L-A in many details except for a core α-helical region that is further predicted to be conserved among many other totiviruses . In particular , we predict the structures of other victoriviruses to be highly similar to HvV190S and the structures of most if not all totiviruses including , Leishmania RNA virus 1 , to be similar as well .
The encapsidated , double-stranded ( ds ) RNA viruses infect a wide range of hosts including bacteria , plants , fungi , insects , and humans and other vertebrates [1] , [2] . Viruses that infect and replicate in fungi ( mycoviruses ) have no known natural vectors and are spread vertically or horizontally by intracellular means [3] , [4] . Despite lacking an extracellular phase [4] , mycoviruses are successfully disseminated through all major groups of fungi [5] , [6] . As such , fungal viruses have been touted as potentially beneficial , biological control agents of pathogenic fungi that infect economically important agricultural crops . This has added significance because fungicides in current use pose health hazards and environmental risks [4] . Encapsidated dsRNA mycoviruses are currently classified in six families . Each member of family Totiviridae has a nonsegmented genome , but members of families Partitiviridae , Megabirnaviridae , Chrysoviridae , Quadriviridae , and Reoviridae have genomes comprising 2 , 2 , 4 , 4 , and 11–12 segments , respectively [4] , [7]–[9] . The capsids of all these viruses have an overall , spherical morphology and are constructed from an icosahedrally symmetric arrangement of one or more capsid proteins . Most of these capsids are single-shelled and range in diameter from ∼300 to ∼450 Å , with the exception of the larger , double-shelled reoviruses ( 600–850 Å ) [4] . All the single-shelled capsids ( except those of chrysoviruses ) , as well as the inner capsids of reoviruses , consist of 120 , chemically identical subunits arranged in a so-called “T = 2” , icosahedral lattice [10] . In addition , all dsRNA viruses , including those that infect fungi , must package one or more virally encoded RNA-dependent RNA polymerase ( RdRp ) molecules that replicate and transcribe the viral genome , since dsRNA cannot function as mRNA [1] . Totiviruses are the simplest encapsidated dsRNA mycoviruses in containing a single genome segment that encodes only two proteins: coat protein ( CP ) and either RdRp or a CP/RdRp fusion [9] . Members of family Totiviridae therefore provide a simple model system for studying mycoviruses , as well as dsRNA viruses more generally . Family Totiviridae comprises five current genera: Totivirus , Victorivirus , Giardiavirus , Leishmaniavirus , and Trichomonasvirus [9] , [11] . Members of the first two infect fungi , whereas the others infect protozoa [9] . Infectious myonecrosis virus ( IMNV ) , a tentative totivirus that infects shrimp [12] is currently unclassified , as are more recently discovered , tentative totiviruses that infect insects and fish [13]–[16] . Of the five genera in family Totiviridae , genus Victorivirus is currently largest , with 14 members and probable members [17] , [18] . Victoriviruses differ from the four other genera in their manner of RdRp translation ( see below ) and in the C-terminal sequences of their CPs . Victorivirus CPs range in mass from 77 to 83 kDa ( 746–780 aa ) and , to our current knowledge , share the unique feature of having a C-terminal region enriched in Ala , Gly , and Pro residues . The combined percentage of these residues in the C-terminal region of HvV190S is 52% , which differs considerably compared to the 24% , 22% , 18% and 17% present in similarly-sized C-terminal regions ( 130-aa stretches ) of representatives from the Totiviridae genera Leishmaniavirus ( LRV1-1 ) , Totivirus ( ScV-L-A ) , Trichomonasvirus ( TVV1 ) , and Giardiavirus ( GLV ) , respectively . Unlike some viruses in the other genera , no confirmed satellite dsRNAs , and consequently no satellite-encoded killer toxins , are known to be associated with victoriviruses . Host genes , however , are known to be upregulated during infections by one victorivirus , Helminthosporium victoriae virus 190S ( HvV190S ) [18] , [19] . Those genes include the victoriocin or vin gene that encodes a broad-spectrum , antifungal protein [20] and a gene that encodes the multifunctional Hv-p68 protein , which exhibits protein kinase , alcohol oxidase , and RNA-binding activities [21] , [22] . HvV190S was first discovered in 1978 in the filamentous fungus Helminthosporium victoriae ( H . victoriae , telomorph: Cochliobolus victoriae ) , the causal agent of Victoria blight of oats [23] . The 5179-bp HvV190S genome includes two large , slightly overlapping , open reading frames ( ORFs ) that encode the CP ( 772 aa; calculated mass , 81 kDa ) and the RdRp ( 835 aa; calculated mass , 92 kDa ) [17] , [24] . Notably , virus-infected H . victoriae exhibits symptoms typical of a disease phenotype [4] , which is unusual for dsRNA mycoviruses , in that most of them do not cause symptoms in their respective hosts . HvV190S may therefore provide an additional , useful model system for studying mycoviruses that have debilitating effects on their hosts [4] . HvV190S and likely other members of the genus Victorivirus [9] are unlike most other members of family Totiviridae in the mechanism by which their RdRps are translated . Members of most genera express their RdRps as a fusion protein with CP consequent to ribosomal frameshifting [25] . HvV190S , in contrast , expresses its RdRp as a separate , nonfused protein , using a coupled termination/reinitiation ( stop/restart ) strategy that involves an AUGA motif in which CP terminates at UGA and translation restarts at AUG to make RdRp . [26] , [27] . The RNA sequence requirements for the stop/restart mechanism include a 32-nt region that contains a predicted pseudoknot and lies close to the downstream AUGA motif [27] . Though the HvV190S genome contains just two ORFs , SDS-PAGE of purified HvV190S virions reveals the presence of three forms of CP , which have been shown by peptide analysis to be closely related and named p88 , p83 , and p78 to reflect their respective Mr values [9] . Expression in both bacterial and eukaryotic systems has shown p88 is the primary translation product , and p83 and p78 are derived from p88 via proteolytic processing at the C-terminus . Also , p88 and p83 are phosphorylated , but p78 is not [28] , [29] . Purified HvV190S virions include 190S-1 and 190S-2 forms , which differ slightly in sedimentation rate and capsid composition [9] , [28] . When separated by two cycles of sucrose density gradient centrifugation , samples of 190S-1 contain similar amounts of p88 and p83 whereas samples of 190S-2 contain similar amounts of p88 and p78 ( Figure S1A ) . The ratio of 190S-1 to 190S-2 particles varies in fungal cultures according to age , with 190S-2 predominating in virion preparations from 14-day or older cultures [29] . However , regardless of culture age , the yield of the separated 190S-1 and 190S-2 is always very low owing to significant losses during the separation protocol [29] and thus is not suitable for structure studies . HvV190S , like Saccharomyces cerevisiae virus L-A ( ScV-L-A , prototype of genus Totivirus ) and presumably all other members of family Totiviridae , has a “T = 2” arrangement of CP subunits [17] , [25] . Currently , the 3D structures of only a few dsRNA mycoviruses have been determined , and the only high-resolution crystal structure is that of ScV-L-A [30] . All other mycovirus structures have been examined by means of transmission electron cryomicroscopy ( cryoEM ) and 3D image reconstruction methods . These include structures of Ustilago maydis virus H1 , another current member of genus Totivirus [31]; three partitiviruses , Penicillium stoloniferum virus F [32] , Penicillium stoloniferum virus S [10] , and Fusarium poae virus 1 [33]; and two chrysoviruses , Penicillium chrysogenum virus [34] , [35] and Cryphonectria nitschkei chrysovirus 1 [36] . Here we used cryoEM and 3D image reconstruction to examine the structures of HvV190S virions and two types of HvV190S virus-like particles ( VLPs ) , called VLPC+ and VLPC . VLPC+ lack the dsRNA genome , whereas VLPC lack both the genome and the RdRp . The 3D structure of HvV190S VLPC was determined by cryoEM in an earlier study at ∼14-Å resolution [37] . We now report the 3D structures of HvV190S virions and both types of VLPs , all at subnanometer resolution ( ∼7–8 Å ) . From these data , we have been able to define the molecular envelopes of the two CP monomers , “A” and “B” , in each asymmetric unit of the “T = 2” capsid . These monomers are morphologically similar to each other as well as to the corresponding monomers of ScV-L-A . However , except for two α-helices in the subunit core and a large β-sheet on one side of the subunit , the folds of HvV190S and ScV-L-A CP are quite different . Comparisons of the predicted secondary structures of several different totiviruses indicate that this helix-rich core may be a highly conserved feature among all totiviruses and perhaps other dsRNA mycoviruses as well .
HvV190S particles were purified from a naturally infected culture of H . victoriae strain A-9 . VLPC+ and VLPC were purified from virus-free strain B-2ss , which was transformed with p190S [27] containing a full-length cDNA of HvV190S dsRNA ( for VLPC+ ) or with p190S containing only the CP ORF ( for VLPC ) . All H . victoriae cultures were grown for 14 days prior to harvest and purification . Western blot analysis , using a CP-specific antiserum , confirmed that all preparations of the three particle types ( virion , VLPc+ and VLPc ) contained the three related capsid proteins p88 , p83 , and p78 ( Figure 1 , left panel ) . The higher relative proportion of p78 compared to p83 reflects the fact that the 190S-2 form ( Figure S1A ) predominates in 14-day cultures ( [29] , also see Introduction ) . A second Western blot , using an RdRp-specific antiserum , confirmed that only virions and VLPC+ contained RdRp ( Figure 1 , right panel ) . Cryo-micrographs of the three particle types show that most of the HvV190S virions and VLPs are roughly the same size and shape , with relatively smooth , featureless profiles and no obvious protrusions ( Figure 2 ) . Most of the virions exhibit approximately uniform density throughout , as expected for particles that contain genome . A few particles in this sample appear to be empty or partially empty ( labeled in Figure 2A ) , and those likely are virions that have lost genome . The fraction of such particles in virion preparations increases with time , as does the fraction of aggregated particles ( not shown ) , and indicates that HvV190S virions are relatively unstable . All VLP samples contained particles with little density inside the confines of a thin capsid ( Figure 2 B , C ) . The absolute size of the HvV190S virions was determined by recording their cryoEM images in samples mixed with HK97 prohead II particles ( Figure 2D ) . The crystal structure of the latter is known [38] , and hence these make an excellent , internal standard for calibrating microscope magnification ( see Materials and Methods ) . Cryo-reconstructions of HvV190S virions , VLPc+ , and VLPC , at estimated resolutions of 7 . 1 , 7 . 5 , and 7 . 6 Å , were computed from 20 , 904 , 16 , 046 , and 6 , 294 particle images , respectively ( Table 1; Figures 3 and S2 ) . When the density maps are rendered at a threshold level set according to the expected mass of the capsid and are color coded to accentuate small radial differences , the outer surfaces of all three capsids appear essentially identical , with a complex distribution of features ( Figure 3 A , D ) . The inner surfaces are also virtually identical and complex ( Figure 3 B , E ) . Each capsid has a relatively smooth , spherical outer profile that is interrupted only by small protrusions at the twelve icosahedral fivefold ( I5 ) vertices where the maximum diameter reaches 462 Å . The outer surface of the capsid drops to a minimum diameter of 356 Å at the icosahedral threefold ( I3 ) and ∼368 Å at the icosahedral twofold ( I2 ) axes . The inside surface of the capsid approximates a sphere of ∼327 Å minimum diameter except for small , mushroom-shaped cavities that extend radially outward along the I5 axes ( Figure 3 B , E , white arrowheads; Figure 3 C , F , black arrowheads ) . The average thickness of the capsid is ∼35 Å but drops to a minimum of just 6 Å at the I3 axes ( Figure 3 C , F ) . Radial density plots of the cryo-reconstructions ( Figure S3 ) show that the capsid region is centered near radius 183 Å in each . The internal features of all three capsids also appear virtually identical , as illustrated in central , equatorial sections of the density maps ( Figure 3 C , F ) . However , close comparisons of these maps reveal the presence of subtle changes that are not easily detected by eye ( see Discussion and Movie S1 ) . The HvV190S virion reconstruction shows the genome packed in four or five concentric , roughly spherical shells ( Figure 3 B , C; Figure S3 ) , each with a maximum average density weaker than the capsid . Also , the level of organization within each shell diminishes with decreasing radius as indicated by progressively lower average density . The average spacing of these shells is ∼30 Å , which compares favorably with genome spacings observed in ScV-L-A and other dsRNA mycoviruses [10] , [12] , [30] , [32] . A gap of ∼17 Å separates most of the inner wall of the capsid from the outermost shell of dsRNA . This includes the regions near the I5 axes where the outermost genome shell is thickest and bulges outwards at the base of each cavity ( Figure 3 B , C ) . The inner surface of the capsid extends to its lowest radius ( 156 Å ) near the I2 axes , and densities spanning the sub-capsid space at those sites suggest contacts with genome ( Figure 3C , dashed red oval ) . Individual subunits in the HvV190S capsid are hard to discern by direct inspection of the density maps ( Figure 3 A , D ) . However , morphological features repeated 120 times on the capsid surface , consistent with a “T = 2” icosahedral arrangement as found in other dsRNA viruses [10] , [32] , [33] , [39] , can be identified upon careful examination . These “T = 2” structures consist of 60 asymmetric dimers , each containing two chemically identical monomers that occupy nonidentical environments , “A” and “B” , in the icosahedron . A-subunits cluster around the I5 axes , and B-subunits around the I3 axes . Close analysis of all three HvV190S cryo-reconstructions showed that at higher radii the capsid surface exhibits a staggered arrangement of two , similarly shaped features in nonidentical environments ( Figure 3 A , D ) . This exposed portion of each subunit contains a narrow “tip” on the proximal side ( near the I5 axis ) and a hook-like “anchor” on the distal side ( near the I2 axis ) . The tips of the A- and B-subunits lie respectively ∼9 and ∼47 Å away from the I5 axis , and the anchor of each A-subunit faces the anchor of a B-subunit from an adjacent capsid vertex region in a quasi-twofold arrangement . The HvV190S capsid , viewed in equatorial sections , is composed of numerous punctate and linear density features ( Figure 3 C , F ) , consistent with the presence of α-helices that lie perpendicular to or in the plane of the sections , respectively . Different , yet complementary views of the HvV190S capsid seen in spherical density projections ( Figure 4 ) show that these same types of features occur throughout the capsid at all radii . They also illustrate that a large portion of the A-subunit outer surface lies at higher radius than any portion of the B-subunit ( Figure 4C ) . The anchors of the A- and B-subunits lie at about the same radius ( Figure 4D ) , which indicates that the A-subunit is tilted relative to B . At lower radii , density features form a tightly interwoven , complex meshwork ( Figure 4 E–H ) , making it difficult to assign specific densities unambiguously to A or B . The highest packing density in the HvV190S capsid appears at radius 183 Å ( Figure 4F ) , consistent with the radial density plot mentioned above . To define more precisely how the A- and B-subunits are oriented relative to each other , we measured the radial positions of three specific structural features in the HvV190S virion reconstruction . First , the proximal tip of B sits ∼18 Å lower than that of A . Second , the distal tip of the B anchor , which lies close to the I3 axis , is ∼7 Å higher than that of the A anchor . These different radial heights of the proximal and distal tips of A and B demonstrate that the two subunits are tilted relative to one another , approximately along a line between neighboring I5 and I3 axes . The two subunits are also tilted relative to one another in the perpendicular direction ( i . e . , along a line between neighboring I3 and I2 axes ) , which positions the portion of the B anchor that lies close to the I2 axis at about the same radius as that of A . This relative arrangement of A- and B-subunits in virions is slightly altered in the VLPs . In both VLPs , both A and B are less tilted than their counterparts in virions , by 1° and 0 . 5° , respectively . Nevertheless , their proximal tips maintain a radial separation of ∼18 Å , and it follows that the A- and B-subunits remain tilted relative to one another in the VLPs . Hence , in all three HvV190S capsids , the A- and B-subunits adopt a staggered arrangement both in radius and in proximity to the I5 axis , though they do exhibit minor changes in their relative alignments . Given the difficulties inherent in assigning densities directly to the A- and B-subunits of the HvV190S capsid , we next tried to use the atomic structure of the A–B dimer of ScV-L-A [30] as a rigid body to model the HvV190S structure and delineate subunit boundaries . However , this process failed , primarily because the CPs of these two totiviruses differ markedly in size ( 680 vs . 772 aa for ScV-L-A and HvV190S , respectively ) and sequence ( see below ) . An initial assessment of the fit of the ScV-L-A dimer into the HvV190S virion density map gave no clear-cut indication of the monomer boundaries and also , as suggested previously [37] , showed no obvious evidence of similar folds of the HvV190S and ScV-L-A CPs . Without ScV-L-A as a guide , we turned to using an ab initio approach to segment out each subunit in the HvV190S virion reconstruction ( see Materials and Methods ) . This method relies on two principal assumptions: that the A- and B-subunits have compact , similarly folded structures and that a full , 120-subunit capsid model generated from the segmented A–B dimer would uniquely account for all non-genome density in the cryo-reconstruction . Results of the segmentation procedure showed that both subunits , as viewed from outside the capsid , have similar , asymmetric profiles that resemble a convex quadrilateral ( Figure 5 A , B ) . These quadrilaterals have dimensions of ∼74–97 Å on opposing long sides and ∼56–67 Å on opposing short sides . The longest dimensions of the A- and B-subunits are about 125 and 124 Å , respectively . Side views ( Figure 5C ) show them to vary in thickness from about 34 to 44 Å , which is consistent with the equatorial section views ( Figure 3 C , F ) and the fact that the long axes of the subunits adopt near-tangential orientations in the capsid . Overall , the 3D shapes of the A and B subunits are quite similar and superimpose well ( Figure 5C ) , indicating that each monomer has about the same tertiary structure . Two of the largest and most obvious deviations occur at the proximal ( Figure 5C , black arrowheads ) and distal ( Figure 5C , red arrowheads ) tips of the two subunits . Additional large differences are seen in corresponding regions on the interior side of the subunits ( encircled in the rightmost view in Figure 5C ) , where a pair of parallel features ascribed to a set of α-helices follows either a curved ( A-subunit ) or straighter ( B-subunit ) path . As seen in a surface view ( Figure 5A ) , extensive intersubunit contacts produce a nearly impenetrable shell . However , close inspection of all three HvV190S cryo-reconstructions reveals two small , solvent-accessible pores at each I2 axis and three , even smaller pores close to each I3 axis . The presence of these pores is confirmed by visual inspection of the grayscale density map , though their sizes and shapes cannot be determined precisely because they are influenced by the final resolution of the map ( the actual size of a pore or channel can only be measured accurately and reliably in maps at atomic or near-atomic resolution ) . Of note , the pores in both VLPs appear smaller than their counterparts in virions . At highest radii in the capsid , the proximal tips of five adjacent A-subunits form a unique set of interactions at the vertices ( I5 axes ) of the icosahedron . The only other A–A contact on the capsid surface occurs via the distal tips of two A subunits across each I2 axis , but this contact is much less extensive than those around each I5 axis . B–B contacts at the capsid surface are extensive but occur only among the B subunits that surround each I3 axis . The most extensive of all contacts at the capsid surface , however , occur between A- and B-subunits , such that each A or B is mostly surrounded by three distinct B- or A-subunits , respectively . The nonidentical environments that the A- and B-subunits occupy in the capsid are quite striking , as illustrated by two examples . First , the proximal tip of the A-subunit interacts with two , symmetry-related tips of neighboring A-subunits around the capsid vertex , whereas the proximal tip of the B-subunit interacts with distinct surfaces of two adjacent A-subunits . Second , the distal tip of each B-subunit interacts with distinct surfaces of two adjacent B-subunits around each I3 axis , whereas the distal tip of each A-subunit interacts with the distal tip of another A-subunit across each I2 axis and also with a distinct portion of an adjacent B-subunit . Having reliable estimates of the segmented volumes of the HvV190S A- and B-subunits , but insufficient resolution in any of the cryo-reconstructions to trace the peptide backbone of either CP , we revisited whether the ScV-L-A crystal structure could guide our analysis and interpretation of the HvV190S data . The CPs of both viruses do in fact have similar , quadrilateral-shaped morphologies , and rigid-body , quantitative fitting of the individual ScV-L-A A- and B-subunit structures into the corresponding , segmented density volumes of the HvV190S virion reconstruction revealed several regions of unassigned densities in HvV190S ( Figure 6 Ai , Bi ) . These regions include the proximal tip of the A-subunit and a portion of the proximal tip of B , the entire distal tips of both subunits , and a large portion of the long edge of both subunits ( right half of each subunit volume seen in Figure 6 ) . In addition , there is unassigned density on the shorter side of the B-subunit ( left half in Figure 6B ) . The presence of such unassigned densities is consistent with the HvV190S CP being 92 aa longer than ScV-L-A CP and also arises because the ScV-L-A models do not include the C-terminal 29 aa in each subunit since they were disordered in the crystal structure [30] . Close inspection of the data showed that most secondary-structure elements in the ScV-L-A subunit models failed to match features in the map for either HvV190S subunit . Nonetheless , three isolated segments in each fitted ScV-L-A model did correlate well with density features in the HvV190S maps . These included two α-helices ( helix 5 , aa 120–139 , and helix 13 , aa 358–383 ) and an antiparallel β-sheet comprising three main strands ( aa 26–38 , 41–54 , and 587–600 ) and three small strands ( aa 302–304 , 487–491 , and 604–606 ) . The α-helices ( colored yellow in subunit A and green in B; Figure 6 Ai , Bi , respectively ) constitute a portion of the central core of each ScV-L-A subunit and correspond closely in size and location to tubular density features in each HvV190S subunit ( more apparent in maps contoured at higher density thresholds; Movie S2 ) . For comparison , we annotated via an automatic procedure ( see Materials and Methods ) the location and length of possible α-helices in the segmented density ( Figure 6 Aiii , Biii ) , and found agreement only with the above two mentioned helices of ScV-L-A ( Figure 6 Aiv , Biv ) . An additional correlation was observed between the β-sheet in each ScV-L-A subunit and a large , plate-like density feature located near the distal end and shorter side of each HvV190S subunit ( Figure 6 Ai , Bi ) . Correspondence of these three elements of the unmodified ScV-L-A subunit model with density elements in the HvV190S virion map , though potentially just coincidental , led us to compare the predicted secondary structure of the HvV190S CP with the known secondary structure of ScV-L-A CP ( Figure 7 ) , with the goal of deriving a homology model for HvV190S . This procedure seemed warranted given that the ScV-L-A structure contains 24% helix , 21% sheet , and 55% random coil , compared with the prediction of 21% helix , 12% sheet , and 67% random coil for HvV190S . Moreover , the validity of the prediction procedure was at least partially substantiated since the predicted secondary structure for the first 651 aa of ScV-L-A CP ( 24 . 4% helix , 20 . 4% sheet , and 55 . 2% coil ) corresponded almost perfectly with the crystal structure . HvV190S CP has 19 predicted helices ( H1–19 ) , which correspond in number but not primary-sequence locations to 18 seen in the ScV-L-A crystal structure , plus one predicted in the missing C-terminal region of ScV-L-A ( Figure 7 ) . However , HvV190S CP has only 13 predicted β-strands ( S1–13 ) compared to 30 plus 1 that are found in ScV-L-A ( Figure 7 ) . We next tried to assign predicted secondary structures in HvV190S to the densities that matched between ScV-L-A and HvV190S in our model-fitting analysis described above . Regarding the large , antiparallel β-sheet in ScV-L-A that matches the plate-like density feature in the HvV190S map ( Figure 6 ) , two of the main β-strands in this sheet in ScV-L-A ( strands 2 and 3; Figure 7 ) appear to match predicted strands S1 and S2 in HvV190S ( aa 30–41 and 50–59 ) ; however , the other four β-strands in this sheet in ScV-L-A ( strands 10 , 18 , 25 , and 26 ) have no obvious correspondence in the predictions for HvV190S . Regarding helix H5 in ScV-L-A , which matches tubular density in the HvV190S map , predicted helices in HvV190S that might correspond are H3 ( aa 108–122 ) or possibly H2 ( aa 84–98 ) ( Figure 7 ) . Alternatively , the tandem stretch of helices 3 , 4 , and 5 in ScV-L-A could correspond with the region of HvV190S that contains both predicted helices H2 and H3 . Lastly , regarding helix 13 , which is the largest helix in ScV-L-A ( aa 358–383 ) , it appears to correspond well to two consecutive predicted helices in HvV190S , H11 ( aa 369–380 ) and H12 ( aa 385–396 ) . Moreover , helix 13 of ScV-L-A superimposes quite well with a long ( ∼30 Å ) tubular density in the HvV190S map . Given this assignment , helix 12 of ScV-L-A may correspond to predicted helix H10 in HvV190S ( aa 350–367 ) , but breaks in density in this and other regions of the HvV190S map made it impossible to determine unambiguously what the CP fold is either locally within the core or elsewhere in the subunit . The net result of our modeling analysis confirmed a previous suggestion [37] that the CP folds of ScV-L-A and HvV190S differ considerably . Also , we saw no evidence to suggest that flexible fitting of the ScV-L-A model to the HvV190S virion map would improve our interpretations of the data . Regardless , and despite the numerous and obvious differences between the CPs of ScV-L-A and HvV190S , our results show that their capsid structures are related and appear to have retained several elements that have survived their divergent evolutionary paths .
As described above , two helix-rich regions in HvV190S CP , encompassing two helices in a 39-aa stretch ( H2–3 , aa 84–122 ) and three helices in a 48-aa stretch ( H10–12 , aa 350–397 ) appear to correspond , respectively , to the helix 3–5 and helix 12–13 core regions in the crystal structure of each ScV-L-A CP subunit ( Figures 6 and 7 ) . This apparent detection of a similar , helical core in these two rather distantly related totiviruses ( prototypes of different genera ) [17] led us to explore whether other totiviruses may contain this conserved core , even though the overall tertiary structures of the CPs may be quite different . We therefore expanded our analysis of predicted secondary structures to include several other Totiviridae members . Comparisons of the predicted secondary structures of six additional victorivirus CPs revealed dramatic conservation of not just the putative H2–3 and H10–12 core components of HvV190S , but also the primary-sequence distribution of nearly every secondary-structure element in the proteins , including the β-strands ( Figure 7 ) . Furthermore , all seven victorivirus CPs that we examined have large C-terminal regions ( ranging from 130 to 136 aa long; Figure 7 ) that are rich in Ala , Gly , and Pro residues and predicted to assume a random-coil structure . This finding is consistent with previous results [17] , including that a 130-aa C-terminal region of HvV190S CP is dispensable for VLP assembly [40] . The overall close correspondence among the predicted secondary structures of the seven victorivirus CPs is consistent with the moderate level of sequence identity that each shares with HvV190S , ranging from 62% to 37% ( Table S1 ) . Considered together , these findings suggest that all victorivirus CPs are likely to adopt nearly identical tertiary structures and to assemble into very similar , smooth-surfaced , “T = 2” capsids . The T = 2 organization of capsid subunits is a highly conserved feature related to the unique life cycle of dsRNA viruses . These T = 2 capsids organize the replicative complex that is actively involved in genome transcription and replication . Hence , the capsids of all victoriviruses most likely have T = 2 arrangements of its subunits . Five additional totitviruses , including representatives of three genera whose members infect protozoa ( GLV , LRV1 , and TVV1 ) and two unclassified viruses ( EbRV1 and IMNV ) , also appear to include a centrally located stretch of residues with predominantly helical content , which might correspond to the shared core helices of ScV-L-A and HvV190S described above ( Figure 7; regions highlighted in yellow ) . In fact , both LRV1 and EbRV1 CPs contain several regions of predicted secondary-structure elements that lie upstream of the putative H12–13 core helices and form a pattern that closely mimics those observed in victoriviruses . These elements include ones corresponding to the H3–5 and S2–3 elements of ScV-L-A and the H2–3 and S1–2 elements of HvV190S . In addition , the C-termini of LRV1 and EbRV1 CPs are predicted to include a substantial amount of random-coil structure ( EbRV-1 more than LRV1 ) . In sum , these findings suggest that the CPs of LRV1 and EbRV1 fold similarly to those of ScV-L-A and HvV190S , and especially to the latter . Further support for the greater similarity to HvV190S includes that the helical-core region of HvV190S shares 25% and 30% sequence identity with the corresponding regions of LRV1 and EbRV1 , respectively ( Table S2 ) , whereas that region of ScV-L-A shares only 21% and 17% identity with those respective viruses . In addition , a recent phylogenetic analysis of the CP and RdRp sequences of 23 different totiviruses indicated that LRV1 and EbRV1 are more closely related to victoriviruses than they are to any other totiviruses [17] . Interestingly , EbRV1 is also the only other known totivirus outside the genus Victorivirus that has an AUGA motif comprising its apparent CP stop and RdRp start codons and therefore might also use a stop–restart strategy to translate its RdRp ( GenBank accession number AF356189 ) . Thus , we predict the overall structures of the LRV1 and EbRV1 capsids share significant similarities to the HvV190S capsid reported here . Our secondary-structure predictions for totiviruses GLV and TVV1 , and tentative totivirus IMNV , on the other hand , show fewer specific correspondences with HvV190S or ScV-L-A , or with one another . Although their CPs contain helix-rich regions ( highlighted in yellow , Figure 7 ) that might correspond to the helical-core region represented by H10–12 in HvV190S , this assignment is not as compelling for GLV and TVV1 . The presence of fiber complexes at the I5 vertices in IMNV revealed by cryoEM moreover clearly distinguishes IMNV from all other totivirus structures studied to date , though its capsid shell is organized in a manner similar to the HvV190S and ScV-L-A capsids [12] . We thus predict in addition from our analysis here that the capsids of GLV and TVV1 will be found to exhibit some distinctive features , as recently determined cryoEM structures do indeed suggest this to be true ( unpublished data ) . The capsids of the HvV190S virions and VLPs were shown to be composed of 120 protein subunits arranged in a “T = 2” capsid , with relatively smooth features , consistent with what was observed in previous studies of the empty capsid shell at ∼14-Å resolution [37] . Given that the encapsidated dsRNA mycoviruses never leave the host cell , the primary functions of the capsid as best we know are to protect the genome and replication intermediates and to participate in the replication and transcription of the dsRNA genome [1] , [31] . Another important function of the capsid is likely to sequester the dsRNA , which is a potent inducer of host defense mechanisms [41] . Most of these capsids have been shown to contain small pores , through which free nucleotides may enter and newly synthesized plus-strand RNA transcripts may exit during transcription , and through which the dsRNA does not normally exit or cellular proteins such as RNases do not normally enter [39] , [42] , [43] . Although the possible role of RNA in assembling these capsids remains unknown , it is currently thought that genomic RNA ( whether single or double stranded ) is not required for particle assembly and packaging of RdRp . Two lines of evidence support this conclusion: First , purified preparations of HvV190S always contain a small fraction of slowly sedimenting , empty capsid component ( the 113S component ) that packages RdRp but not RNA ( Figure S1 B , C; [23] , [44] ) . Second , when virus-free H . victoriae protoplasts are transformed with full-length cDNA of HvV190S dsRNA , the integrated cDNA is transcribed and the full-length transcript is translated into CP and RdRp [9] , [27] . The CP is assembled into VLPs ( VLPc+ ) that package RdRp but not genomic RNA . VLPc+ sediments at a similar rate to the naturally occurring 113S empty-capsid component when subjected to sucrose density-gradient centrifugation [9] , [27] . Close inspections of the equatorial sections of virions and VLPs ( Figure 3 C , F ) reveal subtle shifts in density positions in certain regions of the capsid , especially near the symmetry axes . To aid in seeing and interpreting these differences , we generated a morph movie in Chimera , highlighting the transition between HvV190S virion and VLPC+ . The results in particular exhibit expansions of the VLPC+ capsid near the I2 and I3 axes concurrent with a contraction at the I5 axes , resembling a “breathing” motion ( Movie S1 ) . Since the only known difference in components between the two particle types involves the presence or absence of dsRNA genome , this breathing is likely attributable to contacts between the inner capsid surface and the dsRNA in virions , which are not present in VLPC+ . Such movements are consistent with points of contact between capsid and genome suggested by densities spanning the sub-capsid space at the I2 axes of virions , which possibly pull the capsid inward at those sites ( Figure 3C ) . CP heterogeneity is a property of victoriviruses not shared by other totiviruses [9] , [28] . Preparations of the three HvV190S particle types ( virion , VLPc+ and VLPc ) examined in the current study shared similar amounts of CP forms p88 , p83 , and p78 , with p88 and p78 present in comparable amounts and p83 present in smaller amounts ( Figure 1; see also Figure S1B , which demonstrates that the intensity of the Western blot bands correlates well with the intensity of the corresponding Coomassie-stained bands from similar HvV190S and VLPc+ preparations ) . For simplicity here , we will consider only particles containing p88 and p78 ( 190S-2; Figure S1A ) , the predominant type in virion preparations examined ( though similar considerations would apply to particles containing p88 and p83 ) . How are the two forms of CP distributed in the capsids ? We envision at least four basic , most likely ways in which the uncleaved ( p88 ) and C-terminally cleaved ( p78 ) forms could be arranged . One possible arrangement would have p88 and p78 constituting 60 asymmetric heterodimers , with all p88 subunits in the A position and all p78 subunits in the B position , or vice versa , or with p88 and p78 randomly distributed between A and B positions . A second would have p88 and p78 constituting 30 asymmetric homodimers each , which are randomly distributed in the capsid . A third possible arrangement would have p88 and p78 constituting a random mix of hetero- and homodimers , which are furthermore randomly distributed in the capsid . Lastly , a fourth would have p88 and p78 segregated in different particles . Our current results do not distinguish among these possibilities , although a sorting procedure based on particle mass might have been able to prove the fourth , albeit unlikely possibility . In addition , if the CP C-terminal region ( the part missing from p83 and p78 ) is located in a regular , ordered position within the capsid ( see next paragraph for other possibilities ) , then we might have been able to distinguish subsets of the first possibility in which p88 and p78 are each respectively and consistently located in either the A or the B position . However , although we did not see structural evidence in favor of this consistent segregation of p88 and p78 between A or B positions , we cannot rule out this possibility because the ∼7-Å resolution obtained in this study is likely insufficient to detect these differences . We can also address where the heterogeneous region of HvV190S , i . e . , the C-terminal regions that are present in p88 but absent from p78 , are most likely to be located in the particle , in terms of the different capsid surfaces and radii . This region of p88 is predicted to have a random-coil conformation ( Figure 7 ) , consistent with its high Ala , Gly , and Pro content , and has also been shown to be dispensable for VLP assembly [40] . Susceptibility to cleavage in this region of p88 , to yield p83 and p78 , is moreover consistent with its predicted disorder . Whether cleavage occurs pre-or post-assembly is not presently known . To shed some light on this , purified HvV190S virions were treated with chymotrypsin under conditions that result in gradual/partial digestion of bovine serum albumin ( Figure S4C ) . The finding that protease treatment failed to generate additional cleavages in the C-terminal region of CP forms p88 and p83 ( Figure S4B ) suggests that the C-terminal region is most likely located internally to the CP shell , within the central cavity of the particles in which dsRNA and RdRp molecules are packaged in virions , and thus the cleavages in this region that produced CP forms p83 and p78 that are found in particles most likely occurred before ( or during ) particle assembly . Furthermore , SDS-PAGE analysis of the same purified HvV190S preparation on a 15% polyacrylamide gel did not reveal the presence of polypeptides of 10 kDa ( predicted size of the C-terminal tail; Figure S4A ) . These results further lend support to the contention that CP proteolytic processing occurs pre-assembly . Image analysis of scanned Western blots of purified HvV190S virions have indicated that the CP and RdRp subunits are present at a ratio of ∼65∶1 , supporting the notion that each virion contains 120 CP subunits and , on average , two RdRp molecules [4] , [24] , [39] . An original goal of this study was to locate the HvV190S RdRp within particles . Cryo-reconstruction methods have shown that there are 10–12 RdRp molecules attached to the inner surface of the “T = 2” capsid near the I5 axes in several , large dsRNA viruses , including mammalian orthoreovirus [45]; an aquareovirus [46]; a cypovirus [47] , [48]; simian rotavirus [49]; and bacteriophage φ6 [50] , [51] . Hence , HvV190S , with just 1–2 copies of RdRp per virion , represents an attractive model system to explore how an RdRp functions in the simplest of the encapsidated dsRNA viruses [9] . By imaging samples of HvV190S virions and two types of VLPs that lack genome , but one of which ( VLPC+ ) still contains the RdRp , we hoped to use difference-mapping procedures to locate the RdRp . This effort fell short as described above , however , in that the VLP structures are essentially identical and differ from the virion structure primarily with respect to only small ( ∼2–3 Å ) , rigid-body movements of the A and B capsid subunits ( Movie S1 ) . No additional density features were observed beyond those already ascribed to the 120 CP subunits or the genome . In the end , this lack of density for the RdRp was not surprising given that all three of the HvV190S cryo-reconstructions were computed with imposed icosahedral symmetry , and hence the signal from a component present only 1–2 copies per particle would be reduced at least 60-fold . Attempts to process the HvV190S image data with lower ( fivefold ) or no assumed symmetry have yet to yield conclusive clues about where the RdRp is located . This lack of success to date may simply reflect a need for many more particle images ( at least 12 times more if only fivefold symmetry averaging is employed or 60 times more if no symmetry averaging is used ) to achieve a signal-to-noise ratio close to that obtained in the reconstructions reported here . Considerations in the preceding two paragraphs include an assumption that the HvV190S RdRp is likely to be anchored to the inner surface of the capsid in a regular , stable manner , but this assumption might not be correct . If the RdRp forms only weak or nonspecific interactions with the capsid , then each RdRp could be oriented differently in each virion or VLPC+ , and any particle-based averaging procedure would yield a reduced RdRp signal . It is also possible , as suggested previously for HvV190S [37] , that because all victorivirus RdRps are expressed as separate , non-CP-fused proteins , they might anchor to the genome but not the capsid . Based on findings with other dsRNA viruses , however , we suspect that the victorivirus RdRp is likely anchored to the capsid close to the I5 axis near the base of the cavity where the end of a newly synthesized RNA transcript might accumulate and then exit if , for example , the proximal tips of the A-subunits rotate away from the axis to create a pore that is large enough to allow the RNA to escape from the particle and enter the host cytoplasm .
Purified HvV190S virions were isolated from infected H . victoriae strain A-9 ( ATCC 42018 ) as previously described [27] , [28] . VLPs were isolated from H . victoriae strain B-2 ( B-2ss , ATCC 42020 ) that was transformed with plasmid p190S ( for VLPC+ ) as described [27] , or with a mutant derivative of p190S ( p190S/CP ) lacking the RdRp ORF ( for VLPC ) . Construction of plasmid p190S/CP was based on the previously described transformation/expression vector p190S [27] . A pair of primers , a forward FseI-a-F: 5′ GTCTTTGGCCGGCCAGATGTCGGT 3′ and a reverse primer , CP2608R: 5′ ATATATATCGATTCATTGTCCCTCG 3′ , each containing a restriction site ( in italic font ) and a stretch of CP sequence ( in boldface ) . The primer pair was used to amplify a small fragment of the CP gene using p190S as a template . Primer FseI-a-F contains the restriction site of FseI that is unique and located at nt position 2171 in the CP sequence . A ClaI restriction site , which is not present in HvV190S , was added to the 3′ end of the CP sequence . The amplified fragment of CP is from nt 2171 to 2608 and was cleaved by FseI and ClaI and cloned into FseI/ClaI-digested vector p190S , which contains the remaining larger fragment ( nt 1–2170 ) of the CP sequence . All three particle types were purified from 14-day old stationary cultures grown in potato dextrose broth supplemented with 0 . 5% ( wt/vol ) yeast extracts as described [29] . Briefly , clarified fungal extracts were subjected to two cycles of differential centrifugation followed by rate zonal centrifugation in sucrose density gradients ( 100–400 mg/ml ) . The major band was then withdrawn with a syringe from the side of the tube and diluted with Buffer A ( 50 mM Tris-HCl buffer , pH 7 . 8 , containing 5 mM EDTA and 150 mM NaCl ) . The particles were then concentrated by overnight centrifugation at 40 , 000 rpm in a Beckman 50Ti rotor and the pellets were resuspended in Buffer A . SDS/PAGE analysis showed that all particle types used in the microscopy studies contained similar amounts of p88 , p83 , and p78 ( Figure 1 ) . Transmission electron microscopy ( TEM ) of negatively stained or unstained , vitrified HvV190S samples was performed as described [52] . Negative stain TEM was used to monitor the integrity and homogeneity of all samples . For this , 3 . 5-µl aliquots of each HvV190S particle type ( ∼1–10 mg/ml for V and ∼1–5 mg/ml for VLPC+ and VLPC ) were absorbed to continuous carbon grids that had been glow-discharged for ∼25 s in an Emitech K350 evaporation unit and subsequently stained with 1% aqueous uranyl acetate and rinsed with deionized distilled H2O . Micrographs were recorded on a 4 K2 Gatan CCD camera in an FEI G2 Tecnai ( Polara ) microscope . For cryo-EM , a 3 . 5-µl aliquot of each sample was absorbed to a continuous carbon grid that had been glow-discharged as stated above . An FEI Mark III Vitrobot was used to blot each grid for ∼4 s before plunging it into liquid ethane slush . Frozen grids were then transferred into a pre-cooled , multi-specimen holder , which maintained the specimen at liquid nitrogen temperature . Micrographs were recorded on Kodak SO-163 electron-image film at 200 keV in the Polara microscope under minimal-dose conditions ( ∼24 e/Å2 ) at a nominal magnification of 59 , 000 . The objective lens defocus settings , and the number of micrographs recorded for each particle type are listed in Table 1 . Micrographs were initially screened by eye to select for ones that exhibited minimal drift and without excessive astigmatism , and where the particle distribution and concentration were adequate . Acceptable micrographs were digitized at 6 . 35-µm intervals on a Nikon Supercoolscan 8000 microdensitometer . Programs AUTO3DEM [53] and RobEM ( http://cryoEM . ucsd . edu/programs . shtm ) were used to process the micrographs . This included boxing out individual particles ( 5012 pixels ) , followed by apodization , normalization , and linear gradient correction of these data . For the virion data set , only particles that appeared to be full ( i . e . containing the entire genome ) were boxed and analyzed further . Program ctffind3 was used to estimate the defocus value of each micrograph [54] . Next , the random-model computation procedure [55] was used to generate an initial 3D reconstruction at ∼30-Å resolution from 150 particle images for each specimen . Initial reconstructions obtained in this way served as starting models for the AUTO3DEM program ( version 4 . 01 . 07 , http://cryoem . ucsd . edu/programs . shtm ) , which were used to determine and refine the origin and orientation parameters for all particles in each data set . This process typically required 20 iterations until no further improvement in the resolution of the current reconstruction was achieved . The resolution limit attained for each cryo-reconstruction was estimated by the Fourier Shell Correlation method ( Figure S2 ) and applying a conservative 0 . 5 threshold criterion ( FSC0 . 5 ) [56] . During the course of this study , new functionality was added to AUTO3DEM that helped improve the quality of the final HvV190S reconstructions . First , a new option for re-centering particles was added to the autopp program used for batch processing . Second , two new search modes were added to the PO2R refinement program . Accurate estimates of particle orientation parameters rely on particles being properly centered in the box window , which is rarely if ever perfect for any particle picking method ( either manual or automatic ) . Hence , after the first round of 20 iterations of origin and orientation refinement finished and the estimate for the origin of each particle did not change appreciably , all particles were re-extracted from the original set of digitized micrographs to assure that the center of the particle lies within one half pixel of the center of the new box window . This reboxing procedure was automated by implementing an ad hoc option in autopp . On occasion , for some particles the template-based orientation refinement procedure can get trapped in a local minimum that does not correspond to the optimal solution . Hence , to help assure that the correct orientation is assigned to each particle , a global orientation search is recomputed for all particle images after repeated cycles of PO2R refinement fail to further improve the reconstruction . This global search , which is handled by the porg option in AUTO3DEM , covers the entire icosahedral asymmetric unit at an angular step interval defined by the user ( two degrees in this study ) , in each of the three Euler angles . It differs from the initial global search procedure normally performed with PPFT , primarily because it makes use of the most recent and presumably higher resolution reconstruction as a template , and PO2R is used for the search . The advantage is that the algorithm implemented in PO2R is more accurate than that in PPFT , though PO2R involves a much higher computational load . Given that the HvV190S image data were acquired over many microscopy sessions , the magnifications of the different micrographs varied enough to warrant an additional image processing step , implemented as the porm option in AUTO3DEM . Here , the relative magnification of each micrograph was estimated ( ±2% range scanned at 0 . 1% intervals ) by determining the average for all particles in each micrograph of the radial scale factor that maximized the correlation between each particle image with its corresponding projection of the reconstruction . Based on this analysis , the pixel size assigned to each micrograph was adjusted to assure that the 2D particle transforms were properly scaled and merged in program P3DR [57] to form the 3D Fourier transform and subsequent 3D density map . Each time any of the above procedures was performed ( reboxing , porg , or porm ) , an additional 20 iterations of local refinement were carried out with PO2R . This process led to an overall gain in resolution for the HvV190S virion reconstruction from 7 . 8 to 7 . 1 Å . To this point , all procedures were carried out with reconstructed density maps computed with image data that had been corrected to compensate for both amplitude and phase effects of the microscope contrast transfer function as described [58] . To aid in our analysis and interpretation of the HvV190S capsid structure , the final reconstructed density maps were only computed with phase flipping [59] . The pixel size for the virion map was calibrated as described below , but not for either of the VLP density maps . These were radially scaled against the virion map in RobEM as described [52] , [60] . Features at high spatial frequencies in the virion map were enhanced by application of an inverse temperature factor of 200 Å2 , which was chosen heuristically [61] . We used the program bampweigh in Bsoft to scale the radially-averaged Fourier amplitude spectra for the two VLP density maps to the spectrum computed from the sharpened virion map [62] . RobEM and Chimera [63] were used to visualize and analyze all three density maps . The thresholds used to render the density maps were set to give a capsid volume consistent with 120 copies of an 81-kDa protein . An HvV190S virion sample was mixed with a solution at similar concentration of a known standard , HK97 prohead II [38] to determine the absolute size of HvV190S . A 3 . 5-µL aliquot of this mixture was vitrified and cryo-EM was performed as described above for all other HvV190S samples . Six of the 54 micrographs that were recorded contained enough HvV190S and HK97 prohead II particles to enable us to compute a separate 3D reconstruction of each particle type in each micrograph . Radial density plots were computed for all twelve reconstructions [64] as well as from a density map generated from the crystal structure of the HK97 prohead II ( PDB ID 3E8K [38] ) . The known scale of the HK97 density plot derived from the crystal structure provided an absolute measure of the pixel size for each HK97 reconstruction and hence the magnification of each of the six micrographs ( Table S3 ) . These tests showed that the average magnification of the six micrographs was 59 , 318 , which is within 0 . 5% of the nominal magnification ( 59 , 000 ) of the Polara microscope . Knowledge of the calibrated magnification of each micrograph allowed us to define the absolute scale of the HvV190S reconstruction . This , for example , showed that the highest average radial density in the HvV190S capsid occurred at a radius of 183 . 5 Å ( Table S3 ) . The molecular boundaries of the A and B subunits that comprise each asymmetric unit of the HvV190S capsid were delineated by means of an iterative segmentation protocol specifically developed for this task . All the steps during this process , beginning with the final reconstructed map represented as a surface rendering , were performed using tools available in Chimera . The threshold used to render the HvV190S density map was chosen to account for the total expected molecular mass of the 120-subunit capsid ( 9 , 720 kDa ) . The Segger tool in Chimera [65] was used with its default values to initially segment the map into small “regions” , and only those regions identified as belonging to a single asymmetric unit ( e . g . as defined by the symmetry axes ) were considered for further analysis . Visual inspection of the density map clearly showed that there was a set of similar features on the outer surface of each subunit and this enabled us to manually group the Segger defined regions near these features into two separate subvolumes . Based on this initial estimate , these subvolmes were extended and refined in an iterative manner . At each iteration , this involved masking out densities in the reconstruction ( Segger ) according to the current definition of the subvolumes and then modeling these densities as a grid of markers ( meshmol ) , initially at a coarse ( two voxel ) spacing and later at a finer ( one voxel ) spacing . Next the closest , symmetry-related copies of each grid model ( 3 for A and 4 for B ) were generated ( sym ) and converted into separate density models ( molmap ) . Finally , these density models were visually analyzed , and newly segmented subvolumes were obtained by refining their boundaries . Refinement involved the imposition of two constraints on the models . First , the A and B subunits were assumed to adopt similar tertiary structures . Hence , at each iteration , we superimposed and compared the current solution for each subunit , to verify their agreement . Density features that were present in only one subunit were removed and temporarily left unassigned . As a second constraint , we required that a model of the capsid comprised of both subunits would account for all the capsid density in the reconstructed map . We verified this constraint by imposing icosahedral symmetry on the two subvolumes , and manually inspected all inter-subunit interfaces to identify inconsistencies in our assignments . For example , regions of overlap required re-segmentation , and unassigned density in the original map , which apparently belonged to the A or B subunit , was included in the appropriate subvolume . Following this approach , the boundaries of the A- and B-subunits were iteratively refined , until these constraints were fully satisfied . The crystal structure of the ScV-L-A virus CP ( PDB 1M1C ) [30] was manually docked as a rigid body into each of the A and B subunit maps that were segmented as described above . The Chimera fit_in_map tool was then used to refine independently the rigid body fit of the ScV-L-A model into each subunit . The normalized correlation value of the final fitting was 0 . 37 for both subunits , as provided by Chimera . Density features that could be assigned to α-helical segments were annotated in the A- and B- subunits of the virion as obtained by the segmentation procedure . The locations and lengths of the helices were predicted automatically using the VolTrac algorithm [66] implemented in the Sculptor graphics software [67] . Following the procedure outlined in [66] , a Gaussian-weighted local normalization was applied to the densities , using a standard deviation equal to half the nominal resolution . Then the prediction algorithm was executed with the target number of helices set equal to 20 . All program parameters were set to the default values , except the expansion threshold was set to 75% . Since the first 20 helices ranked by the prediction algorithm differed for the two subunits , we examined visually all the solutions found by the program ( 62 and 59 for A and B , respectively ) , and we retained only those that were common to both ( 33 helices ) . We used PSIPRED and default parameter values ( http://bioinf . cs . ucl . ac . uk/psipred/ ) to predict the secondary structures of the CPs of HvV190S ( p88 , GenBank accession number U41345 . 2 ) and eleven other viruses in the family Totiviridae . These include ( along with the genera to which they are assigned and their GenBank accession numbers ) Victorivirus BfTV1 ( Botryotinia fuckeliana totivirus 1 , AM491608 ) , CeRV-1 ( Chalara elegans RNA virus 1 , AY561500 ) , CmRV ( Coniothyrium minitans RNA virus , AF527633 ) , GaRV-L1 ( Gremmeniella abietina RNA virus L1 , AF337175 ) , SsRV1 ( Sphaeropsis sapinea RNA virus 1 , AF038665 ) , and SsRV2 ( Sphaeropsis sapinea RNA virus 2 , AF039080 ) ; unassigned EbRV1 ( Eimeria brunetti RNA virus 1 , AF356189 ) and IMNV ( penaeid shrimp infectious myonecrosis virus , AY570982 ) ; Giardiavirus GLV ( Giardia lamblia virus , L13218 ) ; Leishmaniavirus LRV1-4 ( Leishmania RNA virus 1–4 , U01899 ) ; and Trichomonasvirus TVV1 ( Trichomonas vaginalis virus 1 , GU08999 ) . The density maps of the cryo-reconstructions of the HvV190S virion , VLPC+ , and VLPC have been deposited in the Electron Microscopy Data Bank at the European Bioinformatics Institute with accession codes EMD-26556 , EMD-26557 , and EMD-26558 , respectively . GenBank accession numbers for the capsid proteins of thirteen related viruses in the family Totiviridae are provided in Tables S1 and S2 .
|
Of the known dsRNA fungal viruses , the best characterized is Saccharomyces cerevisiae virus L-A ( ScV-L-A ) , prototype of the genus Totivirus , family Totiviridae . Until the current study , there have been no subnanometer structures of dsRNA fungal viruses from the genus Victorivirus , which is the largest in family Totiviridae . The 3D cryo-reconstruction presented here of prototype victorivirus Helminthosporium victoriae virus 190S ( HvV190S ) approaches 7-Å resolution and shows the asymmetric unit of the capsid is a dimer comprising two , chemically identical coat-protein subunits organized in a so called “T = 2” lattice . These HvV190S subunits have a similar fold , but one that differs from ScV-L-A in many details except for a core α-helical region that is further predicted to be conserved among many other totiviruses . In particular , we predict the structures of other victoriviruses to be highly similar to HvV190S and the structures of most if not all totiviruses , including Leishmania RNA virus 1 , to be similar as well .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"nucleocapsid",
"viral",
"classification",
"plant",
"biology",
"macromolecular",
"assemblies",
"microbiology",
"viral",
"structure",
"chemical",
"biology",
"plant",
"science",
"rna",
"viruses",
"viruslike",
"particles",
"crops",
"protein",
"structure",
"viral",
"core",
"proteins",
"cryobiology",
"structural",
"proteins",
"chemistry",
"biology",
"biophysics",
"agriculture",
"macromolecular",
"complex",
"analysis",
"biochemistry",
"virology",
"subviral",
"particles",
"computational",
"biology",
"agricultural",
"biotechnology",
"macromolecular",
"structure",
"analysis"
] |
2013
|
Three-dimensional Structure of Victorivirus HvV190S Suggests Coat Proteins in Most Totiviruses Share a Conserved Core
|
Phylogenetic networks represent the evolution of organisms that have undergone reticulate events , such as recombination , hybrid speciation or lateral gene transfer . An important way to interpret a phylogenetic network is in terms of the trees it displays , which represent all the possible histories of the characters carried by the organisms in the network . Interestingly , however , different networks may display exactly the same set of trees , an observation that poses a problem for network reconstruction: from the perspective of many inference methods such networks are indistinguishable . This is true for all methods that evaluate a phylogenetic network solely on the basis of how well the displayed trees fit the available data , including all methods based on input data consisting of clades , triples , quartets , or trees with any number of taxa , and also sequence-based approaches such as popular formalisations of maximum parsimony and maximum likelihood for networks . This identifiability problem is partially solved by accounting for branch lengths , although this merely reduces the frequency of the problem . Here we propose that network inference methods should only attempt to reconstruct what they can uniquely identify . To this end , we introduce a novel definition of what constitutes a uniquely reconstructible network . For any given set of indistinguishable networks , we define a canonical network that , under mild assumptions , is unique and thus representative of the entire set . Given data that underwent reticulate evolution , only the canonical form of the underlying phylogenetic network can be uniquely reconstructed . While on the methodological side this will imply a drastic reduction of the solution space in network inference , for the study of reticulate evolution this is a fundamental limitation that will require an important change of perspective when interpreting phylogenetic networks .
As an example , consider again networks N1 and N2 in Fig . 1 , which display the same trees 𝓣 ( N ) = {T1 , T2 , T3} . ( In the following , 𝓣 ( N ) denotes the set of trees displayed by N . ) By displaying the same trees , these networks display the same clades , the same triples , the same quartets ( triples and quartets are rooted subtrees with 3 leaves and unrooted subtrees with 4 leaves , respectively ) and in general the same subtrees with an arbitrary number of leaves . Therefore , any method that reconstructs a network based on its consistency with collections of such data will not be able to distinguish between networks N1 and N2 . This includes all the methods whose data consists of clusters of taxa ( e . g . , [34] ) , triples ( e . g . , [35] ) , quartets ( e . g . , [36] ) , or any trees ( e . g . , [37] ) . The same holds for many , sequence-based , maximum parsimony and maximum likelihood approaches proposed in recent papers . For maximum parsimony , a practical approach [2 , 29–31] is to consider that the input is partitioned in a number of alignments A1 , A2 , … , Am , each from a different non-recombining genomic region ( possibly consisting of just one site each ) , and then take , for each of these alignments , the best parsimony score Ps ( T∣Ai ) among all those of the trees displayed by a network N . The parsimony score of N is then the sum of all the parsimony scores thus obtained . Formally , we have Ps ( N | A 1 , A 2 , … , A m ) = ∑ i = 1 m min T ∈ 𝓣 ( N ) Ps ( T | A i ) . It is clear that if two networks display the same set of trees ( as in Fig . 1 ) , then their parsimony score with respect to any input alignments will be the same—because they take the minimum value over the same set 𝓣 ( N ) —and thus they are indistinguishable to any method based on the maximum parsimony principle above . As for maximum likelihood ( ML ) , Nakhleh and collaborators [2 , 32 , 33 , 38] have proposed an elegant framework whereby a phylogenetic network N is not only described by a network topology , but also edge lengths and inheritance probabilities associated to the reticulations of N . As a result , any tree T displayed by N has edge lengths—allowing the calculation of its likelihood Pr ( A∣T ) with respect to any alignment A—and an associated probability of being observed Pr ( T∣N ) . The likelihood function with respect to a set of alignments A1 , A2 , … , Am , each from a different non-recombining genomic region , is then given by: Pr ( A 1 , A 2 , … , A m | N ) = ∏ i = 1 m Pr ( A i | N ) = ∏ i = 1 m ∑ T ∈ 𝓣 ( N ) Pr ( A i | T ) Pr ( T | N ) . Note that an important difference with the consistency-based and parsimony methods described above is that any tree T displayed by a network has now edge lengths and an associated probability Pr ( T∣N ) . Unfortunately , this ML framework is also subject to identifiability problems . For example , it does not allow us to distinguish between networks with topologies N1 and N2 in Fig . 1: for every assignment of edge lengths and inheritance probabilities to N1 , there exist corresponding assignments to N2 that make the resulting networks indistinguishable , that is , displaying the same trees , with the same edge lengths and the same probabilities of being observed ( see the last section in the Supporting Information , S1 Text ) . As a result , the likelihoods of these two networks will be identical , regardless of the data , and no method based on this definition of likelihood will be able to favour one of them over the other . We refer to S1 Text for a more detailed discussion about networks with inheritance probabilities and likelihood-based reconstruction . In general , we believe that these identifiability problems affect all network inference methods which seek consistency with unordered collections of sequence alignments or pre-inferred attributes such as clusters , triples , quartets or trees . In this paper , as in the ML framework above , we adopt networks and trees with edge lengths as the primary objects of our study . The primary motivation for this is that this choice makes our results directly relevant to the statistical approaches for network inference , all of which need edge lengths to measure the fit of a phylogeny with the available data . In addition to ML , these approaches include distance-based and Bayesian methods [39] , which are also promising for future work . However , there is another motivation for our choice: accounting for edge lengths solves some of the identifiability problems outlined above , as in some cases it allows to distinguish between networks with different topologies , which would be otherwise impossible to tell apart . For example , consider the three network topologies in Fig . 2 ( top ) , where taxon o is an outgroup used to identify the root of the phylogeny for a , b and c . These networks show three very different evolutionary histories: in N1 taxon b is the only one issued of a reticulation event—in other words the genome of b is recombinant—whereas in N2 and N3 , it is a and c , respectively , that are recombinant . However , N1 , N2 and N3 display the same tree topologies—those of T1 and T2—and thus would be indistinguishable to any approach that does not model edge lengths . If instead edge lengths are accounted for ( e . g . in a ML context ) and the data supports T1 and T2 with the edge lengths in Fig . 2 , then the only network fitting perfectly the data is N2 , with the edge lengths in Fig . 2 ( bottom right ) . It is easy to check that N2 now displays T1 and T2 with the shown edge lengths , whereas no edge length assignment to N1 or N3 can make these networks display T1 and T2 . We note that , throughout this paper , as in classical likelihood approaches , edge lengths measure evolutionary divergence , for example in terms of expected number of substitutions per site . No molecular clock is assumed , meaning that we do not expect edge lengths to be proportional to time . Unfortunately , accounting for edge lengths only solves some of the identifiability problems for phylogenetic networks . Consider networks N1 and N2 in Fig . 3: for any set of edge lengths for N1 , there exist an infinity of edge length assignments for N2 that make these two networks display exactly the same set of trees with the same edge lengths . In the following , we say that networks such as N1 and N2 are indistinguishable . In fact it is not difficult to construct other examples of indistinguishable networks: each time a network has a reticulation v giving birth to only one edge ( i . e . with outdegree 1 ) , then we can reduce by Δλ the length of this edge and correspondingly increase by Δλ the lengths of the edges ending in v , without altering the set of trees displayed by the network . Note that this operation , which we refer to as “unzipping” reticulation v , can result in v coinciding with a speciation node or a leaf when Δλ is taken to equal the length of the edge going out of v . For example in Fig . 3 , one may fully unzip the two reticulation nodes in N1 , thus obtaining the network N′ of Fig . 4 . As expected , N1 and N′ display the same set of trees ( {T1 , T2 , T3} ) and are thus indistinguishable . What is most interesting in this example is that , if we fully unzip the two reticulations in N2 ( the other network in Fig . 3 , also displaying {T1 , T2 , T3} ) , then we eventually end up obtaining N′ again . As we shall see in the following , this is not a coincidence: the unzipping transformations described above lead to what we call the canonical form of a network; under mild assumptions , two networks are indistinguishable if and only if they have the same canonical form ( e . g . N1 , N2 in Fig . 3 have the same canonical form N′; formal definitions and statements in the Results section ) . Here , we propose to deal with the identifiability issues for phylogenetic networks in the following way: since no data will ever enable any of the standard inference methods described above to prefer a network over all of its indistinguishable equivalents , we propose that these methods should only attempt to reconstruct what they can uniquely identify , that is , networks in canonical form . This is a radical shift , not only for the developers of phylogenetic inference methods , who will see a drastic reduction of the solution space of their algorithms , but also for evolutionary biologists , who should abandon their hopes of seeing a network such as N1 or N2 in Fig . 3 being reconstructed by these inference methods . Limiting the scope of network reconstruction to topologically-constrained classes of networks has been a recurring theme and an important goal in the literature on phylogenetic networks . Examples of such classes include galled trees [40 , 41] , galled networks [42] , level-k networks [43] , tree-child networks [44] , tree-sibling networks [45] , networks with visible reticulations [1] . Although the ultimate goal should be to establish what can be inferred from biological data , most of the proposed definitions are computationally-motivated: in general the rationale behind these classes is the possibility of devising an efficient algorithm to solve some formalization of the reconstruction problem . None of these definitions claims to have biological significance . Our goals are more basic: starting from the observation that not all phylogenetic networks are identifiable , since many of them are mutually indistinguishable with most inference approaches , we aim to define a class of networks that is ( existence goal ) large enough that every phylogenetic network has an equivalent ( i . e . indistinguishable ) network within this class and ( distinguishability goal ) small enough that no two networks within this class are indistinguishable . From our standpoint , the computationally-motivated definitions above are at the same time too broad and too restrictive . Too broad , because they determine a set of networks that includes many pairs of indistinguishable networks: for example the three indistinguishable networks in Fig . 2 are all galled trees—and thus belong to every single one of the classes mentioned above ( which are all generalizations of galled trees ) . Too restrictive , because these classes of networks do not include simple networks that it should be possible to reconstruct from real data . For example , Fig . 5a shows a network N with edge lengths that is not tree-sibling , nor has the visible property , and thus is not galled , nor tree-child ( for definitions , see [1] ) , but which in practice should be reconstructible: apart from the lengths of three edges ( x , y , z ) , N is uniquely determined by the trees that it displays ( a consequence of the formal results that we will show in the following ) , meaning that , given large amounts of data strongly supporting each of these ( seven ) trees with their correct edge lengths , any method for network inference properly accounting for edge lengths ( e . g . based on ML ) should be able to reconstruct N , or its canonical form N′ . To the best of our knowledge , only three classes of networks have claims of unique identifiability: reduced networks [46 , 47] , regular networks [48] and binary galled trees with no gall containing exactly 4 nodes [49] . These approaches bear some resemblances to ours , but do not include edge lengths in the definition of a network . Moreover , we argue that these classes of networks are still too narrow to be biologically relevant . We briefly describe and comment these previous works below . Moret et al . [46] defined notions of reconstructible , indistinguishable and reduced networks that resemble concepts that we will introduce here . Although some of their results were flawed [47 , 50] , some of the arguments in this introduction are inspired by their paper . Particularly relevant to the current paper is a reduction algorithm to transform a network into its reduced version . ( However , the exact definition of the reduced version is unclear: as one of the authors later pointed out [47] , “the reduction procedure of Moret et al . [46] is , in fact , inaccurate” and “in this paper we do not attempt to fix the procedure” . ) The concept of reduced version is analogous to that of canonical form here , as the authors claim that networks displaying the same tree topologies have the same reduced version ( up to isomorphism; Theorem 2 in [46] ) . This is somehow a weaker analogue of one of our results ( Corollary 1 ) ; weaker , because it does not claim that , conversely , networks with the same reduced version display the same tree topologies . To have an idea of the difference between our canonical form and the reduced version of Moret and colleagues , in Fig . 6 we compare the canonical form and the reduced version of the same network N1 . ( N1 and its reduced version are taken from Fig . 15 of [46] to avoid possible issues with the reduction algorithm . ) As one can see , the canonical form retains more of the complexity of the original network . Another reduction procedure on network topologies has been studied by Gambette and Huber [49] , who prove that if two network topologies reduce to the same topology , then they must display the same tree topologies . Again , this is analogous to , but somehow weaker than our results , since it only provides a sufficient condition for networks to be indistinguishable ( which in their context means to display the same tree topologies ) . This means that there can be irreducible networks that are indistinguishable ( e . g . those in Fig . 2 ) thus failing to achieve the distinguishability goal . Moreover , Gambette and Huber [49] show that a particular class of network topologies ( binary galled trees with no gall containing exactly 4 nodes ) are uniquely identified by the tree topologies they display . It is clear that this class is too small to achieve the existence goal . Finally , a regular network is a network topology N in which , among other requirements , no two distinct nodes have the same set of descendant leaves ( see [48] for a formal definition and characterizations ) . This requirement implies , among other things , that N cannot contain any reticulation v with outdegree 1 ( v and its direct descendant would have the same descendant leaves ) , which in turn implies that regular networks are special cases of our canonical networks ( the latter however also specify edge lengths ) . In fact regular networks satisfy a property that is analogous to the one we prove here for canonical networks: a regular network N is uniquely determined by the tree topologies that it displays [51] , meaning that there can be no other regular network N′ displaying exactly the same set of tree topologies . Willson [51] shows this constructively by providing an algorithm that , given the ( exponentially large ) set of tree topologies displayed by a regular network R , reconstructs R itself . However , unlike for our canonical forms , for a given network there may exist no regular network displaying the same set of trees ( e . g . consider the topology of N′′ in Fig . 5b ) , thus failing to meet the existence goal . Regularity is in fact a very restrictive constraint for a network . For example , none of the networks in Fig . 5 and Fig . 7 is regular , despite the fact that their topologies are uniquely determined by the trees with edge lengths that they display ( a consequence of our results further below ) . Finally , going back to Fig . 6 , collapsing the edge above taxa c and d in R ( N1 ) yields the regular network displaying the same tree topologies as N1 and N2 . Again , this shows that the canonical form retains more of the complexity of the original network than its regular counterpart .
All the phylogenies considered here—trees or networks—are rooted . This is because we assume that the analysis uses an outgroup ( possibly consisting of multiple taxa , and with no reticulations ) for rooting . For simplicity , outgroup lineages are not included in our phylogenies ( an exception to this is in Fig . 2 ) . Note however that , because our phylogenies have edge lengths , and because omitting the outgroup is just a convention , the omitted lineages must have the same lengths for a network and all the trees it displays . For example , if we wish to omit the outgroup from N2 in Fig . 2 and from the trees that it displays ( T1 and T2 in Fig . 2 ) , then what we obtain are N 2 ′ , T 1 ′ and T 2 ′ in Fig . 7 . This has a notable consequence: the trees displayed by a rooted network with edge lengths may have a root with outdegree 1 ( e . g . T 1 ′ in Fig . 7 ) . For flexibility , we also allow a network to have a root with outdegree 1 . Moreover , we allow multiple lengths for an edge in a network , but not in a tree . For example , in Fig . 6 , network N 2 ′ has an edge with two lengths ( λ7 + λ12 + λ14 and λ7 + λ11 + λ13 + λ14 ) . The motivation behind multiple lengths lies in the observation that , whereas each edge in a phylogenetic tree describing the evolution of non-reticulating organisms trivially corresponds to a unique evolutionary path in the underlying real evolutionary history , when reticulate events have occurred this is not necessarily true: Fig . 8 and Fig . 9 show that some evolutionary scenarios can either be represented using multiedges ( multiple edges with the same endpoints ) or edges with multiple lengths . Although these two options are mathematically equivalent , graphically the second one leads to more compact representations , and this is why we choose to allow multiple lengths rather than multiedges . For our purposes we only need to consider the case where e has a finite set of lengths ( Λ ( e ) = {λ1 ( e ) , … , λk ( e ) } ) . Another unconventional aspect of our networks is the possibility of having nodes with in-degree and out-degree both greater than one . ( See , e . g . , the last common ancestor of c and d in N 2 ′ in Fig . 6 . ) Traditionally , the internal nodes in a phylogenetic network are constrained to belong to one of two different categories: reticulate nodes , with more than one incoming edge and just one outgoing edge , and speciation ( or coalescence ) nodes , with one incoming edge and multiple outgoing edges . Because reticulate and speciation events are clearly distinct , it is reasonable to constrain internal nodes to only fall in the two categories above . In our framework , this requirement is dropped , and some networks , notably those in canonical form , may have nodes that both represent reticulate and speciation events . In this case , it is important to understand that these nodes represent a potentially complex ( and unrecoverable ) reticulate scenario , followed by one or more speciation events . Compare , for example , network N and its canonical form N′ in Fig . 5 , or N2 and N 2 ′ in Fig . 6 . ( In the latter , it is especially instructive to consider the reticulate history above the direct ancestor of taxon e . ) We use network N1 of Fig . 3 to illustrate the NELP property . In N1 there are three distinct weighted paths having as endpoints the root of N1 and the direct ancestor of b . The lengths of these paths are ℓ1 = λ1 + λ6 , ℓ2 = λ2 + λ3 + λ5 + λ8 and ℓ3 = λ2 + λ10 + λ9 + λ8 . Moreover , there is another pair of paths having the same endpoints: those of lengths ℓ4 = λ3 + λ5 and ℓ5 = λ10 + λ9 . Thus N1 has the NELP property if and only if the three numbers ℓ1 , ℓ2 and ℓ3 are all different ( note that this implies that also ℓ4 and ℓ5 are different ) . If edge lengths are taken to represent evolutionary change , rather than time , this is a very mild requirement: when edge lengths are drawn at random from a continuous distribution , the probability that two paths get exactly the same length is zero . On the other hand , the NELP property does not hold for phylogenetic networks where edge lengths are taken to represent time . For these networks , canonical forms may not be unique ( see Fig . 10 for an example of this ) . Even in this case , we believe that inference methods should only consider phylogenetic networks in their canonical form , as this allows to reduce the solution space without any loss in “expressive power”: since every network N has ( at least one ) canonical form that displays exactly the same set of trees—and therefore has the same fit with the data as N—restricting the solution space to canonical forms always leaves at least one optimal network within this space . The real weakness of using canonical forms in a molecular clock context is that if a canonical form is not unique , then it cannot be considered representative of all the networks indistinguishable from it . As an example of this , consider the indistinguishable networks in Fig . 10: none of these is representative of all the others .
Our results are both negative and positive . The bad news is that any method that scores the fit between a network N and the available data—which may be sequences , distances , splits , trees ( with or without edge lengths ) —based on the set of trees displayed by N must face an important theoretical limitation: regardless of the amount of available data from the taxa under consideration , some parts of the network representing their evolutionary history may be impossible to recover—most notably the relative order of consecutive reticulate events ( see , e . g . , Fig . 3 ) . The good news is that , when edge lengths are taken into account , we can set precise limits to what is recoverable: the canonical form of a network N is a simplified version of N that excludes all the unrecoverable aspects of N . In a canonical form , reticulate events are brought as forward in time as possible , causing the collapse of multiple consecutive nodes . ( Compare again network N2 and its canonical form N 2 ′ in Fig . 6 . ) The importance of the canonical form N′ of a network N lies in the fact that , if we restrict our consideration to networks with the NELP property , N′ is the unique canonical network consistent with perfect and unlimited data from the taxa in N . There is an interesting analogy between soft polytomies in classical phylogenetics and collapsed nodes in a canonical network . Both represent lack of knowledge about the order of evolutionary events: speciations or more generally lineage splits in the first case , and reticulate events in the second . However , there is also an important difference between them: whereas in principle polytomies can be resolved by collecting further data from the taxa in the tree ( for example , by extensive sequencing of their genomes [52] ) , the standard network inference methods considered here cannot resolve collapsed nodes in a canonical network , irrespective of the amount of data from the taxa under consideration . This difference is mitigated by the observation that increased taxon sampling may indeed permit to resolve the collapsed nodes , when the new lineages break adjacencies between reticulate nodes . However , such lineages may not always exist or they may be difficult to sample . The present work has several consequences that should be of interest both to the biologists concerned by the use of methods for phylogenetic network inference , and to the researchers interested in the development of these methods . We illustrate these consequences starting from a well-known problem of network inference methods , that of multiple optima . It has been noted before that many of the inference methods that have been recently proposed—especially those solely based on topological features—often return multiple optimal networks: Huson and Scornavacca show a striking example of this ( Fig . 2 in [53] ) , where the problem of finding the simplest network displaying two given tree topologies admits at least 486 optimal solutions . The existence of multiple optimal networks for a given data set is essentially due to two reasons: insufficient data and non-identifiability . For the example of 486 optimal solutions , this large number may be partly due to the fact that the goal was to achieve consistency with only two tree topologies . More data may enable to discriminate among the 486 returned networks . Non-identifiability , which occurs when none of the allowed data can discriminate between two or more networks , is a more serious problem than insufficient data , as it cannot be solved by simply increasing the size of the input sample . Another interesting example appears in a paper by Albrecht et al . [54] , which we reproduce here in Fig . 11 . Here , there are only three optimal networks , essentially differing for which of the three clades {A . bicornis , A . longissima , A . sharonensis} , {A . uniaristata , A . comosa} and {A . tauschii} is considered as a hybrid ( in this example reticulations represent hybridizations ) . This pattern is entirely analogous to that of the three networks in Fig . 2 ( with a , b and c replaced by the three clades above ) , meaning that these three networks are indistinguishable to methods not accounting for edge lengths . Therefore , in this example , the existence of multiple optimal solutions is entirely due to non-identifiability . All this motivates three recommendations: Correspondingly , we recommend that edge lengths should be accounted for in the analyses ( point 1 ) and , for each of the indistinguishable classes resulting from this choice , we identify a canonical network that , for all practical purposes , can be considered to be unique . Most important to the end users , we propose that a canonical network N ̂ is what should be given as the result of the inference , with the caveat that N ̂ is a way to represent a class of networks that are all equally supported ( point 2 ) . In a canonical form N ̂ , the aspects that are not common to all networks in this class are collapsed , as described above . This will help the evolutionary biologist to locate the uncertainties in the phylogeny , and possibly to choose further taxa to resolve them . Finally , we propose that inference methods only attempt to search among—or construct—phylogenetic networks in their canonical form ( point 3 ) . We note that accounting for yet more characteristics of the data may reduce ( or eliminate altogether ) the identifiability issues for phylogenetic networks . In the case of sequence-based methods , one may take into account the natural order of sites within a sequence [11–13 , 55 , 56] . Similarly , for reconstruction methods based on collections of subtrees , one could observe and use the relative position of the different genomic regions supporting the input trees . However , these relative positions must be conserved across the genomes being analyzed , a condition which may hold for recombining organisms ( e . g . individuals within a population or different viral strains ) , but which is not obvious when studying a group of taxa that have undergone reticulate events ( e . g . , hybridization ) at some point in a distant past . The main conclusion of the present study is the following: unless one abandons any optimization criterion that scores a network solely based on the trees it displays , the reconstruction should be carried out in a reduced space of networks: that of the canonical forms defined here . The motivation for this lies in the fact that canonical networks are guaranteed to be uniquely determined , if sufficient data are available . Once a canonical form N ̂ is inferred , it must be kept in mind that even assuming that the inference is free of statistical error , the true phylogenetic network is just one of the many networks having N ̂ as canonical form . Compared to what biologists are used to for phylogenetic trees—where in principle it is always possible to resolve uncertainties—it is clear that this requires an important change of perspective .
In order to prove that any network N has a canonical form , we describe an algorithm to transform N into a canonical network indistinguishable from N . The algorithm simply consists of repeatedly applying to N = ( V , E , φ , Λ ) one of the following two reduction rules , until neither can be executed ( see Fig . 12 ) : Funnel suppression ( R1 ) . Given a funnel v with k ≥ 1 in-edges ( u1 , v ) , ( u2 , v ) , … , ( uk , v ) and out-edge ( v , w ) , remove v and all these edges from N and introduce k new edges ( u1 , w ) , ( u2 , w ) , … , ( uk , w ) . For all i ∈ {1 , 2 , … , k} assign to ( ui , w ) the lengths Λ ( ( ui , w ) ) : = Λ ( ( ui , v ) ) + Λ ( ( v , w ) ) , where the sum of two sets of numbers A and B is defined as A + B = {a + b: a ∈ A , b ∈ B} . Multiedge merging ( R2 ) . Given a collection of multi-edges ( u , w ) with multiplicity k and lengths Λ 1 ′ , Λ 2 ′ , … , Λ k ′ , replace these edges with a single edge with lengths ⋃ i = 1 k Λ i ′ . An example of the reduction of a network to its canonical form is shown in Fig . 13 . Note that , even if the algorithm may temporarily produce multi-edges , the network produced in the end obviously does not have any multi-edge ( otherwise we could still apply rule R2 ) . Proof of part ( i ) of Theorem 1 . We must prove that any network N = ( V , E , φ , Λ ) has a canonical form . For this , we apply the reduction algorithm described above , thus obtaining a sequence N0 = N , N1 , … , Nm , where each Ni+1 is obtained from Ni by applying either R1 or R2 . Neither R1 nor R2 can be applied to Nm . We prove that Nm is a canonical form of N . Although , strictly speaking , Ni may not be a network ( as it potentially contains multi-edges ) , the notion of trees displayed by Ni , and thus that of indistinguishability , trivially extends to these multigraphs . First , note that the algorithm terminates after a finite number of iterations ( m ) . This is true because at each iteration the size of E is reduced by at least one . Moreover , the resulting network Nm is funnel-free , since no reduction of type R1 can be applied to it . What is left to prove is that Nm is indistinguishable from N = N0 . To this end we prove that , at each iteration , Ni and Ni+1 are indistinguishable , i . e . 𝓣 ( Ni ) = 𝓣 ( Ni+1 ) . In other words any tree T is displayed by Ni if and only if T is displayed by Ni+1 . Let T be displayed by Ni . Then T can be obtained by suppressing all suppressible nodes from a tree Ti contained in Ni . We consider three cases . ( 1 ) If none of the edges in Ti is involved in the reduction transforming Ni into Ni+1 , then clearly Ti is still contained in Ni+1 and thus T is still displayed by Ni+1 . ( 2 ) If Ti is involved in a R1 reduction , then it contains a funnel v and it contains one of the in-edges of the funnel , say ( uj , v ) , with length λj ∈ Λj = Λ ( ( uj , v ) ) , along with the out-edge ( v , w ) , with length λ0 ∈ Λ0 = Λ ( ( v , w ) ) . Now , let Ti+1 be the tree obtained from Ti by suppressing the suppressible node v and thus creating a new edge ( uj , w ) with length λj + λ0 . Because the R1 reduction creates a new edge ( uj , w ) with length set Λj + Λ0 , containing the value λj + λ0 , then Ti+1 is contained in Ni+1 . Moreover , it easy to see that T can still be obtained by suppressing all suppressible nodes from Ti+1 . Thus T is still displayed by Ni+1 . ( 3 ) If Ti is involved in a R2 reduction , then it contains one of the edges of a multi-edge ( u , w ) , with a length λ belonging to one of the length sets Λ 1 ′ , Λ 2 ′ , … , Λ k ′ associated to the k copies of ( u , w ) . Thus we have that λ ∈ ⋃ i = 1 k Λ i ′ , which implies that Ti is still contained in Ni+1 and thus T is still displayed by Ni+1 . This concludes the proof of 𝓣 ( Ni ) ⊆ 𝓣 ( Ni+1 ) . In order to prove that , conversely , 𝓣 ( Ni+1 ) ⊆ 𝓣 ( Ni ) , one can proceed in a similar way as above: if T is displayed by Ni+1 , then T can be obtained by suppressing all suppressible nodes from a tree Ti+1 contained in Ni . By considering three cases analogous to the ones above regarding the involvement of Ti+1 in the reduction transforming Ni into Ni+1 , we can prove that in all these cases T is already displayed by Ni . Thus Ni and Ni+1 are indistinguishable , which concludes our proof . □ We note informally that the order of application of the possible reductions in the algorithm above is irrelevant to the end result . To see this , it suffices to show that if two different reductions are applicable to a network , then the result of applying them is the same irrespective of the order of application . As we do not need this remark for the other results in this paper , we do not give a formal proof of it . Lemma 1 . Let N be a network and N′ a canonical form of N obtained by applying the reduction algorithm . If N satisfies the NELP property , then N′ satisfies the NELP property . Proof: We prove that for each basic step of the reduction algorithm—transforming Ni into Ni+1 via a reduction rule R1/R2—if Ni satisfies the NELP property , then Ni+1 also satisfies it . Suppose the contrary; then , Ni+1 contains two distinct weighted paths ρ1 , ρ2 with the same endpoints u and v and same lengths . Because R1/R2 cannot create new nodes , u and v are also nodes in Ni . Moreover , it is easy to see that each weighted path ρ in Ni from u to v gives rise to exactly one weighted path f ( ρ ) in Ni+1 from u to v , with exactly the same length as ρ . Now take two weighted paths in Ni , one in the preimage f−1 ( ρ1 ) and the other in the preimage f−1 ( ρ2 ) . These two weighted paths in Ni are distinct ( as ρ1 ≠ ρ2 ) , have the same endpoints ( u and v ) and the same length . But then Ni violates the NELP property , leading to a contradiction . We thus have that if Ni satisfies the NELP property , then Ni+1 also satisfies it . By iterating the argument above for each step in the reduction algorithm , the lemma follows . □ The proof of Theorem 1 , part ( ii ) , is rather technical . In this section , we introduce a number of new concepts and state the main intermediate results that are necessary to obtain this result . We leave their detailed proofs to S1 Text , together with the obvious definitions of basic concepts such as that of isomorphic networks , sub-network and union of two networks . Definition 3 . ( Root-leaf path , prefix , postfix , wishbone , crack . ) Let N be a network on 𝓧 and ( π , λ ) be a weighted path in N from the root of N to a leaf labelled by x ∈ 𝓧 . Now consider the sub-network P = ( V ( π ) , E ( π ) , φ∣{x} , λ ) on {x} consisting of all the nodes and edges in π and associated labels . Any sub-network of N such as P is called a root-leaf path of N . Given a root-leaf path P and a node v belonging to it , any weighted path formed by all the ancestors [descendants] of v in P is a prefix [suffix] of P . Note that a prefix [suffix] only consists of one node when v is the root [leaf] of P . A wishbone of N is any sub-network of N formed by taking the union of two root-leaf paths that have in common only a prefix . A crack of N is any sub-network of N formed by taking the union of two root-leaf paths that have in common only a prefix and a suffix . Fig . 14 illustrates the definitions above . Note that any root-leaf path P is both a wishbone and a crack , as P is the result of the union of P with itself , and P has a common prefix and a common suffix with P . Moreover , any sub-network R that can be obtained from a root-leaf path by attributing two lengths to one of its edges e is a crack . Finally , note that wishbones and cracks are networks , and thus the notion of isomorphism ( Definition 5 in S1 Text ) can be applied to them . The proof of part ( ii ) in Theorem 1 depends on two important results ( Propositions 1 and 2 below ) , whose proofs can be found in S1 Text . The first states that a network with the NELP property is uniquely determined by the wishbones and cracks it contains . Proposition 1 . Two networks N1 and N2 with the NELP property are isomorphic if and only if they contain the same wishbones and cracks ( up to isomorphism ) . Proposition 1 is interesting on its own as it suggests an enumerative algorithm to verify whether two networks with the NELP property are isomorphic . Unfortunately this algorithm would be impractical , as the number of wishbones ( or cracks ) in a network is not polynomial in the size of the network . Also note that we require N1 and N2 to satisfy the NELP property because there exist non-isomorphic networks containing the same wishbones and cracks: for example the networks in the bottom line of Fig . 10 . The second result that we need is the following: Proposition 2 . Let N1 and N2 be two indistinguishable funnel-free networks , satisfying the NELP property . Then they contain the same wishbones and cracks ( up to isomorphism ) . Proof of part ( ii ) of Theorem 1 . Let N be a network with the NELP property and N′ a canonical form of N obtained by applying the reduction algorithm . By Lemma 1 , N′ satisfies the NELP property . Now suppose that there exists another canonical form of N , called N′′ , satisfying the NELP property . By transitivity , N′ and N′′ are indistinguishable . Because N′ and N′′ are indistinguishable , funnel-free and with the NELP property , N′ and N′′ must contain the same wishbones and cracks ( because of Proposition 2 ) . But then , because of Proposition 1 , N′ and N′′ are isomorphic . □ We note that some of our arguments in S1 Text lead us to conjecture that a funnel-free network satisfying the NELP property cannot be indistinguishable from a funnel-free network violating the NELP property . This claim would allow us to simplify the statement of Theorem 1: networks with the NELP property would be guaranteed to have a unique canonical form ( not just among networks with the NELP property , but among all networks ) . Unfortunately , to this date , we were unable to prove this conjecture . Nonetheless , note that the reduction algorithm returns , for any network with the NELP property , its unique canonical form with the NELP property ( by Lemma 1 ) . It remains to prove the two corollaries at the end of the Results section . The first one states that two networks N1 and N2 satisfying the NELP property are indistinguishable if and only if their unique canonical forms with the NELP property , N 1 ′ and N 2 ′ respectively , are isomorphic . By Lemma 1 , N 1 ′ and N 2 ′ can be obtained by applying the reduction algorithm to N1 and N2 . Proof of Corollary 1 . The if part trivially follows from the transitivity of indistinguishability . As for the only if part , note that ( again by transitivity ) N 1 ′ is indistinguishable from N2 . As it is also funnel-free , N 1 ′ is a canonical form of N2 . Because N2 can only have one canonical form satisfying the NELP property ( by Theorem 1 ( ii ) ) , N 1 ′ and N 2 ′ must be the same network ( up to isomorphism ) . □ As for Corollary 2 , we recall that it states that a canonical network N with the NELP property is uniquely determined by the trees it displays . Proof of Corollary 2 . Let N and N′ be indistinguishable canonical networks satisfying the NELP property . Then , N and N′ are both canonical forms of N satisfying the NELP . But then , by Theorem 1 ( ii ) , N and N′ must be the same network ( up to isomorphism ) . □
|
We consider here an elementary question for the inference of phylogenetic networks: what networks can be reconstructed . Indeed , whereas in theory it is always possible to reconstruct a phylogenetic tree , given sufficient data for this task , the same does not hold for phylogenetic networks: most notably , the relative order of consecutive reticulate events cannot be determined by standard network inference methods . This problem has been described before , but no solutions to deal with it have been put forward . Here we propose limiting the space of reconstructible phylogenetic networks to what we call “canonical networks” . We formally prove that each network has a ( usually unique ) canonical form—where a number of nodes and branches are merged—representing all that can be uniquely reconstructed about the original network . Once a canonical network N ̂ is inferred , it must be kept in mind that—even with perfect and unlimited data—the true phylogenetic network is just one of the potentially many networks having N ̂ as canonical form . This is an important difference to what biologists are used to for phylogenetic trees , where in principle it is always possible to resolve uncertainties , given enough data .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Reconstructible Phylogenetic Networks: Do Not Distinguish the Indistinguishable
|
In the present paper , we quantify , with a rigorous approach , the nature of motor activity in response to Deep Brain Stimulation ( DBS ) , in the mouse . DBS is currently being used in the treatment of a broad range of diseases , but its underlying principles are still unclear . Because mouse movement involves rapidly repeated starting and stopping , one must statistically verify that the movement at a given stimulation time was not just coincidental , endogenously-driven movement . Moreover , the amount of activity changes significantly over the circadian rhythm , and hence the means , variances and autocorrelations are all time varying . A new methodology is presented . For example , to discern what is and what is not impacted by stimulation , velocity is classified ( in a time-evolving manner ) as being zero- , one- and two-dimensional movement . The most important conclusions of the paper are: ( 1 ) ( DBS ) stimulation is proven to be truly effective; ( 2 ) it is two-dimensional ( 2-D ) movement that strongly differs between light and dark and responds to stimulation; and , ( 3 ) stimulation in the light initiates a manner of movement , 2-D movement , that is more commonly seen in the ( non-stimulated ) dark . Based upon these conclusions , it is conjectured that the above patterns of 2-D movement could be a straightforward , easy to calculate correlate of arousal . The above conclusions will aid in the systematic evaluation and understanding of how DBS in CNS arousal pathways leads to the activation of behavior .
Deep Brain Stimulation is currently being used in the treatment of Parkinson’s Disease , Disorders of Consciousness ( DoC ) and clinical depression [1–3] . The possibility that Deep Brain Stimulation ( DBS ) could be used to enhance brain arousal is a subject of immense interest , e . g . , in traumatic brain injury ( TBI ) . For example , in a human patient who had suffered DoC for more than seven years , DBS of the central thalamus was used successfully to aid in the recovery of his consciousness ( Schiff et al ( 2007 ) ) [4] , Schiff ( 2010 ) [5] . From a fundamental neuroscientific point of view , this has been conceptualized as an elevation of generalized CNS arousal ( Pfaff , 2006 ) [6] . In the mouse , locomotion is the most elementary of behavioral responses , and in this work we utilize such movement patterns as the basis of our inference relating DBS parameter changes to behavioral effects ( Leshner and Pfaff ( 2011 ) [7] , Benjamini et al ( 2011 ) [8] , Quinkert et al ( 2010 , 2011 , 2012 ) [9–11] ) . In the present study , the stimulations occur in the central thalamus ( cf . Schiff et al [4] ) , over fixed 10 min intervals , every three hours , eight per day ( four in light , four in dark ) . There are both stimulated and control ( electrodes implanted but nonstimulated ) mice . What is observed are the x- and y-coordinates of location , per second , over three days , with there being 12 hours of light , 12 hours of dark . The range of DBS parameters consist of three amperages and four frequencies , and were applied to each mouse over the three days ( reported by Quinkert et al [9] ) . A common motor activity summary statistic is Total Activity , i . e . , the total distance traveled over some fixed time interval ( e . g . , 10 min ( 600 sec ) ) , or equivalently , Mean Activity ( or Mean Speed ) : Total Activity divided by the number of time points ( e . g . , 600 ) . Two questions that arise in the use of such statistics are , first , is there is a loss of important information in such Mean ( or Total ) Activity summarization; for instance , are important angular changes in direction or differences in the spatial range of movement , lost ? Secondly , how does one calculate a standard error ( or make a probabilistic assessment ) for any statistic derived from such time-varying location data , doing so in a manner that can be justified ? One needs to appropriately account for both local time-correlations , as well as broader circadian changes , otherwise the calculations may not be representative . In a control mouse , or a stimulated mouse if there were no DBS effect , our basic assumption is that the time-varying processes constructed from the motor activity , are piecewise stationary . We will verify and apply piecewise stationarity for division into 3-hr segments , although longer periods could also be justified . We show that the autocorrelations within such a stationary segment die out after 20 minutes; segments of motor activity separated by 20 min can be assumed to be uncorrelated ( or , in the present context , it is reasonable to assume independence ) . In the case of the stimulated mouse , we show that autocorrelations also die out after 20 min for regions sufficiently separated from a stimulation interval . The first stage of the analysis is to show that the stimulation is effective for at least one of the DBS parameter values . In this analysis , all calculations are on segments separated by 20 min; that is , our statistics for each animal are based on the ±80 minutes , centered at each stimulation time , leaving at least 20 min between any of the time intervals on which calculations are to be made . The statistic calculated on each ( independent ) segment results in a null hypothesis of no effect . A False Discovery Rate ( FDR ) thresholding then establishes that there is at least some DBS effect . Once such a DBS effect is established , piecewise stationarity is used in a slightly different manner , since stationarity is being briefly perturbed by the stimulations to new steady-states , with possibly different stationary mean and/or the covariance structure for the 10 min stimulation intervals , than for nearby non-stimulated time intervals . That is , after establishing that there is some DBS effect , the determination of the relationship of DBS parameters to motor activity will need to be based on the collection of 10 min stimulation intervals , with the means and/or covariance structures for distinct such intervals possibly varying with the DBS parameter values . Differences in the stationarity structure for the different 10 min stimulation intervals , as we will see , result in significant differences in the variances for the calculated statistics , and hence does not allow for a traditional ANOVA formulation or nonparametric method . We will also consider differences in Mean Activity between light and dark , and piecewise stationarity will be the basis of that model . As part of our analysis , we analyze speed and angular patterns and also decompose the movement into its randomly occurring epochs of 0- , 1- , and 2-dim movement . One-dimensional ( linear ) movement appears to be not unlike background noise in the present context , which is why the decomposition was formulated . The main ideas of the present paper important to neuroscience are: ( Hypothesis I ) The stimulation is effective; ( Hypothesis II ) It is 2-D movement , not 1-D movement , that occurs in response to stimulation , and there is a detectable relationship between DBS amplitude and frequency and the resulting movement; and , ( Hypothesis III ) It is 2-D movement , not 1-D movement , that differs between light and dark; and , stimulation in the light initiates a manner of movement ( 2-D movement ) more commonly seen in the ( non-stimulated ) dark . To address these three hypotheses , we consider seven motor processes and their resulting statistics . For each , we calculate a forward 10 min mean ( i . e . , forward in time from each given sec ) and a right minus left difference in 2 min means ( at each sec ) ; the later can detect rapid changes . The focus of several ( 3–4 ) of the statistics ( those that are one-dimensional ) is not on their power of information extraction , but rather the opposite; these statistics have basically no power and should be subtracted off , in order to obtain better statistics ( those that are purely two-dimensional ) .
All animal procedures were in compliance with National Institutes of Health guidelines and approved by the Rockefeller University Institutional Animal Care and Use Committee . Details of methodology regarding neurosurgery and behavior have been published ( Quinkert et al [9] ) . Briefly , mice were singly housed with food and water available and were subjected to a 12 h light/dark cycle . Stainless steel monopolar electrodes ( 0 . 3 mm diameter ( Plastics One ) ) were insulated using polymide with 0 . 2 mm stripped from the electrode tips , with electrodes implanted bilaterally in the central thalamus . Stimulation was programmed and delivered by a four-channel stimulus generator ( Multichannel systems STG2004 ) . Stimulation epochs lasted for 10 min and occurred every 3 h over the course of 3 days . All stimulations were biphasic with a pulse width of 0 . 1 ms on both anodic and cathodic phases of the pulse . Three pulse amplitudes were applied , each for one day: 75 , 100 and 125 μA . Four pulse frequencies were selected from 50 , 125 , 175 and 225 Hz . The mice were euthanized and then a histological assessment of electrode placement was made , following data collection . Here , data are analyzed from novel points of view . Nine mice were studied , with electrodes implanted in each; stimulation was applied in six mice ( Mice 1–6 ) , with three others used as controls ( without stimulation , Controls 1–3 ) . Four of the six stimulated mice ( Mice 1–4 ) showed significant responses to the stimulations , whereas Mice 5–6 did not appear to respond at all; it was established via histological investigation that differences in electrode placement explained the nonresponsiveness of Mice 5–6 . They were included in the initial analysis as a matter of completeness . The number of stimulations ( Nstim ) was 8 × 3 = 24 for mice 2 , 4 , 5 and 6 , and was 23 for mice 1 and 3 ( the recording of the 24-th stimulation period was not complete ) . Home cage activity data was collected by a 3D infrared monitor ( Accuscan Instruments ) , which records the locations at the times of a detected change . The times between changes varied from the millisecond scale to that of multiple seconds; the data were interpolated to the one second scale , for computational purposes . In the present mouse experiments , what is observed are the time-evolving locations ( x- and y-coordinates ) of the animal . The stimulation data is observed over 3 days , with N = 3*24*3600 = 259200 seconds . The stimulation times are denoted as {Sk , k = 1 , … , Nstim} , with Nstim being the number of stimulations . From the location data r ( ti = ( x ( ti ) , y ( ti ) ) , ti = 1 , … , N , one can calculate the ( discrete-time ) Velocity , Speed and Angle Direction processes: r _ ( t i ) = ( x ( t i ) , y ( t i ) ) V _ ( t i ) = ( V x ( t i ) , V y ( t i ) ) = r _ ( t i + 1 ) − r _ ( t i ) , , t i = 0 , … , N − 1 M ( t i ) = x 2 ( t i ) + y 2 ( t i ) , U _ x = ( 1 , 0 ) , θ ( t i ) = angle ( V _ ( t i ) , U _ x ) ( 1 ) The angle θ ( ti ) is defined with respect to the positive x-axis ( Ux ) and is uniquely defined in [−π , π ) , with −π identified with π ( i . e . , the angles are on the unit circle ) . Since the location data are fully recoverable from the latter two and the initial location: r _ ( t i ) = ∑ j = 0 i − 1 M ( t j ) × ( c o s ( θ ( t j ) ) , s i n ( θ ( t j ) ) ) , ( 2 ) the Magnitude and the Direction of Angle processes are the basis for our analysis . Hence , we model the changing patterns of the following motor activity processes: M ( ti ) : Speed , or Magnitude of the velocity , and by summing over any time interval one can obtain the Total Activity ( i . e . , Total Distance ) . θ ( ti ) : Angle of direction as a function of time ( relative to the fixed x-axis ) . Moreover , one can decompose these angles into two groups: θ ( P ) ( ti ) : those that are Multiples of π/2 ( e . g . , continuation in same direction , a reversal or a perpendicular move ) , including movement parallel to the walls; and θ ( NP ) ( ti ) : those that are Non-Multiples of π/2 ( here , e . g . , movement into the interior , non-parallel to a wall ) ; Movement Pattern over a w = 30 sec window , calculated in a forward direction starting at each second , and decomposed into three groups ( uniquely defined at each time point ) : ( D ( 0 ) ( ti ) = 0 , for zero-dimensional movement ) D ( 1 ) ( ti ) : length of line segment containing the 1-dim movement ( if points in time all lie on a line , and are not constant ) . One can determine if the movement over the 30 sec window beginning at ti , is one-dimensional , and if so , to calculate the length of its 1-D domain . D ( 2 ) ( ti ) : area of the two-dimensional convex hull generated by the points ( if they do not lie on a line ) . Because the two-dimensional path can , and often does , cross itself , defining the 2-D domain is not straightforward or necessarily well defined . A natural definition is the use of the convex hull . The convex hull of the points ( x ( ti+r ) , y ( ti+r ) , r = 0 , 1 , … , w − 1 ) , over a moving window of length w = 30 seconds , is determined and its two-dimensional area D ( 2 ) ( ti ) is calculated . ID ( ti ) : identifies at each time whether the movement is zero- , one- or two-dimensional , by 0 , 1 or 2 ( based upon ( i ) and ( ii ) , above ) . The Speed M ( ti ) , when non-zero , can be decomposed into the Speed at times at which the movement is two-dimensional or one-dimensional: M ( ti ) = M2D ( ti ) + M1D ( ti ) M2D ( ti ) = M ( ti ) if ID ( ti ) = 2 , and 0 otherwise; M1D ( ti ) = M ( ti ) if ID ( ti ) = 1 , and 0 otherwise; One can define A ( ti ) to be the Total Activity over the 10 min window ( w1 = 600 ) starting at ti ( i . e . , the sum of M ( tj ) over the interval ) . For simplicity , we will use the Mean Activity A ¯ ( t i ) rather than Total Activity , in that the derivation of standard errors is more direct . The difference is merely one of scale . Based upon ( d ) above , summing over M2D ( ti ) and M1D ( ti ) , respectively , A ¯ ( t i ) can be decomposed into the sum of 2-Dim and 1-Dim Mean Activity: ( zero-dim movement adds zero ) A ¯ ( t i ) = 1 w 1 × ∑ j = 0 w 1 − 1 M ( t i + j ) = 1 w 1 × ∑ j = 0 w 1 − 1 M 2 D ( t i + j ) + 1 w 1 × ∑ j = 0 w 1 − 1 M 1 D ( t i + j ) = A ¯ 2 D ( t i ) + A ¯ 1 D ( t i ) ( 3 ) As we will see , it is the 2-Dim activity that differs most significantly between light and dark , as well as that which predominates in response to stimulation . That is , we will show that it is the statistics that extract 2-Dim information that are informative concerning responsiveness to DBS stimulation , and not those that are 1-dimensional . In Figs . 1–2 , there are two panels ( A and B ) in each , one for high ( A ) and low ( B ) parameter values; Fig . 1 is during the dark , Fig . 2 is during light , and they are for Stimulated Mice 1–2 , respectively . In the first row of each panel are displayed the actual time-varying two-dimensional ( x- and y- ) position ( per sec ) , over a sequence of four 10 minute intervals , starting 20 min prior to a stimulation , and including the 10 min stimulation interval and the 10 min interval following it . In the second and third rows are the individual x- and y-coordinate patterns , from which one can infer movement parallel or non-parallel to a wall . In the remaining rows are the above-described time-evolving motor processes , appropriately plotted; in row 4 are the time changing patterns of D ( 1 ) ( ∙ ) ( length , red ) and D ( 2 ) ( ∙ ) ( area , blue ) ; in row 5 are the analogous plots for Speed: M1D ( ∙ ) ( 1D , red ) and M2D ( ∙ ) ( 2D , blue ) ; and , in row 6 , are the time-evolving θ ( P ) ( ∙ ) ( multiple of pi/2 , green ) and θ ( NP ) ( ∙ ) ( non-multiple of pi/2 , blue ) . The seventh motor process , Mean Activity M ( ∙ ) , being the sum of M1D ( ∙ ) and M2D ( ∙ ) , was not plotted , for simplicity . The figures very much depict the motivation for the 1D and 2D decompositions , and their use in quantifying the stimulation response . We will utilize two basic statistical calculations , defined below in expression ( 4 ) : ( 1 ) a forward mean over a moving window of time ( forward meaning that the value at time ti is for the window starting at ti; the window width being w1 = 600 sec ) ; and ( 2 ) a difference in means ( i . e . , that to the right minus that to left ) , for a moving window of time ( window width being w2 = 120 sec , on each side ) . The latter statistic acts as a high-pass filter , detecting the rapid onset of movement . We denote the two statistics , respectively , as X¯ ( ∙ ) and X¯ R L ( ∙ ) : X¯ ( t i ) = 1 w 1 × ∑ j = 0 w 1 - 1 X ( t i + j ) and X¯ R L ( t i ) = 1 w 2 × ( ∑ j = 0 w 2 - 1 X ( t i + j ) - ∑ j = - w 2 - 1 X ( t i + j ) ) ( 4 ) where X generically represents any of the following seven motor activity processes: A 1 D ( t i ) , D ( 1 ) ( t i ) , θ ( P ) ( t i ) , θ ( N P ) ( t i ) , A ( t i ) , A 2 D ( t i ) , D ( 2 ) ( t i ) ( 5 ) The first three are 1-Dim statistics ( 1-D Activity , 1-Dim Length , Multiple of pi/2 angle ) . The next two are 2-dimensional Non-Multiple of pi/2 angle , Total ( 1-D and 2-D , combined ) Activity , but are incomplete in certain ways ( Results ) . The final two 2-dimensional statistics ( 2-D Activity , 2-Dim Area ) are those that are of greatest potential interest . In Fig . 3 , for Stimulated Mouse 1 , the time-varying ( per sec ) statistics were averaged over a moving 10 min window starting at each second , moving second by second; the statistics were: mean activity , fraction of non pi/2 and fraction of pi/2 angles , mean area ( of 2-D movement ) and the mean length ( of 1-D movement ) . The red asterisks denote the stimulation times , and are plotted at the height of the recorded response at the stimulation time to accentuate the magnitude ( or lack of ) in response to the stimulus . For the angle processes , the statistics calculated are fractions , rather than means . Let X ( ti ) generically represent any of the following seven Motor Activity Processes given in expression ( 5 ) , above . Our basic assumption is that , in the non-stimulated animal , or in the stimulated if there were no DBS effect ( Hypothesis I ) , there is piecewise stationarity ( in time ) for motor processes under consideration , calculated from the location data r ( ti = ( x ( ti ) , y ( ti ) ) , ti = 1 , … , N . Below , we show such piecewise stationarity for the speed ( M ( ti ) ) process , but others ( e . g . , the angle process ) can similarly be shown . Specifically , we assume that time can be decomposed into a finite set of time segments , for which the process of interest is assumed to be representable as a strictly stationary process on each segment . On different time segments , the structural parameters of stationarity ( mean , variance , autocovariances , spectral density ) are allowed to differ . We justify below the use of a decomposition into 3 hr time segments ( although longer segments could be justified ) . Once it is shown that there is ( some ) DBS effect , piecewise stationarity will no longer hold in the same form , in that the local stationarity at the stimulation intervals is being perturbed by the stimulations to new brief steady-states on these 10 min stimulation intervals , or for possibly longer ( as it returns to its original steady-state ) . In testing Hypothesis II , only the 10 min stimulation intervals will be used ( they are now stationary at the perturbed stead-state , potentially different for different stimulation levels ) ; in testing Hypothesis III , the stimulation intervals will be excluded . In modeling piecewise stationarity , as in stationarity , one can model from either the time- or frequency-domain . In the present work , we have chosen to use a time-domain approach , primarily because it made the analysis for the three hypotheses , as a whole , more unified . There has been a great deal of work on modeling time-varying spectra ( see Ombao et al ( 2001 ) [12] , Huang et al ( 2004 ) [13] ) . We first establish the overall pattern of changing stationarity . In Fig . 4 , there are four panels ( A–D ) . In A , left , for a single control mouse , the local mean is calculated for the speed ( M ( ti ) ) process , over a moving window of width 10 min ( blue ) , 1 hr ( green ) and 3hr ( red ) ; on the right , are the time-varying means for all three controls , over a moving window of width 3 hrs . Mouse movement typically consists in random bursts of movement , of random lengths , interspersed with low or no movement periods , of random lengths . One can view such patterns as being doubly stochastic , with the first level describing whether there is or is not movement . The mean patterns reflect such behavior , where the values locally can be quite variable . In B , there are two rows . The first displays the sample autocovariance functions ( over 1 hr = 3600 sec ) ) for all three controls , starting at the 4th hr and at the 6th hr , in light ( left ) and dark ( right ) , six functions in total . The second displays the same , calculated at the 8th and 10th hrs . The 4 and 6 hrs represent early behavior in the 12 hrs , whereas the 8 and 10 hrs represent late behavior for the 12 hrs . Panel C displays the two autocorrelation functions calculated from the mean of the , respective , early ( blue ) and late ( red ) autocovariance functions . The horizontal bands ( ± 1 . 96 / 3600 × 6 ) , which at a single lag is a confidence band about zero , are a standard time series practice to assess a “loss of correlation in time . ” The result of Panels A–C of Fig . 4 is that , for the control mouse , the means are relatively stable during light and during dark , differing for the two; the same is true for the autocovariance structure , being relatively stable during light and during dark , but differing for the two . The autocorrelation patterns indicate that it is reasonable to assume that the statistics calculated on intervals that are separated by 20 min or more can be assumed to be uncorrelated ( or , as we will , as independent ) . In the bottom panel ( D ) of Fig . 4 , the autocovariances and autocorrelations are examined , for the stimulated mice 1–4 , at times in between the broad range of stimulation intervals that are separated by 3 hrs . The autocovariances were calculated for 1 hr , starting 80 min after a stimulation interval . Again , one sees the same decay to negligible levels after 20 min , and hence , even in the stimulated case , intervals sufficiently separated can be assumed to be uncorrelated ( again , in the present case , we will assume independence ) . Specifically , in Hypothesis II , where statistics are only calculated on the 10 min stimulation intervals , separated by 3 hrs , it is reasonable to assume uncorrelation ( or , again , independence ) . Moreover , in Hypothesis III , in comparing 1D and 2D Mean Activity between light and dark , we will make calculations on three 3hr segments , separated by 1 hr , within light and within dark each , excluding the stimulation intervals and an additional 20 min following the stimulation . Hypothesis III concerns whether the means of the two ( light , dark ) are also different , in addition to their autocovariances . In Fig . 5 , top row are speed ( per sec ) data for a Control mouse ( hence , non-stimulated ) . The left column displays 12-hrs in light and the right column 12-hrs in dark . In the second row are sample autocovariance functions of the top row data ( per sec ) , calculated over 3-hr periods , starting at hrs 4 , 5 , 6 and 7 . The most common approach to testing stationarity , has been spectral ( squared modulus of the Fourier Transform ) , using the cumulative periodograms . The third row displays the estimated log spectral densities ( using Thomson’s multitapering in the construction ) , for the four 3-hr time intervals . From the auto covariance functions , a dark versus light comparison , shows variances that differ ( by a factor of more than five ) . However , within light and within dark , individually , the four variances show virtually no difference . Because of this , within light and within dark , we normalize the cumulative periodograms to a maximum of one ( i . e . , dividing by the variance ) . To test the equivalence of the four spectra , the Diggle-Fisher test ( 1991 ) [14] was performed , which is as follows . If the different sample spectra were all estimating the same true spectra , then shuffling the four spectral values at a given frequency , should statistically produce equivalent estimates . This is the basis of the test . One calculates the maximum difference at each frequency in the cumulative periodograms , and then the maximum of those over all frequencies . This is done for the actual observed cumulative periodograms and compared to the results for all the shuffling . The resulting histogram of the maximum difference ( over all frequencies ) of the shuffled cumulative periodograms , over 1000 shuffles , are displayed in the fourth row . The red asterisks denote the observed spectral maximum differences . The P-values for testing the hypothesis of stationarity during light is . 54 and during dark is . 75 . Various other tests of stationarity have been proposed ( e . g . , Priestley and Subba Rao ( 1969 ) ) [15] , often as analogues of a Kolmogorov-Smirnov like test , which have proven difficult to use in practice . Our basic assumption is that , over a 3-hr period , there exists a strictly stationary process that describes the chosen motor process ( e . g . , the speed or angle processes ) , and that it satisfies a mixing condition , specifically ϕ-mixing ( see Billingsley ( 1968 ) ) [16] , i . e . , that there exists a function ϕ ( ∙ ) , such that limn→+∞ϕ ( n ) =0 , and for any two events , F1 and F2 , F2 dependent upon the information up to time m , F1 dependent upon the information up to time m+n , n ≥ 0 and P ( F2 ) > 0 , they satisfy ∣P ( F1∣F2 ) −P ( F1 ) ∣ ≤ ϕ ( n ) . We also assume that its autocovariances are absolutely summable ( and hence a continuous spectral density exits ) . Phi ( ϕ ) -mixing is a very weak assumption which describes the rate at which time-dependency dies out . From Figs . 4–5 , this is a very reasonable assumption of the dying out of the dependencies . Since X¯ ( ∙ ) and X¯ R L ( ∙ ) are finite linear filters of X ( ∙ ) , they also are strictly stationary and satisfy the same ( form of ) mixing condition . Let FX¯ ( t j ) denote the marginal distribution of X¯ ( ∙ ) at a ( arbitrary ) single time point tj , which by strict stationarity is the same at all times tj ( in the segment of stationarity ) . Under the assumption of ϕ-mixing , the empirical distribution function Fn ( ∙ ) : F n ( x ) = 1 n ∑ j = S k - 80 * 60 S k + 80 * 60 I [ X¯ ( t j ) ≤ x ] , where n is the number of terms in the sum , is asymptotically equivalent to FX¯ ( ∙ ) ( see Billingsley ( 1968 ) ) [16] . Specifically , there is uniform weak convergence , with s u p x n ( F n ( x ) − FX¯ ( x ) ) converging to a Gaussian process ( indexed by x ) . Hence , we have that limn→+∞Fn ( x ) =FX¯ ( x ) , uniformly in x . As a consequence , probability calculations under Fn are asymptotically equivalent to those under FX¯ . Since Sk is a fixed time ( the k-th stimulation time onset ) , if there were no effect due to the stimulation at time Sk , then X¯ ( S k ) is random with the probability distribution FX¯ . Evidence against the hypothesis that there is no effect due to the stimulation at time Sk , can hence be measured by the probability of observing a value greater than or equal to X¯ ( S k ) under Fn , i . e . , a P-value for each k , k = 1 , … , Nstim . This comparison can be interpreted as a permutation test except that the permutations are restricted to being translations ( or rotations if viewed on a circle ) . This restriction is a direct consequence of the test adhering , as is necessary , to the piecewise stationarity . The same results apply to the empirical distribution function for X¯ R L . Thus , for Hypothesis I , the question is ( where X¯ represents any of the seven motor processes ) : Is X¯ ( S k ) , for a fixed k , k = 1 , … , Nstim , significantly greater than most X¯ ( t j ) , for tj in the ±80 min ( 160 min ) period centered at Sk ? If so , this is ( probabilistic ) evidence that the stimulation caused a change in motor activity . The same question applies to X¯ R L ( S k ) . In each case , one calculates the proportion of the values that lie above X¯ ( S k ) and X¯ R L ( S k ) , respectively . Representative P-values ( k = 1 , … , NStim ) are presented in Fig . 6 for Mouse 1 and Control 1 . If one wishes to make an assessment of the effect due to a given amperage level ( i . e . , the level over a given day ) , one can apply a False Discovery Rate ( FDR ) procedure ( presented in Results ) ( Benjamini and Hochberg ( 1995 ) [17] , Benjamini and Yekutieli ( 2001 ) [18] ) . In the case of Hypothesis II , once Hypothesis I has been affirmed ( that there are motor effects due to the stimulations ) , then one must establish the standard error for the statistics , using only that particular 10 min stimulation interval , since the structure on that interval can now be quite different from that of neighboring non-stimulation intervals or other 10 min stimulation intervals ( corresponding to different DBS parameter values ) . Hence for the Mean Activity ( and its 2D and 1D components ) , obtained by summing over the 10 min stimulation interval , one needs to base its standard error calculation on the stationarity of the speed over same 10 min stimulation interval . For a stationary time series with a covariance function that dies out sufficiently fast so that the spectral density exists and is continuous ( which occurs under our assumptions ) , the asymptotic variance of a sample mean ( sample size n ) of the process is: lim m → + ∞ ( γ ( 0 ) + 2 ∑ h = 1 m − 1 ( 1 − h m ) γ ( h ) ) ( 6 ) and the standard approximation to this , is to replace the covariances with their sample estimates , using a number of terms ( m ) of an order less than n . A standard practice is to use as the number of terms ( m ) , the square-root of n ( see Shumway and Stoffer ( 2011 ) [19] ) . Hence , the approximation can be applied to the various motor processes ( expression ( 5 ) ) , e . g . , A ¯ ( S k ) , A ¯ 2 D ( S k ) and A ¯ 1 D ( S k ) , A ¯ 1 D ( S k ) , k = 1 , 2 , … , Nstim . One might ask if it is more accurate to fit a time series ARMA ( p , q ) model to each stimulation 10 min interval , estimate the ARMA parameters and to then use covariance estimates based upon these finite number of parameters . For finding standard errors of most time series parameter estimates , that approach is superior; but for the sample mean , because of its linear construction , there is no improvement , in that the use of the sample covariances in the above construction , produces an asymptotically efficient estimator ( Grenander and Rosenblatt ( 1957 ) [20] , Priestley ( 1981 ) [21] ) . In testing Hypothesis II , the above standard errors for the motor processes are utilized ( Results , Fig . 8 ) . In order to identify the effect of dark versus light on one- and two-dimensional movement patterns , we break , for each of the 3 days , the 12 hrs ( 720 min ) of light and dark each into three 3hr segments , with 1 hr between: 30 to 210 , 270 to 450 , 510 to 690 . We do this for each of the nine mice . This breaks the 3 days into a sequence of eighteen 3hr segments , each separated by 1hr . In addition , certain times are excluded: those times during a stimulation and the 20 min following a stimulation ( for the Stimulated Mice ) , and the ±30 min at a light/dark transition are excluded by construction ( they are within the 1 hr separating the segments ) . On each segment we calculate the 1D and 2D Mean Activity ( A ¯ 1 D , A ¯ 2 D ) . We assume piecewise stationarity for the 3 hr segments . The sample autocovariances and means are calculated on each 3 hr segment . We do this for the 1D Mean Activity and the 2D mean Activity , separately . The variances of the means for each 3hr segment are calculated , as they were for the 10 min stimulation intervals , by applying expression ( 6 ) above . We then average these 3hr-based means over the dark and over the light , and take the difference . What we wish to test is whether or not there are differences in the means in light versus dark , for 1D and for 2 D Mean Activity . Let Y1Dim , D−L and Y2Dim , D−L denote these differences in means . Thus , we obtain values Y1Dim , D−L , EstVar ( Y1Dim , D−L ) , Y2Dim , D−L and EstVar ( Y1Dim , D−L ) . The test statistics are: Y 1 D i m , D − L / EstVar ( Y 1 D i m , D − L ) and Y 2 D i m , D − L / EstVar ( Y 2 D i m , D − L ) ( 7 ) For each of the nine mice , a P-value can be calculated for the dark/light comparison , for 1D and 2D Mean Activity , and FDR analysis of the P-values is applied . These values are given in Fig . 9A–B , as part of the testing of Hypothesis III .
The important ideas of the present paper are: ( Hypothesis I ) The stimulation is effective . The methodology confirms that the resulting movement at the stimulation times ( under appropriate DBS parameters ) is in fact stimulation-driven , delineating it from merely being coincidental endogenously-driven movement; ( Hypothesis II ) It is 2-D movement , not 1-D movement , that occurs in response to stimulation , and , the effects due to the three amperages ( 75 , 100 , 125 μA ) are statistically increasing in value and distinguishable . Moreover , there is a significant synergism at the combination of 125 μA and 125 Hz . In terms of the four frequencies , the effect due to 50 Hz was less than that for each of 125 , 175 , 225 Hz . There is not a significant response in 1-D movement to the stimulation; and , ( Hypothesis III ) It is 2-D movement , not 1-D movement , that differs between light and dark , and finally , stimulation in the light initiates a manner of movement ( 2-D movement ) more commonly seen in the ( non-stimulated ) dark . In order to establish the above ideas , we utilized two basic statistical calculations , defined above in Methods: ( 1 ) a forward mean over a moving window of width 10 min; and ( 2 ) a difference in right minus left means with moving window of width 2 min . The calculations are applied to each of seven motor activity processes: A 1 D ( t i ) , D ( 1 ) ( t i ) , θ ( P ) ( t i ) , θ ( N P ) ( t i ) , A ( t i ) , A 2 D ( t i ) , D ( 2 ) ( t i ) . The first three are 1-Dim statistics ( 1-D Activity , 1-Dim Length , Multiple of pi/2 angle ) ; we show that these statistics , are best removed from the resulting statistics , leaving only 2-dimensional components . The next two are 2-dimensional ( Non-Multiple of pi/2 angle , ( Total ( 1-D and 2-D , combined ) Activity ) , but lack certain strengths , e . g . , the Total Activity still has the 1-dim activity as a component , and the angle calculation has 2-dim information but has no velocity magnitude information . The final two 2-dimensional statistics ( 2-D Activity , 2-Dim Area ) are those that have the greatest strength . ( For the angle processes , the statistics calculated are fractions , rather than means . ) For Hypothesis I , to test that there is a response to ( at least some of ) the stimulations , the asymptotic distributions of our test statistics , under the assumption of piecewise stationarity , were derived in Methods . Statistically , one is going to calculate a single value , based upon that particularly given 10 min interval of stimulation . The starting point of each stimulation interval is surrounded by a ±80 min larger interval , with these larger intervals all separated from one another by at least 20 min . Hence , calculations on each ( based upon results of Methods ) are independent of one another . One wishes to show that the value of the statistic calculated for the 10 min stimulation interval is significantly different from the same calculation at an arbitrary point in the surrounding interval . The mouse is being stimulated by environmental cues all the time . One needs to show that the motor activity at the stimulation time was not just coincidental endogenously-drive movement , but rather significantly different from such . One cannot permute because of time dependency , and bootstrapping in a stationary context is difficult and involves various heuristic choices ( Kunsch ( 1989 ) [22] , Lahiri ( 2003 ) ) [23] ) . However , one can imagine , for any value in the surrounding 160 min interval , that one makes the same calculation for a 10 min interval starting at any point in the 160 min . Calculations at the limits of the 160 min interval , are made in a wrap-around manner ( i . e . , the interval is viewed as a circle ) . Such wrapping around is a standard time series/Fourier procedure that has no asymptotic effect; so doing allows us to make the calculations and keep the different 160 min intervals sufficiently separated . The sample cdf of all such translations ( rotations ) is asymptotically derived in Methods , under the stated conditions of ϕ-mixing and absolute summability of autocovariances , which are very weak assumptions . That is , the appropriate test statistic is the restriction of permutations to just those that are translations ( i . e . , rotations , if viewed as a circle ) ; only the translations adhere to the local time invariance of stationarity . In Fig . 6 , representative calculations , including the final P-values of the test statistics , are displayed for one stimulation mouse ( Panels A–B ) and one control mouse ( Panels C–D ) . There are four rows for each mouse . In the first row are shown the raw data on which each of four statistics are to be calculated; the second row shows the collection of values for each statistic , based upon the translations , plus the observe value of the statistic is displayed as a red asterisk . The third and fourth rows show , respectively , the resulting probability histogram and cumulative distribution function , with the observed value displayed as a red asterisk and the P-value is indicated . In Fig . 7 , Hypothesis I is tested , and the P-values for all mice and motor process statistics are summarized; the results of Fig . 6 are contained within these . Because the calculations are all made on distinct intervals separated by 20 min or more , as described in Methods , the calculations can be assumed to be uncorrelated or , more specifically , independent . For a given mouse , there are 24 ( or 23 for two ) stimulation intervals; each stimulation interval produces a null hypothesis for no DBS effect at those particular parameters ( Amperage , Hz , light/dark ) . For a chosen statistic , a P-value is obtained for each interval , for each mouse . In this multiple testing setting , we use a False Discovery Rate ( FDR ) thresholding at q = . 05 , to assess the evidence of significance . First , though , the first eight stimulation intervals are for amperage 75μA . This value was included as a baseline , with little expectation of response , but rather to be used as a reference point for the primary two: 100 , 125 μA . If one is assessing the strength of evidence , it does not seem appropriate to include these eight in any overall assessment , but to be considered separately . [It is analagous to including experiments with a placebo to determine if there is some physiological response to an agonist . ] For simplicity and compactness of display , we have shown , in some of the subplots , the results simultaneously for all seven of the motor processes . In Fig . 7 , displayed are the sorted P-values ( uncorrected ) for m null hypotheses , and the thresholding function ( j/m ) * ( q/cv ) ; in the independence case , cv = 1 , and in the correlated case , cv is the sum of reciprocal indices ( Benjamini and Hochberg ( 1995 ) [17] , Benjamini and Yekutieli ( 2001 ) [18] ) . As stated above , our calculations are independent for distinct intervals . In Fig . 7 , first row , we display the three amperages 75 , 100 and 125 μA , separately . One can argue that the three are different experimental conditions . In the second row , middle plot , the 100 and 125 μA null hypotheses are combined . As has been stated previously , one aspect of the work is to show that the three 1-dimensional statistics do not reflect DBS responsiveness , and should in fact be removed within appropriate statistics . The two statistics: Non-multiple of pi/2 ( contains 2D information , but lacks information about velocity magnitude ) and the Mean Activity ( combining 1D and 2D ) both lacked strength . The remaining two statistics are those of primary interest: 2D Mean Activity ( A ¯ 2 D ( S k ) ) and the Mean 2D Area ( D ¯ ( 2 ) ( S k ) ) . We are not trying to choose between the seven statistics; each identifies separate and distinct information . However , we include a plot ( second row , rightmost ) of the null hypotheses for the two primary statistics , which , since they are correlated , we use FDR in the correlated case . In the second row , leftmost plot , we consider the 3 controls . In the third row we display the results for the two nonresponsive mice ( Mice 5–6 ) , on the left . The other two subplots consider the Stimulated Mice 1–4 , and the second form of the calculations: Difference between the Right and Left Means , at each point; this can be viewed as a high pass filter or as a change-point detector . An expression of the strength of evidence in a given subplot is the number of significant hypotheses . One statement of the strength of evidence for at least some detectable DBS effect , would be the combined 100 , 125 μA plot for the Stimulated Mice 1–4 , which is the middle plot of row 2 . There , the four 2-dimensional statistics each have between 20 and 25 significant tests out of the combined 62 ( 32+30 ) . In all of the cases: Stimulated Mice 1–4 , Controls 1–3 , and Nonresponsive Mice 5–6 , the three 1-dimensional statistics never go below the threshold function . The results in Fig . 7 , with respect to Hypothesis I , are that: ( i ) there is no measurable response , for any of the seven processes , to the stimulations at the lowest ampere level of 75 μA; this was not unexpected , in that it was chosen to hopefully identify a baseline level; ( ii ) the ( total ) Mean Activity , Mean Activity 2D , the Fraction of Non-Multiples of pi/2 Directions and the Area 2D , represent measurable stimulation responses at the 100 and 125 μA; and , ( iii ) the Mean Activity 1D , Fraction of Multiples of pi/2 Directions and the Length 1D , show virtually no response . Once one has shown that there are motor responses specifically due to stimulation , one then proceeds to test ( Hypothesis II ) that these responses can be related to the stimulation parameters . To calculate a standard error for the stimulated response , one is restricted to using only the data in the 10 min stimulation interval itself , in that the distribution of the test statistic is now known to be different outside the stimulation interval . A time series method was presented in the Methods to calculate the standard errors for the statistics over each of the 10 min stimulation intervals ( intervals separated from another by 3 hrs ) . One can combine these means and ( unequal ) standard errors across animals , to obtain overall means and standard errors . The availability of these standard errors allows one to make multiple comparison calculations , in order to assess differences in the responses with respect to amplitude ( μA ) versus frequency ( Hz ) and versus light or dark ( L/D ) . Because the standard errors differ significantly across amplitude and frequency , traditional methods such as Analysis of Variance or the nonparametric Kruskal-Wallis test , both of which require a constant variance , cannot be applied . There is a certain degree of robustness , but the differences here are significantly beyond that ( a factor 5–10 , at times ) ( Scheffe ( 1959 ) ) [24] . An alternative is still available . The statistical basis for our analysis is a multiple comparisons test , Dunnett’s C test , for which the variances are allowed to be unequal , as are the sample sizes ( Dunnett ( 1980 ) [25] . In Fig . 8A , the data , consolidated over Mice 1–4 , are displayed for four different statistics: 1D and 2D Mean Activities , Mean 2D Area and the RL Mean Difference of 2D Area . In Fig . 8B , a summary of the multiple comparisons ( at α = . 05 ) is presented for the the 2D and 1D Mean Activities . The results of the other two 2D statistics are similar to those of the 2D Mean Activity . In summary , for Hypothesis II , the mean activities due to the three amperages were statistically distinguishable ( and increasing with respect to amperage ) , whether one collapsed over L and D or compared within L or within D . From Hypothesis I , one knows that 75 μA could not be delineated from non-stimulation , and now it is shown , statistically , that 100 μA results in an increase in activity , and 125 μA in an even greater increase . As for comparisons of frequencies , they differed as to whether L and D were combined or not . A general statement ( at α = . 05 ) is that in dark , 50 Hz was statistically distinguishable from all three higher values ( 125 , 175 and 225 Hz ) . If one considers nonlinear interactions between amperage and frequency , there is a significant synergistic effect at the combination of 125 μA and 125 Hz , an important , practical conclusion . Again , all of the above results were based upon multiple comparison procedures . Hypothesis III is that stimulation initiates a pattern of movement that is more common to dark . Specifically , we show that 2-D movement bursts are more natural in the dark and that stimulation of sufficient strength in the light initiates a 2-D movement burst of the form that occurs in the non-stimulated dark state . The analysis for this hypothesis is based upon a comparison of the data , for all nine mice . To compare the 2-Dim and 1-Dim Mean Activities during light and during dark , test statistics were constructed ( in expression ( 7 ) ) ( comparisons of the means during dark with those during light ) and their standard errors were calculated , based upon estimated autocovariance functions ( see Methods ) . P-values were calculated based upon the test statistics . For two-dimensional movement , using nine mice ( six stimulated ( two , non-responsive ) and 3 non-stimulated controls ) , eight of the nine were statistically significant at . 05 for the Day-light comparisons . For one-dimensional movement , only one of the nine was significant ( at . 05 ) ( Fig . 9A ) . The strength of the evidence in such a multiple testing setting ( nine null hypotheses each for 1D and 2D ) was evaluated by a FDR thresholding at q = . 05 , with eight of the nine hypotheses being significant for 2D , and none of the nine hypotheses were significant for 1D ( Fig . 9B ) . The evidence is very strong for the day/light difference in two-dimensional movement . For one-dimensional movement , it is a matter of interpretation; at most , it would suggest a very weak circadian effect . Finally , in Fig . 9C , evidence is given that stimulation during light initiates movement representative of non-stimulated nocturnal movement ( which from Fig . 9A–B is 2-D movement ) . Plotted are the fractions of 2-D movement during non-stimulated light and non-stimulated dark ( for Mice 1–6 , Controls 1–3 ) , and the fractions during stimulated light and stimulated dark ( for Mice 1–4 ) . The plots reveal that similar changes occur during stimulated light ( Mice 1–4 ) and non-stimulated dark , as compared to non-stimulated light , providing additional evidence that stimulation in light initiates movement that is more naturally nocturnal . We have shown that Deep Brain Stimulation does initiate motor activity in response to stimulation , distinguishing it from what otherwise might have been just the coincidental occurrence of the continuously occurring stop-and-start movement of the mouse . We established that there is an increasing level of response to an increasing level of amperage , and increased response to the three higher frequencies . Significant synergism at the 125μA , 125 Hz combination was uncovered . The responses were shown to be those corresponding to 2-dim movement , not 1-dim movement . That 2-dim movement is much more common in dark than in light , was quantified . An important identification was that DBS stimulation in light initiates a level of 2-dim movement similar to the level of 2-dim movement in dark , generally ( i . e . , in non-stimulated dark ) . [We have also calculated from the incremental ( per sec ) changes in angle and the time-evolving winding number , and verified that stimulation initiates spatial movement and not mere spinning in place . ( not included in the Results ) ]
The statistically discoverable principles of the impact of DBS on Generalized Arousal ( GA ) behavior are not well understood , yet , at the same time , its use in the treatment of a diverse range of diseases is rapidly expanding . The present paper makes an important and vital contribution by identifying , with statistical criteria , the relationship of DBS parameters to induced motor activity ( which serves as a correlate for behavior ) . In the present work , we have established that DBS does stimulate movement and have quantified the degree of responsiveness to the stimulation parameters ( amperage and frequency ) . Stimulation at 100 μA produced a significant increase in activity above 75 μA ( shown to be equivalent to baseline ) and 125 μA a significant increase above 100 μA . A key conclusion was that there is a highly significant synergism at the combined 125μA and 125 Hz levels; 125 μA was the highest current , but 125 Hz was mid-level in the ( 50 , 125 , 175 , 225 Hz ) range . This could have a significant impact on the practical use of DBS as a treatment for a variety of diseases . A key concept in the present modeling was the identification of the importance of two-dimensional versus one-dimensional movement . It is 2-D movement , not 1-D , which responds to DBS stimulation . It is 2-D movement , not 1-D , which differs between light and dark . The proportion of 2-D movement in the dark is much greater than the proportion in light . The present experiments revealed that the proportion of 2-D movement initiated by stimulation in light during was similar to that of 2-D movement in non-stimulated dark . These factors , as whole , suggest that stimulation activates in light a manner of movement ( 2-D movement ) that is more commonly , nocturnal . Not only do the above conclusions have important practical consequences for brain arousal , but the methodology developed to draw such inferences , itself , has broad potential . The methodology is applicable to studies , broadly , for which the data consists of measurements of animal motor activity over time . One cannot apply traditional statistical methods to time-dependent processes , unless the time-varying structure is taken into account . Moreover , as a general statement , if one observes a time-dependent process , for which the time-dependency itself is changing ( e . g . , circadian rhythm ) , then , without additional knowledge , very little can be inferred about the underlying structure . If one observes a large number of animals ( e . g . , 15–20 ) , under identical conditions , one can potentially avoid the time-varying issue . However , if a large number of animals is not observed , then a much different approach is required . One must begin the modeling at the level of the individual animal , and build up from there . The key assumption ( justified in Methods ) is that of local stationarity , specifically , piecewise stationarity . In such a setting , methods such as shuffling or permutation tests are not valid and bootstrap methods are difficult to implement . If one uses the data in a manner that does not adhere to the time-dependency , one can often end up mistakingly inferring that there is an experimental effect , when in fact what was implicitly being tested ( and rejected ) was that the data was IID ( which it is not ) . In the present approach , a method was developed to compare the activity in the stimulation interval to that in neighboring intervals , in a manner consistent with the inherent local stationarity of mouse motor activity , which is highly influenced by light and dark . The developed methods , utilizing piecewise stationarity , allowed one to calculate statistics of motor activity over a segment of time , and , most importantly , to obtain accurate and justified standard errors for those statistics , again for an individual animal . The methods then allowed one to combine across individual animals , reaching the level of desired inference ( drawing conclusions based upon the full data ) . Because the variances at different times of day and/or different DBS parameters , are significantly different ( e . g . , at times , a factor of 5–10 ) , Analysis of Variance and nonparametric tests are not applicable , but other multiple comparison tests ( Dunnett’s C test ) are applicable . The conclusions of the present paper will aid in our understanding of the manner by which the CNS arousal pathways initiate various forms of behavior . In addition , the methodology developed for this work provides the experimentalist with justified methods for testing hypotheses in the common neuroscience framework in which animal motor activity is measured .
|
Brainstem and thalamic regulation of arousal has been studied experimentally since the mid 20-th century . Today , Deep Brain Stimulation ( DBS ) is used in the treatment of movement disorders , chronic pain , clinical depression , amongst others . At present , the proper choice of DBS parameters ( frequency and strength of the electric stimulation ) , and how those parameters should be modified as conditions change , are not well understood . In this work , using motor activity as the observed response , a statistical framework is developed for such study , and a quantitative relationship between parameter values and response is established . Within this framework , a possible correlate of arousal , the rapid onset of spatial ( two-dimensional ) movement , is uncovered and also studied . One long-term hope for techniques such as DBS are that they could assist in the treatment of disorders of consciousness , by supplementing or replacing ( e . g . , in Traumatic Brain Injury ) what should ordinarily be the appropriate endogenous stimulation .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Stochastic Modeling of Mouse Motor Activity under Deep Brain Stimulation: The Extraction of Arousal Information
|
A variety of adverse conditions including drought stress severely affect rice production . Root system plays a critical role in drought avoidance , which is one of the major mechanisms of drought resistance . In this study , we adopted genome-wide association study ( GWAS ) to dissect the genetic basis controlling various root traits by using a natural population consisting of 529 representative rice accessions . A total of 413 suggestive associations , containing 143 significant associations , were identified for 21 root traits , such as maximum root length , root volume , and root dry weight under normal and drought stress conditions at the maturation stage . More than 80 percent of the suggestive loci were located in the region of reported QTLs for root traits , while about 20 percent of suggestive loci were novel loci detected in this study . Besides , 11 reported root-related genes , including DRO1 , WOX11 , and OsPID , were found to co-locate with the association loci . We further proved that the association results can facilitate the efficient identification of causal genes for root traits by the two case studies of Nal1 and OsJAZ1 . These loci and their candidate causal genes provide an important basis for the genetic improvement of root traits and drought resistance .
Rice ( Oryza sativa L . ) is the staple food crop that feeds a large segment of the world’s population [1] . Owing to the climate change and shortage of freshwater , drought has been the most critical environmental stress influencing agriculture worldwide , particularly regarding the productivity of field crops [2] , especially for rice . To improve the drought resistance of rice and increase its yield under drought stress conditions is of great significance . Plant roots play an important role in the absorption and translocation of water and nutrients . The vibrant root system would allow crops to gain more water , and improvement of the root system architecture will contribute to drought avoidance in crops [3] . Some root traits are associated with plant productivity under drought stress conditions , such as fine root diameters , specific root length , and considerable root density [4] . In recent decades many researchers have tried to uncover the genetic basis of root traits in rice , aiming to improve the drought resistance of rice and to increase its yield under drought stress conditions . Traditionally , linkage mapping has been commonly employed to detect quantitative trait loci ( QTL ) for complex traits , including root architecture traits . In 1990s , linkage mapping of root traits in rice was conducted for the first time by using a population of 203 recombinant inbred lines ( RILs ) which were treated with drought stress at different stages , and five parameters of root morphology were investigated [5] . Since then , many QTLs related to root traits , such as root penetration ability , root growth rate under drought stress conditions , deep root morphology , and root thickness , were identified by linkage mapping in different populations [6–8] . A critical study on linkage mapping were conducted for drought resistance-related root traits , fitness , and productivity related traits in a population of 180 RILs , and 36 and 38 QTLs for different root traits were identified under normal and drought stress conditions , respectively [9] . In a subsequent study , a QTL controlling root volume in rice , qFSR4 , was fine mapped to a region of 38-kb on chromosome 4 , and Narrow leaf 1 ( Nal1 ) has been assumed to be the candidate gene for this QTL [10] . In another RIL population , 84 additive-effect QTLs and 86 pairs of epistatic QTLs were detected for six root traits at five different stages [11] . Meanwhile , several root related QTLs in rice have been verified in the past few years [12] . Sta1 , a QTL controlling the stele transversal area in root of rice , was fine mapped to a 359-kb interval on chromosome 9 [13] . ( DEEPER ROOTING 1 ) DRO1 [14] , DRO2 [15] , and DRO3 [16] , three major QTLs in the control of deep rooting of rice , were identified under normal conditions using a basket method; qSOR1 , a major QTL controlling the soil-surface rooting of rice in paddy fields , was mapped to an 812-kb interval on the long arm of chromosome 7 [17]; qRL6 . 1 [18] and qRL7 [19] , two major QTLs associated with the root length of rice , were identified in hydroponic conditions . Remarkably , the causal gene of the QTL DRO1 has been cloned , and DRO1 is involved in the regulation of deep rooting by affecting root growth angle [3] . The next-generation sequencing technology coupled with the growing number of genome sequences opens the opportunity to redesign genotyping strategies for more effective genetic mapping and genome analysis [20] . In recent years , genome-wide association study ( GWAS ) has been widely used as a powerful tool to reconnect a trait back to its underlying genetics [21] . Several association studies have been conducted for root traits in rice and other crops . The genetic architecture of aluminum tolerance in rice was analyzed through GWAS and bi-parental linkage mapping of the relative root growth of the total root system [22] . A GWAS for root related traits was carried out in a panel of 167 japonica rice accessions by using a hydroponic cultivation system at the seedling stage in rice [23] . Another GWAS involving the ratio of deep rooting by a modified ‘basket’ method in the field was performed in a population of 237 rice varieties , coupled with linkage mapping for the same trait in 180 recombinant inbred lines [24] . In maize , 268 marker-trait associations were detected in a GWAS for 22 seedling root architecture traits using 384 inbred lines [25] . In barley , 11 putative QTL for root related traits were found in a GWAS using a unique diversity set [26] . However , considering the complexity and plasticity of root development in crops , the genetic basis of root traits at the adult and seed maturation stages , especially for a population grown in soil under drought stress conditions , remains to be elucidated . In this study , the genetic architecture of root traits at the seed maturation stage under normal and drought stress conditions was investigated by GWAS using a panel of 529 rice accessions collected worldwide . We adopted a protocol for drought treatment by planting and stressing rice plants grown in individual polyvinyl chloride ( PVC ) tubes in which each genotype was stressed to the same extent at the same developmental stage [9] . The results showed that 225 of 264 loci identified by GWAS overlapped with reported root related QTLs and 11 reported root related genes were located in the corresponding region . In addition , two candidate genes , Nal1 and OsJAZ1 , identified by association analysis , were confirmed to control the corresponding root traits by genetic experiments . These results revealed a complete genetic control of the root system in rice at the reproductive stage under both normal and drought stress conditions , which could be an important basis for the genetic improvement of root traits and drought resistance .
In order to systematically dissect the genetic basis of root traits of rice grown in soil , a natural population containing 529 diverse rice accessions [27] were evaluated for various root traits under normal and drought stress conditions in PVC tubes . The population exhibited a distinctive population structure and was mainly classified into indica subpopulation ( 295 accessions ) including ind I , ind II , and indica intermediate ( ind ) , and japonica subpopulation ( 156 accessions ) including tej , trj , and japonica intermediate ( jap ) ( S1 Table ) . Drought stress was applied at the booting stage , and the plants were recovered when all of the leaves became fully rolled . A total of 21 root related traits , such as maximum root length under normal ( MRLN ) and drought stress conditions ( MRLD ) , volume of the deep roots ( > 30 cm ) under normal ( RVDN ) and drought stress conditions ( RVDD ) , and dry weight of the deep roots ( > 30 cm ) under normal ( RWDN ) and drought stress conditions ( RWDD ) , were measured at the seed maturation stage ( Table 1 ) . These traits were classified into four categories: root length , root volume , root weight , and deep root rate . For most of the traits , a large range of variation was detected , with the coefficients of variation ( CV ) varying from 0 . 236 for MRLN to 1 . 643 for RVDD . Most of the root traits showed a normal distribution ( S1 Fig ) while RVDD , RVDN , RWDN and RWDD showed skewed distributions , mainly due to the presence of accessions with very short roots ( < 30 cm ) . Correlation analysis among the 21 traits suggests that many traits were correlated ( S2 Fig ) , and the correlation coefficients between several trait pairs were very high , such as trait pairs of MRLN and DVRN , RVSN and RWSN , and RVSD and RWSD . Most of the traits under drought stress treatment were correlated with the same trait under normal conditions ( with a correlation coefficient r > 0 . 5 ) , such as the pair of MRLD and MRLN ( r = 0 . 63 ) , RVTD and RVTN ( r = 0 . 76 ) , and the pair of RWTD and RWTN ( r = 0 . 70 ) . In order to examine the effect of drought stress treatment on root growth without the interference of correlation , we compared the three deep root rate traits , including deep root length rate ( DLR ) , deep root volume rate ( DVR ) , and deep root weight rate ( DWR ) , under normal and drought stress conditions ( Table 2 ) . All the three traits were increased significantly in the whole population after drought stress treatment . The result is consistent with the conclusion that drought stress may promote the growth of deep roots [3 , 4] . However , the effect of drought stress on root growth varied in different subpopulations . For example , DWR was significantly increased after drought stress treatment in the japonica subpopulation but not in the indica subpopulation . While in different japonica subpopulations and different indica subpopulations [28] , the effect of drought stress treatment was complex . This result implies that the genetic basis of root traits in different subpopulations may vary significantly . In order to reduce the effect of population structure , a mixed linear model ( FaST-LMM ) [29] was used for the following association study . To identify root trait related association loci , we used a reported genotypic dataset consisting of about 6 . 4 million SNPs generated for the 529 accessions [27] to conduct GWAS for the 21 root related traits . Using a Bonferroni correction based on the effective numbers of independent markers [30] , the P-value thresholds were set at 1 . 21×10−6 ( suggestive ) and 6 . 03×10−8 ( significant ) . In our study , a total of 413 suggestive associations with 373 lead SNPs were identified , and among them 143 associations ( 133 significant SNPs ) exceeded the significant threshold ( S2 Table ) . In rice , an association locus has been defined as a chromosomal region in which the distance between the adjacent pairs of associated SNPs is less than 300 kb [27] . According to this definition , a total of 264 suggestive loci containing 373 suggestive SNPs were detected , and 110 of which were significant loci containing 133 significant SNPs ( S3 Table ) . A summary and comparison of the numbers of significant loci for the four categories of root traits under different growth conditions showed that the number of significant loci for root weight was more than that for the other three categories of root traits ( Fig 1 ) . In addition , more significant loci were detected under normal conditions than drought stress conditions in all types of root traits , except for root length . We further checked the loci detected by multiple traits . The result showed that , among the 264 suggestive association loci , a total of 73 loci were detected simultaneously by more than two different traits ( S4 Table and S3 Fig ) . In particular , a locus on chromosome 5 , GWAS NO . 104 , was detected by 8 different traits including RWTN , RVDD , RVTD , RWSN , RWDD , RWTD , RVDN , and RWDN . These results are in agreement with that many of the root traits were highly correlated ( S2 Fig ) . In the Tropgene DB ( rice data ) ( http://tropgenedb . cirad . fr/en/rice . html ) , a large number of QTLs for root traits ( deep root rate , root length , root volume , and root weight ) in rice have been deposited [31] . We picked up a total of 269 QTLs from this database to compare with the association loci in this study . As the most recent update of the database was in the year 2010 , another 36 QTLs [11 , 14–16 , 18 , 19 , 32 , 33] for the four categories of root traits reported in recent years were also included for comparison ( S5 Table ) . Considering the relevance among deep root rate , root length , root volume , and root weight , we analyzed the overlapping of the association loci and QTLs as a whole firstly , without distinguishing the four types of root traits . The results showed that out of the 264 suggestive loci , 225 loci ( 85 . 2% ) were located in the region of the QTLs ( Fig 2A and S5 Table ) . The large proportion of association loci overlapping with QTLs suggests that the GWAS is effective in the genetic dissection of root traits in rice . We further analyzed the overlapping of the association loci and QTLs for the four categories of root traits . The overlapping of the association loci and QTLs for root weight was greater than for the other three types of traits ( Fig 2B ) , with 104 of 166 association loci ( 62 . 7% ) for root weight locating in the regions of QTLs for root weight ( S4 Fig ) . The detailed information on the overlapping of the association loci and QTLs for root volume ( 22 . 3% of the association loci overlapping with the QTLs ) , root length ( 43 . 5% of the association loci overlapping with the QTLs ) , and deep root rate ( 24 . 3% of the association loci overlapping with the QTLs ) are presented in S5–S7 Figs , respectively . These results suggest that the significant association loci for root weight related traits may be more repeatable than the loci for the other three types of root traits in the comparison of the two mapping methods . Among all of the 264 suggestive loci , 11 reported root related genes were closely linked to the lead SNPs of each suggestive or significant loci for eight different traits ( RWDD , RWDN , RVDN , RVTN , DRVD , RWSN , RVDD , and RWTN ) ( Fig 3 and S6 Table ) . The details of the GWAS results ( Manhattan and quantile-quantile plots ) for the other 13 traits are presented in S8 Fig . The DEEPER ROOTING 1 ( DRO1 ) , a gene which increases the deep root ratio by controlling root growth angle [3] , was detected in the GWAS NO . 198 locus and linked with the trait for RWDD ( Fig 3A ) . This gene is located 14 kb upstream of the lead SNP sf0916321114 . The WUSCHEL-related homeobox gene WOX11 , a gene increasing drought resistance of rice by controlling root hair formation and root system development [34–36] , is located in the GWAS NO . 161 locus with the lead SNP sf0729023405 for the trait of RWDN ( Fig 3B ) . And within the same locus , OsAPY , another gene with a similar genetic effect of controlling root hair formation and root development [37] , is located very close to WOX11 on chromosome 7 ( Fig 3B ) . Also to be mentioned , a cytokinin-responsive gene RR2 , the common target of WOX11 and ERF3 [34] , was also associated with the trait of RWDN , in the GWAS NO . 43 locus ( Fig 3B ) . Another WUSCHEL-related homeobox gene WOX4 [38] , located in the GWAS NO . 98 locus , was also associated with RWDN ( Fig 3B ) . OsPID , the ortholog of PINOID in rice involved in the control of polar auxin transport and adventitious root development [39] , is located in the GWAS NO . 263 locus for the trait of RWDN ( Fig 3B ) . Dwarf and gladius leaf 1 ( DGL1 ) [40] , a gene involved in crown root development , microtubule organization , and gibberellin signaling , is located in the GWAS NO . 19 locus which was simultaneously detected for the traits RVDN , RVTN , and DRVD ( Fig 3C–3E ) . A casein kinase I protein OsCKI1 , regulating lateral root development in rice [41] , is located in the GWAS NO . 44 locus associated with RWSN ( Fig 3F ) . OsRHL1 , a gene controlling root hair formation and development [42] , is located in the GWAS NO . 120 locus associated with RVDD ( Fig 3G ) . Among the closely associated candidate genes , there were several reported genes with no evidence for their functions in the control of root traits . Two of them , Nal1 and OsJAZ1 , were selected for functional confirmation as case studies . Nal1 has been reported to control leaf width [43] , spikelet number [44] , photosynthesis rate [45] , and yield [46] . And it has also been identified by GWAS for leaf traits [47] and panicle number [48] . In this study , a region including Nal1 was detected by GWAS for RWDD ( dry weight of the deep roots under drought stress conditions ) , with the P-value of 3 . 12×10−7 ( Fig 3A ) , and Nal1 is located 78-kb downstream of the lead SNP sf0430940007 . Previously , we fine mapped a pleiotropic QTL qFSR4 to a region of 38 kb , which controls flag leaf width , spikelet number , and root volume at the reproductive stage in rice . In this region , Nal1 has been assumed to be the most promising candidate gene [10] . We sequenced the promoter ( about 2 kb before the ATG ) and CDS region of Nal1 in the two parent varieties Zhenshan 97B ( ZS97B ) and IRAT109 . Sequence comparison identified one SNP and ten Indels in the promoter region ( Fig 4A ) . In the CDS region , a 5985-bp retrotransposon insertion , located in the junction site of the first intron and the second exon of Nal1 in the genomes of Koshihikari and Nipponbare [45 , 49] , was also found in the genomes of ZS97B and IRAT109 ( Fig 4B ) . Except for this insertion , four SNPs were identified in the exons , leading to either synonymous mutations or changes of amino acids with similar biochemical properties , which were unlikely to affect the function of the Nal1 protein ( Fig 4B ) . However , the expression levels of Nal1 in the near isogenic line ( NIL ) with a background of IRAT109 ( Nal1IRAT109 , qIR ) were significantly greater than in the NIL with the background of ZS97B ( Nal1ZS97B , qZS ) in root , leaf , and panicle ( Fig 4C ) . Therefore , we assumed that the phenotypic difference might be caused mainly by the difference in the expression level of Nal1 . To confirm Nal1 as the causal gene for the pleiotropic QTL qFSR4 controlling the root traits , we performed a genetic complementation analysis . The 1 , 749-bp full-length CDS of Nal1 from the NIL of qIR , driven by the 2 , 059-bp promoter from the NIL of qIR , was transformed into the NIL of qZS . The positive complementary line COM4 with single copy of the transgene ( Fig 4D ) showed significantly greater root volume ( Fig 4F ) , wider flag leaves , and bigger spikelet number ( S9 Fig ) compared to the negative transgenic control ( NC ) . This result suggests that Nal1 is the causal gene of the pleiotropic QTL qFSR4 . To further confirm whether the variation in the expression level of Nal1 leads to the phenotypic difference , we constructed an overexpression vector and an RNA interference vector using the cDNA of Nal1 from the rice variety Nipponbare . Two independent overexpression lines and RNA interference lines of Nal1 were used for root traits investigation under normal and drought stress conditions . The expression levels of Nal1 in these lines under normal conditions at the seedling stage were presented in Fig 5A and 5F . Compared with the segregated negative transgenic control plants ( OE6-13 ( - ) ) , overexpression plants ( OE6-3 ( + ) ) exhibited a larger root size ( Fig 5B ) , with significantly increased root volume and root weight under normal growth conditions at the seed maturation stage ( Fig 5C ) , and the difference was also observed under drought stress conditions ( Fig 5D and 5E ) . To the contrary , RNA-interference plants ( Ri12-11 ( + ) ) exhibited significantly smaller root volume and root weight under both normal ( Fig 5G and 5H ) and drought stress ( Fig 5I and 5J ) conditions , compared to the segregated negative transgenic control plants ( Ri12-5 ( - ) ) . The same difference in root phenotypes was observed for another overexpression line ( OE3 ) and another RNA interference line ( Ri13 ) ( S10 Fig ) . These results further confirmed that Nal1 is the causal gene for the QTL qFSR4 controlling root volume . Since Nal1 was identified to control root volume , we examined the association of sequence variation in Nal1 with the root volume trait RVTN ( Fig 5K ) . In the 2 kb promoter region and the entire coding region , a total of 78 SNPs were extracted from RiceVarMap database [50] , and 28 SNPs were used for association analysis after excluding minor variants ( frequency < 0 . 05 ) ( S7 Table ) . Among them , six SNPs were associated with the RVTN , surpassing the Bonferroni-adjusted P-value ( 0 . 05/total markers = 1 . 78×10−3 ) in a mixed linear model . Except for three associated SNPs in the promoter , the other three SNPs were in the intron or 3' UTR region . Haplotype analysis of the six associated SNPs revealed two major haplotypes ( S7 Table ) , and Hap2 had a greater RVTN than Hap1 , with a P-value of 7 . 5×10−7 ( Fig 5L ) . Considering the difference in the expression level of Nal1 observed in the qIR and qZS NILs , this result implies that sequence variations in the promoter region of Nal1 may mainly contribute to the phenotypic difference in RVTN . A locus with a lead SNP sf0433137256 on chromosome 4 was identified for RWSN ( dry weight of the shallow roots under normal conditions ) with a P-value of 6 . 4×10−9 ( Fig 3F ) and for RWTN ( dry weight of the total roots under normal conditions ) with a P-value of 1 . 58×10-8 ( Fig 3H ) . A reported gene OsJAZ1 ( EG2 ) , which has a role in the regulation of spikelet development in rice [51] , is located only 14 kb upstream of the lead SNP . This gene caught our attention because it has a strong expression level in root [52] , and it is one of the few genes with an obvious root-enriched expression pattern in the region around the lead SNP according to data from RiceXPro database ( http://ricexpro . dna . affrc . go . jp/ ) [53] ( S11 Fig ) . Association analysis between OsJAZ1 and RWTN was conducted to detect the causal variant . A total of 143 SNPs in the 2kb promoter region and the entire coding region of OsJAZ1 from RiceVarMap database were used for study after excluding minor variants ( frequency < 0 . 05 ) ( S8 Table ) . It was shown that only one SNP ( P = 1 . 94×10−5 ) , located in the promoter region , surpassed the Bonferroni-adjusted P-value ( 0 . 05/total markers = 3 . 49×10−4 ) in a mixed linear model , and it was regarded as a suggestive association with the RWTN ( Fig 6A ) . Meanwhile , seven SNPs in the promoter region and intron region were identified as marginally suggestive . Further haplotype analysis of the above eight associated SNPs revealed two major haplotypes ( S8 Table ) , and Hap2 had significantly greater RWTN value than Hap1 ( P = 1 . 1×10−3 ) ( Fig 6B ) . To examine whether OsJAZ1 is functionally related to root traits , we checked the phenotype of OsJAZ1-overexpression ( OsJAZ1-OE ) plants . The OsJAZ1-OE plants exhibited significantly longer roots and more crown roots than the wild-type Zhonghua11 ( ZH11 ) at the seedling stage ( S12A and S12B Fig ) . The relative expression level of OsJAZ1 in the leaves of OsJAZ1-OE plants was about ten times greater than in ZH11 , while the difference in the roots was about thirty times ( Fig 6C ) . The OsJAZ1-OE plants also showed significantly larger root size at the tillering stage ( Fig 6D ) and the seed maturation stage ( Fig 6E ) . The average root dry weight of the OsJAZ1-OE plants was significantly greater than that of ZH11 at these two stages ( Fig 6F ) , which is consistent with the association result that the locus containing OsJAZ1 was detected for the trait of root weight . These results suggest that OsJAZ1 may regulate root development at various developmental stages of rice .
Plant roots are of great significance in drought avoidance . However , phenotyping of the root system in soil is still a big challenge , especially for plants under drought stress conditions . DRO1 [14] , DRO2 [15] , and DRO3 [16] , three major QTLs involved in the deep rooting of rice under normal conditions , have been fine mapped by planting rice in hemispherical baskets placed in hydroponic solution . The hemispherical basket method was adopted in the GWAS for deep root ratio analysis in the field [24] . Association mapping of root traits in a japonica rice panel has also been conducted using a hydroponic root phenotyping system with glass beads to support the plants [23] , but root phenotyping under drought stress conditions was seldom addressed . Although the hydroponic method can overcome the problem of root invisibility which exists with the soil method , natural root architecture can hardly be revealed [54] . In this study , root traits of rice under normal and drought conditions were measured using the PVC tube method [9] . The PVC tube , 1 m in height and 20 cm in diameter , was designed specifically for measuring the root traits of rice under normal or drought stress conditions . Using this method , we could measure several important root traits that are closely related to drought avoidance in the field , such as the maximal root length , root dry weight , root volume , and deep root rate . Despite the labor-intensive root washing , root phenotypes of plants grown in soil-filled large pots are more similar to the actual root system in the field than the hydroponic method . Nevertheless , the PVC tube method also has its disadvantages in root phenotyping . First , it is hard to monitor the dynamic growth of root traits unless a more powerful detection system is developed . Second , sandy soil should be carefully washed away from the roots before measuring , which inevitably impairs the natural root architecture . Therefore , a non-destructive root phenotyping system is eagerly expected . Nowadays , several research groups have successfully established different kinds of non-destructive root phenotyping systems such as RooTrak system [54] . However , a high throughput root phenotyping technology for plants grown in natural soil conditions remains a big challenge . Some root traits showed differences under normal conditions and drought stress conditions . Due to the effect of population structure , the change tendency of deep root rate traits after drought stress treatment varied not only between the indica and the japonica subpopulations but also in the different indica subpopulations ( ind , ind I , and ind II ) and the different japonica subpopulations ( jap , tej , and trj ) ( Table 2 ) . Comparison of the association loci under normal and drought stress conditions suggests that some loci can be detected under both normal and drought stress conditions . The genetic control for some traits under normal and drought stress conditions is partially overlapped . For example , for RVTN and RVTD , two very close associations were detected on chromosome 4 ( Figs 3D and S8I ) , and these two associations were regarded as the same association locus ( GWAS NO . 78 ) . Some loci were detected only under normal conditions or drought stress conditions . These results suggest that the genetic basis of root traits under normal and drought stress conditions is largely different for some traits , such as the maximal root length , root volume , and deep root weight . For an example , several obvious peaks were detected for deep root weight under normal conditions ( RWDN ) ( Fig 3B ) , but few signals were detected for the trait under drought stress conditions ( RWDD ) ( Fig 3A ) . It is noteworthy that a significant locus ( GWAS NO . 198 ) was detected on chromosome 9 with a lead SNP ( sf0916321114 ) . However , this locus was not detected under normal conditions for RWDN . Interestingly , DRO1 is located at 14 kb upstream of the lead SNP . DRO1 has been reported to be involved in cell elongation of the root tip in response to gravity , and greater expression of DRO1 could increase deep rooting by increasing the root growth angle , thus improving drought avoidance and maintaining yield under drought stress conditions . The single 1-bp deletion in the fourth exon of DRO1 in IR64 results in the introduction of a premature stop codon [3] . In the 529 accessions used in this study , however , only two accessions have the 1-bp deletion ( S9 Table ) . Therefore , this 1-bp deletion is unlikely to be the causal variant for the root traits in this panel . Compared to many dicot species such as Arabidopsis , which has a primary root iteratively branching to generate several orders of lateral roots , the root system of cereal crops such as rice is more complex [55] . The complexity of the genetic control of root architecture is as complex as the trait itself [4 , 9] . In this study , 73 loci were detected for two and more traits ( S4 Table ) , but 191 loci were detected only for one trait , which may suggest that some loci have pleiotropic effects on root traits while most of the root traits are controlled by distinct genetic loci . On the other hand , considering the relatively low LD decay in rice , one association locus in this study was defined as a 200 kb region containing more than ten genes , thus it is rather difficult to pinpoint the causal genes for these loci . However , the combination of QTL information , expression profile , and prediction of gene function could help to narrow down the candidate genes , just like the two case studies we presented . The pleiotropic QTL qFSR4 affects flag leaf width , spikelet number , and root volume in rice , and it has been fine mapped to a 38 kb region in which Nal1 was assumed to be the most likely candidate gene [10] . Nal1 has been reported to be involved in auxin polar transport [43] and cell division [56] , and its regulation in the development of leaves and adventitious roots may be via modulating the expression of the PIN and CRL genes [57] . In this study , the association analysis revealed a locus for RWDD , which overlapped with the QTL qFSR4 . We confirmed the function of Nal1 in the control of root traits by sequence and expression level comparison , genetic complementation , and haplotype analysis . Overexpression and RNA-interference lines of Nal1 resulted in significant changes in root volume and root weight under both normal and drought stress conditions compared to the control plants , demonstrating the authentic function of Nal1 in the control of root traits . The Nal1 allele from the japonica rice cultivar showed better performance in regulating adventitious root development at the seedling stage [57] , which is consistent with our result that the japonica allele of Nal1 is superior when compared to the indica allele with respect to the control of root volume at the adult stage . Another case study for OsJAZ1 , a candidate causal gene for the locus detected by RWSN and RWTN on chromosome 4 , suggests that the association results could provide opportunities to identify novel genes or known genes with new functions in the control of root traits . OsJAZ1 has been reported to be involved in spikelet development in rice [51] . However , no JAZ proteins have been reported to regulate root traits under normal or drought stress conditions in rice , although some JAZ proteins in Arabidopsis have been shown to regulate root development under JA treatment [58] . Here , we found OsJAZ1 was involved in root development with genetic evidence of OsJAZ1-overexpression lines at the seedling , tillering and reproductive stages . It would be very interesting to further reveal how OsJAZ1 regulates both the underground organ ( root ) and the reproductive organ ( spikelet ) . Besides the tens of reported root-related QTLs or genes , many unreported loci or genes for root traits were detected in our association analysis . With the help of expression profiling databases and bioinformatic analysis , we could narrow down the potential candidate causal genes for root traits in the significant association loci . For example , 28 possible candidate genes with obvious drought-responsive and/or root-specific expression patterns were identified for some of the significant association loci ( S10 Table ) . Nevertheless , subsequent genetic experiments are necessary to confirm the functions of these genes in root development or drought avoidance . In conclusion , our study provides a relatively comprehensive analysis of the genetic architecture of root traits in rice . We employed a GWAS with 529 rice accessions for root traits at the seed maturation stage under normal and drought stress conditions , and 225 of 264 loci identified by GWAS overlapped with reported root related QTLs . Many known root-related genes were located in the significant association loci . Importantly , case studies of two genes , Nal1 and OsJAZ1 , demonstrate the feasibility of mining for candidate genes by GWAS . The association loci and causal genes identified in this study provide an important foundation for revealing the molecular mechanism of root development and genetic improvement of rice root or drought resistance in the future .
A total of 529 rice accessions including 202 from the China Core Collection and 327 from the World Core Collection were used for the association analysis ( S1 Table ) . This panel of rice accessions is essentially the same as the panel of 533 accessions as previously described [29] , except three accessions ( C126 , W196 , and W232 ) with severe heterozygosity and one ( W190 ) with a low mapping rate ( 10% ) omitted . For GWAS of root traits , 529 rice accessions were grown in PVC tubes ( 1 m in height , 20 cm in diameter ) described by Yue et al [9] , with one plant per tube and six plants per accession ( three for normal conditions and three for drought stress conditions ) . The PVC tubes were arranged in the field facilitated with a moveable shelter at the experimental station of Huazhong Agricultural University ( 114 . 33°E , 30 . 35°N ) . The average air temperature was 30 . 3°C and the average relative air humidity was 67 . 5% during the rice growth period . The arrangement of the PVC tubes and rice accessions are shown in S13A and S13B Fig . Considering that the heading date varied in the whole population , the 529 accessions were sown in several batches each with a relatively close heading date . At the beginning of the tillering stage , 1 g of urea ( dissolved in water ) was applied to each tube . The plants were fully irrigated every day until the drought stress treatment was applied . At the booting stage , drought stress was applied to three of the blocks with the other three blocks used as a control . To apply drought stress treatment , water was added to the full capacity of the tubes , and the plugs on the tubes were removed , allowing the slow drainage of water in the tubes through the small holes on the tubes . Rain was kept off by closing the shelter . When all the leaves of a stressed plant became fully rolled , watering was applied to the full capacity of the tubes . When the full water capacity maintained for one day , the second cycle of drought stress was applied to the plant until all of the leaves became fully rolled again . After the second round of drought stress treatment , watering was resumed for the rest of the life cycle . A total of 21 root related traits were phenotyped in this study . The root traits were measured at the seed maturation stage of the plants . To measure these traits , the plastic bag containing the soil and roots was pulled out from the PVC tube and laid out on a soil-washing table with a 2-mm sieve . After removing the plastic bag , the soil was washed away carefully and the length of the longest root was scored as the maximum root length ( in centimeters ) . The washed roots which were free from the soil were shown in S13C Fig . Then the roots were cut into two parts at 30 cm from the basal node of the plant . The volume ( in milliliters ) of the roots from the two parts was measured in a cylinder using the water-replacement method . The dry weight ( in grams ) of the roots from the two parts was measured with a balance after air-drying the roots . The root mass above 30 cm was designated as shallow roots while the root mass below 30 cm was designated as deep roots , from which a number of indices were derived . The flow of measurement is shown in S13D Fig . The abbreviations and descriptions of the corresponding traits are listed in Table 1 . A total of 529 accessions including nine subpopulations were collected to construct this association panel . For the 21 traits used for GWAS , we adopted a mixed-model approach using the factored spectrally transformed linear mixed models ( FaST-LMM ) program , with 4 , 358 , 600 SNP across the entire rice genome ( minor allele frequency ≥ 0 . 05; the number of accessions with minor alleles ≥ 6 ) . The suggestive and significant P-value thresholds of the entire population were respectively 1 . 21×10−06 and 6 . 03×10−08 . The linkage disequilibrium ( LD ) statistic r2 was calculated by Plink based on haplotype frequencies . More detailed information about our association analysis was referenced in the recent study [29] . Candidate association analysis of Nal1 and OsJAZ1 was performed with TASSEL version 5 [59] . LD plots were generated with Haploview4 . 2 , and LD was indicated using r2 values between the pairs of SNPs multiplied by 100 ( white , r2 = 0; shades of gray , 0< r2 <1; black , r2 = 1 ) . The physical region of root-related QTLs was determined by the physical position of the left border marker and the right border marker , which were obtained by searching markers information in the GRAMENE database ( http://www . gramene . org/ ) . The physical region of each association locus was defined as 200 kb around each lead SNP . An overlapping locus was claimed if the physical region of the association locus is overlapped with the physical region of any reported QTL for root traits . The genotypes of Nal1 and OsJAZ1 in the 529 rice samples were obtained from the RiceVarMap database ( http://ricevarmap . ncpgr . cn/ ) . The haplotypes were classified based on all of the SNPs with an MAF > 0 . 05 in a candidate gene . The haplotypes containing at least ten rice accessions were used for comparative analysis . One-way ANOVA and Student’s t-test was employed to compare the differences in root traits among the haplotypes [60] . The QTL qFSR4 was narrowed down to a region of approximately 38-kb flanked by markers FSR-75 and FSR-78 [10] , and the progeny of a recombinant plant WHD10-74 were genotyped with the marker RM17483 , and the plant with homozygous genotype as ZS97B was used as NIL Nal1ZS97B ( qZS ) while the plant with homozygous genotype as IRAT109 was used as NIL Nal1IRAT109 ( qIR ) in our study . For the complementation test of Nal1 , a 2 , 059-bp Nal1 promoter fragment was amplified from the NIL qIR with the addition of restriction sites for KpnI and EcoRI , and the 1 , 749-bp Nal1 full CDS fragment was amplified from the cDNA of the qIR line with the addition of a restriction site for KpnI at both ends . The two fragments were cloned into the binary vector pCAMBIA2301 to generate the transformation plasmid for the complementation test . The resulting transformation plasmid was introduced into NIL qZS by Agrobacterium-mediated transformation [61] . The copy numbers of marker gene ( G418 ) for transformation were determined by Southern blot . The genotype was detected with PCR using the primer pair based on the sequence difference in the promoter region of ZS97B and IRAT109 . For the Nal1 overexpression transgenic plants , the full-length cDNA of Nal1 was amplified from the japonica cv . Nipponbare . The sequence-confirmed PCR fragment was ligated into pCAMBIA1301U which was digested with KpnI , based on the Gibson assembly principle [62] . The construct was introduced into the japonica cv . Zhonghua11 by Agrobacterium-mediated transformation [61] . For the OsJAZ1 overexpression transgenic plants , the full-length cDNA of OsJAZ1 was cut from a clone from the full length cDNA library of the indica cv . Minghui63 , and cloned into the pCAMBIA1301H vector driven by the OsLEA3-1 promoter LEAP [63] . The construct was introduced into the japonica cv . Zhonghua11 by Agrobacterium-mediated transformation [61] . For the Nal1 interference transgenic plants , the 454-bp length cDNA of Nal1 was amplified from the japonica cv . Nipponbare by RT-PCR . The sequence-confirmed PCR fragment was recombined into the pANDA vector [64] , by the gateway system . The construct was introduced into japonica cv . Zhonghua11 by Agrobacterium-mediated transformation [61] . To investigate the transcript levels of Nal1 in different tissues at different stages , seeds of the NILs qZS and qIR were germinated on normal 1/2 strength MS medium in Petri dishes . Before germination , rice seeds were surface-sterilized in 75% ethanol for 5 min , followed by a 10 min incubation with 0 . 15% HgCl2 , and then washed four to five times with sterile water . After germination , the seedlings were transplanted to 10 cm high sterile plastic box with a 5 cm deep normal 1/2 strength MS medium in it . The roots and leaves of qZS and qIR seedlings were sampled at the two-leaf stage , then the qZS and qIR seedlings were transplanted in the field . Flag leaves , the second leaves , and the third leaves of qZS and qIR were sampled at the heading stage . Panicles of qZS and qIR were sampled before heading and after heading . To investigate the transcript levels of OsJAZ1 and the root phenotype of OsJAZ1 overexpression and wild-type plants Zhonghua11 ( ZH11 ) , the positive transgenic plants were selected by germinating on 1/2 strength MS medium containing 25 mg/L hygromycin B ( Roche ) . The wild-type seeds were germinated on normal 1/2 strength MS medium . Before germination , the rice seeds were surface-sterilized in 75% ethanol for 5 min , followed by a 10 min incubation with 0 . 15% HgCl2 , and then washed four to five times with sterile water . After germination , the overexpression seedlings were transplanted to 10 cm × 10 cm square plastic Petri dishes with a 0 . 5 cm deep normal 1/2 strength MS medium in it ( 14 seedlings each , three repeats ) , with control plants in the same dish ( half of each ) . Root length , seedling length , and crown root number of the overexpression and wild-type plants were investigated at the two-leaf stage . The roots and leaves of overexpression and wild-type plants were sampled after phenotype investigation . The seedlings were transplanted to pots with sandy soil in it until the tillering stage , at which point the seedlings were carefully washed out from the sandy soil and the roots were cut for root weight measurement after air drying . Then , the seedlings were transplanted to PVC tubes ( each with one overexpression plant and one wild-type plant ) , and the root phenotype of the overexpression and wild-type plants were investigated at the seed maturation stage using the same method as mentioned above . To investigate the phenotype of the NILs of qFSR4 and complementary plants , the seeds of qZS , qIR , COM4 , and NC were sown in the nursery field and the seedlings were transplanted to PVC tubes at the four-leaf stage . Each tube contained two plants , one qZS plant and one qIR plant , or one COM4 plant and one NC plant . Each group was repeated ten times ( tubes ) for normal growth and drought stress treatment , respectively . Leaf width was measured for three flag leaves of each plant at the heading stage , and the panicles were harvested for counting spikelet and seed numbers after maturation . The root traits were measured by carefully washing soil away using the method described above . To investigate the root phenotype of the Nal1 overexpression and RNA-interference plants , the seeds of transgenic positive plants ( OE3-16 ( + ) , OE6-3 ( + ) , Ri12-11 ( + ) , and Ri13-6 ( + ) ) and negative controls ( OE3-19 ( - ) , OE6-13 ( - ) , Ri12-5 ( - ) , and Ri13-15 ( - ) ) were sown in the nursery field , leaves of each line were sampled at the four-leaf stage for investigating the transcript levels of Nal1 . Then the seedlings were transplanted to PVC tubes . Each tube contained two plants ( overexpression or RNAi plant and corresponding control ) . Each group was repeated ten times ( tubes ) for normal growth and drought stress treatment , respectively . Drought stress treatment was applied at the booting stage as described above . The traits of root volume and root weight were investigated as described above . Total RNA was extracted using Trizol reagent ( Invitrogen ) . The first-strand cDNA was reverse transcribed using M-MLV reverse transcriptase ( Invitrogen ) according to the manufacturer’s instructions . Quantitative PCR was conducted on a 7500 Real-Time PCR System ( Applied Biosystems ) using SYBR Premix ExTaq ( TaKaRa ) according to the manufacturer’s instructions . The rice Ubiquitin gene was used as the internal control . The relative expression level was determined as reported previously [65] .
|
Drought stress is a key environmental factor that severely reduces crop yield all over the world . The root system plays a critical role in the drought avoidance of crops , but the genetic basis of the root system in soil conditions has seldom been investigated in rice . We analyzed the genetic control of 21 root traits via genome-wide association study ( GWAS ) of a natural population , and we identified 110 significant association loci containing many reported and unknown candidate genes related to root development . We presented the case studies of two genes , Nal1 and OsJAZ1 , to demonstrate the high efficiency of the identification of genes in the control of root traits based on the association and candidate gene analyses . Our results would provide a foundation not only for elucidating the genetic and molecular basis of root development but also for improving the drought avoidance of rice .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"biotechnology",
"genome-wide",
"association",
"studies",
"ecology",
"and",
"environmental",
"sciences",
"quantitative",
"trait",
"loci",
"plant",
"physiology",
"plant",
"science",
"rice",
"genetically",
"modified",
"plants",
"materials",
"science",
"experimental",
"organism",
"systems",
"genome",
"analysis",
"trait",
"locus",
"analysis",
"plant",
"pathology",
"plant",
"ecology",
"plants",
"macromolecules",
"genetic",
"engineering",
"materials",
"by",
"structure",
"research",
"and",
"analysis",
"methods",
"polymers",
"polymer",
"chemistry",
"grasses",
"genomics",
"genetically",
"modified",
"organisms",
"chemistry",
"genetic",
"loci",
"plant",
"defenses",
"plant",
"and",
"algal",
"models",
"plant",
"resistance",
"to",
"abiotic",
"stress",
"polyvinyl",
"chloride",
"ecology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"plant",
"biotechnology",
"plant-environment",
"interactions",
"organisms",
"human",
"genetics"
] |
2017
|
Genetic control of the root system in rice under normal and drought stress conditions by genome-wide association study
|
Leaf angle is an important agronomic trait and influences crop architecture and yield . Studies have demonstrated the roles of phytohormones , particularly auxin and brassinosteroids , and various factors in controlling leaf inclination . However , the underlying mechanism especially the upstream regulatory networks still need being clarified . Here we report the functional characterization of rice leaf inclination3 ( LC3 ) , a SPOC domain-containing transcription suppressor , in regulating leaf inclination through interacting with LIP1 ( LC3-interacting protein 1 ) , a HIT zinc finger domain-containing protein . LC3 deficiency results in increased leaf inclination and enhanced expressions of OsIAA12 and OsGH3 . 2 . Being consistent , transgenic plants with OsIAA12 overexpression or deficiency of OsARF17 which interacts with OsIAA12 do present enlarged leaf inclination . LIP1 directly binds to promoter regions of OsIAA12 and OsGH3 . 2 , and interacts with LC3 to synergistically suppress auxin signaling . Our study demonstrate the distinct effects of IAA12-ARF17 interactions in leaf inclination regulation , and provide informative clues to elucidate the functional mechanism of SPOC domain-containing transcription suppressor and fine-controlled network of lamina joint development by LC3-regulated auxin homeostasis and auxin signaling through .
Rice is one of the most important crops in the world and breeding rice varieties with ideal architecture is a vital strategy for improvement of grain yields [1 , 2] . Leaf is the main organ for photosynthesis and its development is crucial for the yield potential . Leaf inclination indicates the angle between leaf blade and culm [3] , and studies have shown that erect leaf facilitates the penetration of sunlight and enhances the photosynthetic efficiency [4 , 5] , which is suitable for dense planting . Unbalanced development of collar cells at adaxial or abaxial sides , development of mechanical tissue and mechanical strength , formation of vascular bundle and cell wall compositions also have been pointed out to affect the leaf angle [2 , 3 , 6 , 7] . Recent systemic analysis of the dynamic developmental processes of lamina joint through cytological observation showed that cell differentiation , division and elongation , cell wall thickening , and programmed cell death ( PCD ) , are closely correlated with leaf angle and regulated by a complex network , consisting of various factors , especially protein kinases and hormones [8] . Indeed , studies by using mutants or transgenic approaches indicate that altered biosynthesis or signaling of brassinosteriods ( BRs ) lead to the change of leaf inclination , such as BR-deficient mutant dwarf4-1 [9] , ebisudwarf ( d2 ) [10] , dwarf1 ( brd1 ) [5] , BR signaling mutant d61-1 , 2 ( weak mutant alleles of OsBRI1 ) [11] , or rice plants with reduced expression of OsBZR1 [12] . Similarly , plants with suppressed auxin signaling by overexpressing miR393a/b that suppress expression of receptor OsTIR1 [13] , or with reduced auxin levels including mutant lc1 [6] or plants overexpressing GH3 family members OsGH3 . 2 , OsGH3 . 5 , and OsGH3 . 13 [7 , 14 , 15] , present increased leaf inclination . It is noticed that BR stimulates while auxin suppresses the leaf inclination through regulating the cell division or elongation at adaxial side of lamina joint and auxin coordinates with BR to control the lamina joint development [6 , 7 , 16] . In addition , ethylene may participate in BR-induced leaf inclination [17] and repressed expression of a gibberellin signaling negative regulator , SPINDLY , leads to increased leaf angles [18] . Many transcription factors ( TFs ) are involved in leaf inclination regulation . OsWRKY1 and MADS-box proteins OsMADS22 , OsMADS55 and OsMADS47 negatively regulate leaf inclination [19–21] . Ectopic expression of LAX PANICLE ( LAX ) , a basic helix-loop-helix TF , leads to increased bending of lamina joint [22] . Deficiency of OsLIGULELESS1 ( OsLG1 , a SBP domain-containing TF ) results in defects in auricle , ligule , and lamina joint [23] . Recently , a genome-wide association study shows that rice TFs OsbHLH153 , OsbHLH173 and OsbHLH174 involve in flag leaf angle regulation [24] . In addition , the Aux/IAA family members interact with AUXIN RESPONSE FACTOR ( ARFs ) to suppress auxin signaling [25] and regulate the leaf blades [26] . Overexpression of OsIAA1 , OsIAA4 , OsARF19 lead to the increased leaf angle [7 , 16 , 27] , while deficiency of OsARF11 , an ortholog of Arabidopsis ARF5 , results in reduced leaf angle [28] . However , the detailed mechanism , especially how Aux/IAA is regulated during lamina joint development and which distinct Aux/IAA-ARF interaction regulates leaf inclination is unknown yet . By systemic analysis of a rice mutant with enlarged leaf angle , we showed that leaf inclination3 ( LC3 ) , a SPOC domain-containing protein that is speculated to facilitate protein-protein interactions in transcription repression complex [29] , interacts with a HIT zinc finger domain-containing TF LIP1 ( LC3-interacting protein 1 ) to suppress the auxin signaling and homeostasis genes , hence to regulate the cell elongation at adaxial side of lamina joint and thus leaf inclination . These results provide informative clues on the fine-controlled network regulating lamina joint development .
Our previous studies by analyzing the global transcriptome of developing lamia joint showed that gene leaf inclination3 ( LC3 , Os06g39480 ) is down-regulated from stage 2 to stage 6 during lamina joint development . Further analysis of the corresponding knockout mutant , lc3 , revealed the obviously increased leaf angles under LC3 deficiency [8] . LC3 encodes a novel SPOC domain-containing protein and the underlying functional mechanism is thus detailed studied . Analysis of the transcription pattern of LC3 by quantitative real-time RT-PCR ( qRT-PCR ) confirms the reduced expression of LC3 along with lamina joint development , while LC3 is relatively highly expressed in pistil , spikelet and seeds at early stage ( Fig 1A ) . Further promoter-reporter gene fusion analysis ( a 3-kb promoter regions of LC3 was fused to the β-glucuronidase gene ) consistently show that LC3 is transcribed at adaxial side of lamina joint , glume and pollen , pistil and seeds ( Fig 1B ) . Based on the significantly decreased LC3 expression in lc3 [8] , transgenic lc3 plants with complemented expression of LC3 , driven by its native promoter , was generated ( Fig 1C , left panel ) . Phenotypic observation and measurement show the restored leaf inclination ( Fig 1C ) , which confirms the role of LC3 in regulating leaf inclination . To clarify the cytological change of lc3 mutants , lamina joint paraffin section was conducted . Observations of the longitudinal sections show the increased cell width , while unaltered cell number and cell length of the second layer parenchyma cells at adaxial side of lc3 lamina joint ( Fig 1D ) . Further observations of transverse sections consistently show the increased cell length and unaltered cell layer numbers at adaxial side ( Fig 1E ) , and no change of cell length and cell layers at abaxial side of lc3 mutant lamina joint ( S1 and S2 Figs ) . Overall , the excessive cell elongation at adaxial side of lamina joint results in the enlarged leaf angle of lc3 . Auxin and brassinosteroids play crucial roles in regulating lamina joint development and thus leaf inclination . To investigate the functional mechanism of LC3 , expression level of auxin and brassinosteroids signaling related genes and some reported genes regulating leaf angles were examined . qRT-PCR analysis reveals the decreased level of auxin signaling related genes ARF2 , IAA6 , IAA9 , unaltered expressions of BR-related genes , increased expressions of LAZY1 [30] and TAC1 [31] , and interestingly , dramatically increased levels of OsIAA12 and OsGH3 . 2 in lc3 mutant ( Fig 2A ) . Previous studies showed that overexpression of OsGH3 . 2 did result in the increased leaf angles [14] , similar to LC1 ( OsGH3 . 1 ) overexpressing plants [6] . We thus focus on the effect of OsIAA12 and relevant regulatory mechanism . Transgenic rice plants overexpressing OsIAA12 driven by a maize ubiquitin promoter were generated and analysis of the positive lines ( Fig 2B , left panel ) at 10 days after heading showed that OsIAA12 overexpression indeed leads to the increased leaf inclination ( Fig 2B ) . Analysis of the longitudinal sections of lamina joint reveals the increased cell width of the second layer parenchyma cells at adaxial side of OsIAA12-overexpressing lines ( Fig 2C ) , which is same to that of lc3 . Consistently , observations of the transverse sections of flag leaf lamina joint show that though there is no change in cell layer numbers in adaxial or abaxial regions ( S3 and S4 Figs ) , increased length of adaxial cells was detected ( Fig 2D ) . These results indicate that LC3 regulates lamina joint development possibly through OsIAA12 and auxin signaling . AUX/IAA proteins interact with ARFs to suppress the auxin signaling , and which ARF cooperates with OsIAA12 to regulate the leaf inclination distinctly is studied . Previous studies on the IAAs-ARFs interacting networks indicated the interaction between OsIAA12 and OsARF17 [32] , which was confirmed by the yeast two-hybrid assays ( Fig 3A ) . Further analysis by Split-YFP assay through expressing N-terminal YFP fused OsIAA12 ( nYFP-OsIAA12 ) and C-terminal YFP fused OsARF17 ( OsARF17-cYFP ) in tobacco leaf epidermal confirmed the OsIAA12-OsARF17 interaction in nucleus in planta ( Fig 3B ) . Being consistent , transient expression of OsIAA12-RFP and OsARF17-GFP fusion proteins in rice protoplasts showed that OsIAA12 co-localizes with OsARF17 in nucleus ( Fig 3C ) . To confirm the role of OsIAA12-OsARF17 interaction in leaf inclination regulation , plants deficiency of OsARF17 were generated by Crispr/Cas9 approach ( OsARF17-Cas9 ) . Six independent transgenic lines were obtained and four of them were homozygous with either insertion or deletions at 5’ end of OsARF17 ( Fig 3D , upper panel ) . Phenotypic observations and measurement of T2 generations showed obviously enhanced flag leaf angles ( Fig 3D , bottom panel ) . Analysis of paraffin section revealed similar cytological change as lc3 mutant and transgenic plants overexpressing OsIAA12 ( Fig 2C and 2D; S3 and S4 Figs ) , suggesting that LC3 regulates cell elongation at adaxial side and leaf inclination through suppressing the OsIAA12 expression , which regulates OsARF17 effects by protein interaction . Examination of the transcriptions of OsIAA12 and OsGH3 . 2 showed the suppressed expression of OsIAA12 and OsGH3 . 2 under LC3 overexpression and restored expression in lc3 plants with complemented expression of LC3 ( Fig 3E ) , further confirming the regulation of OsIAA12 and OsGH3 . 2 by LC3 . In addition , expression of OsARF17 is unaltered under OsIAA12 or LC3 overexpression ( S5 Fig ) , which is consistent with that Aux/IAA proteins repress ARF transcription factors via direct protein-protein interaction . LC3 encodes a SPOC-domain containing protein and localizes widely in cells ( mainly in nucleus , Fig 4A ) . Previous reports showed that Spilt ends ( Spen ) protein family members compose an N-terminal RNA recognition motifs ( RRM ) domain and a conserved C-terminal SPOC domain [33 , 34] . RRM domain regulates chromatin modification by recognizing and binding to DNA/RNAs specifically [34] , while SPOC domain is proposed to facilitate protein-protein interactions in the transcription repression complex [29] . However , the underlying mechanism is unclear yet . In animals , Spen family members are reported to involve in neuron development , immune responses [29] and sex determination [35] , which is less clarified in plants . Phylogenetic analysis shows that there are three identified proteins close to LC3 ( S6 Fig ) , including Arabidopsis FPA that controls flowering time [36] . OsRRMh and OsRRM , two rice homologues of AtFPA , that control flowering , fertility , and architecture [37 , 38] . Protein structural analysis shows that compared to OsRRM , OsRRMh and AtFPA , the conserved RRM domain is absent in LC3 ( Fig 4B ) , suggesting the distinct function of LC3 . It is speculated that LC3 possibly interacts with other factors , which help to recognize DNA or RNA sequence and cooperate with LC3 to repress the transcription of downstream genes . Yeast two-hybrid screening was thus conducted to isolate the candidate proteins that interact with LC3 . Four proteins possibly interacting with LC3 were identified and designated as LIPs ( LC3-interacting proteins ) . Further analysis confirmed the interaction between LC3 and LIP1 , a HIT zinc finger domain-containing protein ( Fig 4C ) . Transcription pattern analysis showed that LIP1 presents similar expression pattern as LC3 during lamina joint development ( Fig 4D ) . Observation of fluorescence in rice protoplasts expressing LIP1-GFP/LC3-YFP revealed the similar localization of LIP1 and LC3 ( Fig 4E ) . Furthermore , transient expression of LC3-RFP and LIP1-GFP fusion proteins in rice protoplasts showed that LC3 co-localizes with LIP1 both in nucleus and cytoplasm ( Fig 4F ) . Split-Luciferase assay confirmed the interactions between LIP1 and LC3 in vivo ( Fig 4G ) , indicating that LIP1 may coordinate with LC3 to regulate the leaf inclination . As LC3 lacks the RRM domain , it is hypothesized that LC3 may repress the downstream genes OsIAA12 and OsGH3 . 2 through interacting with LIP1 , which recognize the binding sequences . Yeast one-hybrid analysis of OsIAA12 promoter ( fragments -1710 to 0 before ATG ) showed that LIP1 binds to later region ( -914 to 0 before ATG ) but not the forward one , and LC3 binds to neither region ( Fig 5A ) . In addition , by using 10-day-old transgenic seedlings expressing LC3-GFP , analysis of chosen five fragments in later region of OsIAA12 promoter by quantitative chromatin immunoprecipitation ( ChIP ) -PCR indicated the enrichment of four DNA fragments ( Fig 5B ) , confirming that LC3 binds to OsIAA12 promoter through interacting with LIP1 . To further confirm the repression effect of LC3-LIP1 on downstream genes , two effector constructs carrying LC3 and LIP1 fusion GFP , were transiently expressed with a luciferase reporter ( LUC ) construct containing ~2 . 7-kb promoter of OsIAA12 in rice protoplasts . Measurement showed that LUC expression was significantly reduced in the presence of LC3 or LC3-LIP1 , and no differences in the presence of single LIP1 ( Fig 5C ) , suggesting that LIP1 alone does not present inhibition effect , and LC3 and LIP1 cooperatively suppress the expressions of downstream genes . Similarly , ChIP-PCR assays of different DNA fragments in OsGH3 . 2 promoter showed the enrichment of two DNA fragments ( Fig 5D ) , indicating the binding of LC3 to OsGH3 . 2 promoter as well . These results suggest that LIP1 orchestrates with LC3 to repress the OsIAA12 and OsGH3 . 2 expressions . There is no change of leaf angles under LC3 overexpression ( S7 Fig ) , indicating that LC3 functions to maintain the normal leaf inclination . To testify the function of LIP1 in lamina joint development , plants deficiency of LIP1 in background of LC3 overexpression were generated by Crispr/Cas9 approach ( LIP1-Cas9 in LC3-ox ) . Eighteen independent transgenic lines were obtained and four of them were homozygous with either insertion or deletions at 5’ end of LIP1 ( Fig 6A ) . Observation and analysis of the leaf inclination of plants in fields showed the increased leaf inclination under LIP1 deficiency ( Fig 6A ) , indicating the crucial roles of LIP1 in mediating the LC3-LIP1 effects . In addition , the increased expression of OsIAA12 under LIP1 deficiency ( Fig 6B ) further demonstrate that LIP1 and LC3 synergistically inhibit the transcription of OsIAA12 expression .
SPOC-domain is speculated to facilitate protein-protein interactions in the transcription repression complex . Although SPOC domain-containing proteins are demonstrated to involve in regulation of various developmental processes , functions of them in plants are rarely reported . On the other hand , though auxin signaling/biosynthesis related genes are shown to affect the lamina joint development , the upstream regulation is still poorly understood . We functionally characterize a novel rice SPOC domain-containing protein leaf inclination 3 ( LC3 ) , whose deficiency ( lc3 mutant ) presents enhanced leaf angle due to the excessive cell elongation at adaxial side of lamina joint , and demonstrate that LC3 controls leaf inclination by regulating auxin signaling through interacting with LIP1 , a HIT zinc finger domain-containing transcriptional factor . It is therefore proposed that LC3 interacts with LIP1 to cooperatively suppress the expression levels of OsIAA12 and OsGH3 . 2 , resulting in the suppressed auxin signaling and homeostasis , to maintain the normal lamina joint development ( Fig 6C ) . Our findings not only identify a novel factor regulating leaf inclination through auxin signaling and homoeostasis , but also reveal the function and underlying mechanism of a novel SPOC domain-containing protein . Previous reports showed that RRM domain of SPOC domain-containing protein functions to recognize and bind to DNA/RNAs . Deficiency of RRM domain suggests that LC3 acts as a transcription repressor through interacting with other factors . Indeed , LIP1 , a HIT zinc finger domain-containing TF , recognizes specific DNA sequence and forms a heterodimer with LC3 through interaction to suppress the transcription of downstream genes , especially OsIAA12 and OsGH3 . 2 . These illustrate the mechanism how LC3-LIP1 heterodimer represses the expression of auxin signaling and homeostasis related genes . In addition , it’s the first time to characterize the function and relevant mechanism of a SPOC-domain containing protein lacking RRM domain , which expands the knowledge on regulating the expression of downstream target genes in addition to RRM domain . A mouse Spen-like protein , MINT , binds to homeoprotein Msx2 to co-regulate osteocalcin [34 , 39] , and our results provide another example showing how SPOC-domain containing protein functions through interacting with a HIT zinc finger domain-containing protein , suggesting that SPOC-domain containing protein may interact with distinct TFs to suppress the transcription of specific genes , which provides novel insights for the functions of SPOC-domain containing proteins . The underlying mechanism how LC3-LIP1 represses downstream target genes expression and whether there are other factors involving in the regulation , still need further investigations . In human , SHARP ( SMRT/HDAC1-associated repressor protein ) , a spen protein , interacts with co-repressor SMRT ( silencing mediator for retinoid and thyroid receptors ) and NCoR ( nuclear receptor corepressor ) , and these co-repressors repress transcription by recruiting a large complex containing histone deacetylase ( HDAC ) activity [29 , 40] . In mice , Znhit1 binds to HDAC1 and suppresses CDK6 expression by decreasing the histone H4 acetylation level in its promoter region [41] . Whether LC3 interacts with histone deacetylase or any other factors to repress the downstream gene transcription and hence regulates the distinct developmental processes needs further studies . Plant phytohormone IAA plays crucial roles in lamina joint development , however , the upstream regulations of the key negative regulator Aux/IAAs during the process and which distinct IAA-ARF interaction is involved in lamina joint development control are unclear . We at first time demonstrate that a SPOC domain-containing protein LC3 regulates auxin signaling by directly suppress OsIAA12 and auxin homeostasis through OsGH3 . 2 . Aux/IAAs bind to ARFs to suppress its function [42] and as a multi-member family ( there are 25 ARFs and 31 Aux/IAA proteins in rice ) , studies have revealed a complex interacting network of Aux/IAAs-ARFs that participate in regulation of various aspects of plant growth and development . Although it is known that each IAA protein can interact with different ARFs and each ARF protein can be suppressed by different IAAs to perform the diverse and specific functions [43] , distinct functions of each interaction pair and how IAA-ARF interaction regulates leaf inclination remain to be elucidated . Our studies demonstrate the specific role of OsIAA12-OsARF17 interaction , which will help to illustrate the auxin effects in lamina joint development . Interestingly , other Aux/IAAs proteins ( i . e . OsIAA20 , OsIAA21 , and OsIAA31 ) interact with OsARF17 besides OsIAA12 , and OsIAA12 can also bind with OsARF21 [32] , whether other Aux/IAAs-ARFs interactions regulate leaf inclination need further investigation . Further studies of the downstream genes of OsARF17 will expand the understanding of the detailed mechanism of OsIAA12-OsARF17 regulation in lamina joint development . In addition , the expression level of neither LC3 nor LIP1 is influenced by exogenous IAA treatment ( S8 Fig ) , what kind of factors regulate LC3-LIP1 complex and hence lamina joint development will be interesting to be investigated . GH3 family members encode an indole-3-acetic acid-amido synthetase that conjugates free IAA to various amino acids [44 , 45] . In addition to the regulation of GH3 gene by ARFs , which is conserved among dicot and monocot plants [7] , our results provide further understanding of GH3 gene regulations by other regulators .
Rice Zhonghua11 ( ZH11 , Oryza sativa japonica variety ) plants , lc3 mutant , and transgenic lines were grown in Shanghai and Lingshui ( Hainan Province ) under standard paddy conditions . Seedlings used to isolate protoplasts were grown in MS medium at 28°C with 12h-light/12-h dark cycle . To analyze the expression pattern , lamina joints of flag leaf were collected from 60- , 65- , 70- or 80-day-old plants ( stages 2 , 4 , 5 , 6 according to definition by Zhou et al . , 2017 ) . Leaf , root , and stem were collected from 7- or 20-day-old seedlings . Seeds ( 3 , 6 , or 9 days after pollination ) , anther , pistil and spikelet were sampled . For auxin treatment , 7-day-old seedlings were immersed in liquid 1/2 MS ( Murashige and Skoog ) medium containing IAA ( indole-3-acetic acid , 10 μM ) for 2 h and the collars were collected for qRT-PCR ( quantitative real-time RT-PCR ) analysis . Leaf angle measurement and paraffin section were performed using plants 10 days after heading . Collected leaf lamina joints were photographed , and angle between sheath and leaf was measured by ImageJ program . At least 30 leaf angles of individual plants were measured . For paraffin section analysis , leaf lamina joints were fixed in FAA solution ( 45% ethanol , 5% acetic acid , and 12 . 5% formaldehyde in water ) for 24 h and dehydrated in a graded ethanol series and xylene-ethanol solution . Samples were embedded in paraffin ( Sigma ) for 1 day , then sections were cut ( 10 mm ) and deparaffinized in xylene , hydrated through a graded ethanol series , and stained with Toluidine Blue . Extra stain was flushed and sections were dehydrated by a graded ethanol series again . Sections were microscopically observed and photographed , and cell number and cell size were calculated using ImageJ software . Entire LC3 gene sequence including the 3-kb promoter region was amplified using primers LC3-3/LC3-4 and subcloned into pCAMBIA2300 for complementation study . Coding sequence of LC3 was amplified by primers LC3-13/LC3-14 and subcloned into pCAMBIA1300 driven by maize ubiquitin promoter for overexpression analysis . LC3 promoter region was amplified using primers LC3-11/LC3-12 and subcloned into pCAMBIA1300+pBI101 vector [46] to drive the β-glucuronidase ( GUS ) gene . A binary vector pCAMBIA2300 carrying OsIAA12 coding sequence amplified by primers IAA12-3/IAA12-4 and driven by Zea mays ubiquitin promoter was constructed for overexpressing OsIAA12 . Transgenic rice with OsARF17 mutation was generated by Crispr/Cas9 [47] with pOs-sgRNA using primers ARF17-1/ARF17-2 . The gene editing construct for LIP1 deficiency via Crispr/Cas9 was designed using primers ( LIP1-5/LIP1-6 and LIP1-7/LIP1-8 ) as previously described [48] . Confirmed constructs were transformed into rice by Agrobacterium-mediated transformation . Sequences of used primers were listed in S1 Table . Various tissues were collected from the confirmed positive transgenic lines and incubated in substrate buffer ( pH 7 . 0 NaH2PO4 , 0 . 1M; EDTA , 10 mM; K4Fe ( CN ) 6 , 0 . 5 mM; K3Fe ( CN ) 6 , 0 . 5 mM; 1% Trition X-100; 40 mg/mL X-Gluc ) . Examined samples were vacuumed and kept at 37°C overnight , then washed with 75% ethanol and observed . Total RNAs were extracted by Trizol reagent ( Invitrogen ) and used to synthesize cDNA through reverse transcription ( Toyobo ) . qRT-PCR was conducted in a total volume of 20 μL containing 10 μL SYBR Premix Ex-Taq , 0 . 2 μL cDNA , primers ( 0 . 2 mM ) and 8 . 3 μL double distilled water . Rice Actin gene was used as an internal control and transcription levels of LC3 , LIP1 , OsIAA12 , OsGH3 . 2 , OsARF17 were examined using primers LC3-1/LC3-2 , LIP1-9/LIP1-10 , IAA12-1/IAA12-2 , GH3 . 2-1/GH3 . 2–2 , ARF17-9/ ARF17-10 . Other primers used in qRT-PCR analysis were listed in S1 Table . All examinations were conducted with three biological and technological replicates . Coding sequences of LC3 , LIP1 , OsIAA12 , and OsARF17 were amplified using primer LC3-5/LC3-6 , LIP1-1/LIP1-2 , IAA12-5/IAA12-6 , and ARF17-3/ARF17-4 and subcloned into pGADT7 and pGBKT7 vectors respectively ( Clontech ) . Confirmed constructs were co-transformed into yeast AH109 strain . Transformed yeast clones were diluted 101 , 102 , 103 times , grown on synthetic dropout ( SD ) medium in the presence or absence of histidine with different concentrations of 3-amino-1 , 2 , 4-triazole ( 3-AT ) according to the manufacturer’s instructions ( Matchmaker user’s manual , Clontech , California ) , and observed after 4 days . pGBKT7-LC3 vector was transformed into yeast strain Y2H Gold and used as bait in yeast-two hybrid screening analysis . The prey cDNAs derived from a rice cDNA library constructed from rice seedlings at different stages of ZH11 were transformed into yeast strain Y187 . Mating bait and prey plasmid transformants were rotated at low speed for 20 h , then grown on synthetic dropout ( SD ) medium absence of histidine . Identified candidate prey cDNA was isolated from yeast cells and transformed into Escherichia coli for sequencing . Full cDNA sequence was amplified and cloned into pGADT7 , which was co-transformed with pGBKT7-LC3 into yeast strain AH109 to verify the interaction . Coding sequence of OsIAA12 , OsARF17 , LC3 , LIP1 were amplified using primers IAA12-7/IAA12-8 , ARF17-5/ARF17-6 , LC3-9/LC3-10 , LIP1-3/LIP1-4 , LC3-15/LC3-16 and subcloned into pBI221-RFP [49] or pA7 vectors ( C-terminus fusion with GFP or YFP ) respectively . Resultant constructs expressing OsIAA12-RFP , OsARF17-GFP , LC3-YFP , LIP1-GFP , LC3-RFP fusion proteins were transformed into rice protoplasts and fluorescence was observed by confocal laser scanning microscope ( FV10i , OLYMPUS ) after 16 h . Coding sequences of OsIAA12 and OsARF17 were amplified using primers IAA12-9/IAA12-10 , ARF17-7/ARF17-8 and subcloned into pCAMBIA1300S-YN and pCAMBIA1300S-YC vector by Infusion kit ( Clontech ) separately . Resultant constructs were transformed into Agrobacterium tumefaciens strain GV3101 , which were used to infiltrate the leaves of 6-week-old tobacco plants . After infiltration for 48 h , YFP fluorescence was observed using a confocal laser scanning microscope ( FV10i , OLYMPUS ) . Fusion proteins nYFP-OsHAL3 and cYFP-OsHAL3 were used as a positive control [50] . Coding sequences of LC3 and LIP1 were amplified and subcloned into the Gateway vector nLUC and cLUC respectively . After infiltration of tobacco leaves for 48 h , excess luciferin was sprayed on leaves and kept in dark for 10 min to eliminate the background fluorescence . Relative LUC activity was measured by a low light cooled CCD imaging apparatus at -70°C . Experiments were repeated three times for each assay . Coding sequence of LC3 and LIP1 were amplified using primers LC3-7/LC3-8 and LIP1-1/LIP1-2 , and subcloned into pGADT7 ( Clontech ) . OsIAA12 promoter regions Pro2 ( -1709 to -915 bp before ATG ) and Pro3 ( -914 bp to 0 before ATG ) were amplified using primers IAA12-11/IAA12-12 , IAA12-13/IAA12-14 and subcloned into pHIS2 vector . Resultant constructs were transformed into yeast strain Y187 . Yeast transformants were grown on synthetic dropout ( -Leu/-Trp/-His ) medium containing 175 mM 3-AT for 3 days and observed . Experiments were repeated three times . ChIP-PCR assays were performed according to previous description [51] . Genomic DNAs extracted from 10-day-old transgenic seedling expressing LC3-GFP were digested into small pieces and crosslinked with formaldehyde . Resultant DNA fragments were sonicated to be ~200 bp in length . Chromatin immunoprecipitation were performed using anti-GFP antibody ( ab290; Abcam ) , and Normal rabbit lgG ( 10500C; Invitrogen ) was used as a negative control . Samples collected before immunoprecipitation were ‘input DNA’ . Immunoprecipitated and input DNA were purified with PCR purification kit ( Qiagen ) and amplified using primers covering around 150-bp region of OsIAA12 or OsGH3 . 2 promoters by qPCR to examine the ChIP enrichment . Sequences of primers ( IAA12-17 ~ IAA12-28 , and GH3 . 2–3 ~ GH3 . 2–12 ) are listed in S1 Table . Fold-enrichment was calculated by normalizing the amount of a target DNA fragment against the respective input DNA samples . Experiments were repeated three times . For effector constructs , coding regions of LC3 and LIP1 were amplified using primers LC3-9/LC3-10 and LIP1-3/LIP1-4 and subcloned into vector pA7 ( C-terminus fusion with GFP ) . A ~2 . 7-kb DNA fragment of OsIAA12 promoter was amplified by primers IAA12-15/IAA12-16 and subcloned into a modified pGreen0800 vector to generate the reporter construct . Effector and reporter constructs were co-transformed intro rice protoplasts . Dual-luciferase transcriptional activity assay was performed as previously described [52] . Experiments were biologically repeated three times . Ten-day-old ZH11 seedlings were used to isolate protoplasts and 100 μL protoplasts suspension ( containing ~2×105 protoplasts ) were transfected with plasmid ( 5–10 μg DNA ) and 110 μL PEG solution . Transformation mixture was incubated in darkness for 15 min at 28°C , then diluted by 1 mL W5 solution ( NaCl , 154 mM; CaCl2 , 125 mM; D-Glucose , 5 mM; KCl , 5 mM; MES , 2 mM , pH 5 . 7 ) and centrifuged at 100 g for 2 min . Protoplasts were suspended in WI solution ( Mannitol , 0 . 5 M; KCl , 20 mM , MES , 4 mM , pH 5 . 7 ) and transferred into multi well plates and incubated at 28°C for 16 h . To construct a phylogenetic tree of SPOC-domain protein , homolog sequences in A . thaliana , O . sativa were obtained at the TAIR Web site ( http://www . arabidopsis . org ) and Rice Genome Annotation Project ( http://rice . plantbiology . msu . edu ) . Alignment of available sequences was performed with CLUSTALX 1 . 83 . The phylogenetic tree was constructed with MEGA 3 [53] using the neighbor-joining algorithm with 1001 bootstrap replicates . All relevant data are within the paper and its Supporting Information files except for the sequence data which is available from the rice genome database RICEGE ( http://signal . salk . edu/cgi-bin/RiceGE ) or GenBank databases ( https://www . ncbi . nlm . nih . gov/genbank/ ) under the following accession numbers: LC3 ( Os06g0595900 ) , LIP1 ( Os10g0520700 ) , OsIAA12 ( Os03g0633800 ) , OsARF17 ( Os06g0677800 ) .
|
Leaf angle is a major trait of ideal architecture of crops that associates with photosynthetic efficiency and yields . Studies of the underlying mechanism will greatly help to improve the crop yield . Phytohormones especially auxin and brassinosteroids play crucial roles in regulating the leaf inclination , however , the upstream regulatory network is still unknown . Here , we functionally characterize a novel SPOC domain-containing protein LC3 ( leaf inclination3 ) in lamina joint development through regulating auxin signaling . LC3 deficiency results in the excessive cell elongation at lamina joint adaxial side and hence the enlarged leaf angle . LC3 acts as a transcription suppressor through interacting with LIP1 ( LC3-interacting protein 1 , a HIT zinc finger domain-containing transcription factor ) , which directly binds to the promoters of auxin signaling and homeostasis related genes . Our studies provide new insights in the functional mechanism of SPOC domain-containing proteins and help to elucidate how auxin signaling is regulated during lamina joint development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"biotechnology",
"plant",
"anatomy",
"engineering",
"and",
"technology",
"gene",
"regulation",
"hormones",
"plant",
"science",
"rice",
"plant",
"hormones",
"genetically",
"modified",
"plants",
"experimental",
"organism",
"systems",
"plants",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"bioengineering",
"grasses",
"genetically",
"modified",
"organisms",
"animal",
"studies",
"proteins",
"gene",
"expression",
"leaves",
"biochemistry",
"plant",
"biochemistry",
"eukaryota",
"plant",
"and",
"algal",
"models",
"adaxial",
"surface",
"protein",
"domains",
"genetics",
"biology",
"and",
"life",
"sciences",
"auxins",
"plant",
"biotechnology",
"organisms"
] |
2018
|
SPOC domain-containing protein Leaf inclination3 interacts with LIP1 to regulate rice leaf inclination through auxin signaling
|
Histone deacetylase 8 from Schistosoma mansoni ( SmHDAC8 ) is essential to parasite growth and development within the mammalian host and is under investigation as a target for the development of selective inhibitors as novel schistosomicidal drugs . Although some protein substrates and protein partners of human HDAC8 have been characterized , notably indicating a role in the function of the cohesin complex , nothing is known of the partners and biological function of SmHDAC8 . We therefore employed two strategies to characterize the SmHDAC8 interactome . We first used SmHDAC8 as a bait protein in yeast two-hybrid ( Y2H ) screening of an S . mansoni cDNA library . This allowed the identification of 49 different sequences encoding proteins . We next performed co-immunoprecipitation ( Co-IP ) experiments on parasite extracts with an anti-SmHDAC8 antibody . Mass spectrometry ( MS ) analysis allowed the identification of 160 different proteins . SmHDAC8 partners are involved in about 40 different processes , included expected functions such as the cohesin complex , cytoskeleton organization , transcriptional and translational regulation , metabolism , DNA repair , the cell cycle , protein dephosphorylation , proteolysis , protein transport , but also some proteasome and ribosome components were detected . Our results show that SmHDAC8 is a versatile deacetylase , potentially involved in both cytosolic and nuclear processes .
Schistosomiasis is a neglected tropical parasitic disease of major public health importance [1–3] caused by blood flukes of the Schistosoma spp . , with 258 million people requiring treatment worldwide , and 780 million at risk of infection ( http://www . who . int/mediacentre/factsheets/fs115/en/ ) . There is currently no effective vaccine against schistosomiasis [4] and the treatment of the disease relies on a single drug , praziquantel ( PZQ ) . Because of the intensive use of PZQ , field observations show the appearance of schistosome strains resistant to PZQ [5 , 6] . Thus , the development of new drugs is imperative . We have used a “piggy-back” strategy that consists in identifying orthologues of proteins already targeted in other pathologies , like cancer . Among these , we have chosen to target enzymes involved in epigenetic processes [7] and in particular , histone deacetylases ( HDAC ) , which are among the most studied epigenetic targets . Schistosoma mansoni possesses three class I HDACs ( SmHDAC1 , 3 and 8 ) and four class II HDACs , two class IIa and two class IIb enzymes [8 , 9] . We have shown [10] that Trichostatin A ( TSA ) , a pan-inhibitor of HDACs , induces hyperacetylation of histones , deregulates gene expression and induces the death of schistosome larvae and adult worms in culture . Schistosome HDACs are therefore promising targets for the development of new drugs against schistosomiasis , especially SmHDAC8 [11 , 12] . Indeed , it is the only schistosome class I HDAC for which the structure of its catalytic pocket differs significantly from that of the human orthologue [13] , allowing the development of selective inhibitors that are toxic for schistosome larvae ( apoptosis and death ) and adult worms ( changes in the reproductive organs , separation of worm pairs and arrest of egg laying ) in culture [13–15] . Moreover , transcript knockdown of SmHDAC8 leads to markedly reduced parasite viability and fecundity [13] suggesting that this enzyme is essential to parasite growth and development . Human HDAC8 ( hHDAC8 ) catalyzes the deacetylation of lysine residues within histone and non-histone proteins but its biological role has long remained elusive [11] . The two best-characterized non-histone substrates are the estrogen-related receptor ( ERRα ) and the structural maintenance chromosome 3 protein ( SMC3 ) . hHDAC8 interacts directly with ERRα in vivo and deacetylates ERRα in vitro , increasing its DNA binding affinity [16] . In the case of SMC3 , a member of the cohesin complex , hHDAC8 is involved in its deacetylation that allows the recycling of the cohesin complex , and hHDAC8 mutations are linked with the Cornelia de Lange syndrome [17 , 18] . Two recent studies using more systematic approaches identified novel hHDAC8 substrates . The first [19] detected seven proteins all of which are nuclear proteins , including SMC3 , and this approach failed to identify histones or ERRα as substrates . The second study [20] detected 19 novel hHDAC8 substrates , not all of which were nuclear . Among them , the cohesin complex component SMC1A , but also all the protein substrates identified by Olson [19] were predicted by this approach . Clearly , each approach provides different information about the hHDAC8 substrates . In order to characterize the protein partners of the 11 human HDACs , Joshi et al . [21] performed a global study of their interactions by establishing T-lymphoblast cell lines stably expressing Enhanced Green Fluorescent Protein ( EGFP ) -tagged hHDACs combined with proteomics and functional studies to identify hHDAC-containing protein complexes . They showed that in these cells hHDAC8 interacts with 15 proteins: four related to the cell cycle including SMC1A and 3 , three related to protein and ion transport and 8 with other/unknown functions . However , although these proteins are members of hHDAC8-containing complexes , the methodology used cannot distinguish whether or not they interact directly with the enzyme . Moreover , human acetylome analysis [22] interestingly reveals that among these partner proteins , SMC1A , SMC3 , SA2 , SEC16A and NUP98 are acetylated but only SMC1A and SMC3 have been identified as hHDAC8 substrates [19 , 20] . SmHDAC8 shows major structural differences compared to hHDAC8 , most obviously in the presence of insertions within the catalytic domain that form loops at the surface of the protein [13] and therefore represent potential surfaces for protein-protein interactions . We therefore sought to determine whether SmHDAC8 interacts with the same or different proteins than hHDAC8 . Since it is not possible to overexpress a tagged “bait” protein in schistosomes we decided to use two different methods to identify protein partners , yeast two-hybrid screening ( Y2H ) and co-immunoprecipitation experiments coupled to mass spectrometry ( Co-IP/MS ) . The former technique characterizes direct protein-protein interactions and the latter identifies proteins that may interact directly or are members of protein complexes that interact with the target protein . They therefore yield different results , but in combination give an overall picture of the cellular processes in which SmHDAC8 participates . Our results show that SmHDAC8 is a versatile deacetylase , potentially involved in both cytosolic and nuclear processes , and contribute to the understanding of its status as a therapeutic target .
A yeast two-hybrid ( Y2H ) S . mansoni adult worm ( 6 week-old male and female worms ) cDNA library that consists of the Gal4-activation domain ( Gal4-AD ) , amino acids 768–881 , fused with S . mansoni adult worm cDNA was used . The cDNA library was constructed according to the manufacturer's instructions ( Matchmaker Library Construction and Screening Kit , Clontech ) using the pGADT7 plasmid containing the LEU2 reporter gene . The cDNA library was transformed into Saccharomyces cerevisiae AH109 strain containing HIS3/ADE2/LacZ reporter genes , under the conditions recommended by the supplier ( Yeast Protocols Handbook , Clontech ) . The cDNA library was screened with a bait construct corresponding to the Gal4-DNA binding domain ( Gal4-DBD ) fused with the full-length coding sequence of SmHDAC8 ( EF077628 ) [8] , amplified using oligonucleotides SmHDAC8 Fw ( 5'-GCTCGAATTCATGTCTGTTGGGATCG-3' ) and SmHDAC8 Rw ( 5'- ACCTCGAGGATCCCATACCAGTTAAATTATA-3' ) , and cloned into the EcoRI and BamHI restriction sites of the pGBKT7 vector bearing the TRP1 reporter gene . The S . cerevisiae Y187 strain containing the LacZ reporter gene was transformed with the bait construct and mated with the AH109 strain overnight . After incubation , diploid yeasts were plated on selective medium lacking adenine , histidine , leucine and tryptophan and the plates were incubated at 30°C for 5 days . Positive clones were confirmed both by restreaking on selective medium and by a liquid LacZ assay ( Yeast Protocols Handbook , Clontech ) . Each selected positive clone was cultivated in medium lacking leucine . Cells were harvested by centrifugation ( 2 , 000 g , 20 min ) and disrupted with glass beads . Plasmid extraction was performed with the Nucleospin plasmid kit ( Macherey-Nagel ) according to the manufacturer’s instructions . Plasmids extracted from yeast cells were then transformed in Subcloning Efficiency DH5α competent cells ( Invitrogen ) . Prior to sequencing ( Eurofins Genomics ) , about 400 individual clones were pre-screened by digestion with the restriction enzymes HindIII and EcoRV and electrophoresis on agarose gels in order to select clones with unique restriction profiles and weed out duplicates . After extraction of plasmids , each clone was transformed in the AH109 strain to confirm that the interaction was robust . The S . cerevisiae Y187 strain was transformed with the bait construct SmHDAC8 pGBKT7 and mated with the AH109 strain overnight . After incubation , diploid yeasts were plated first on selective medium lacking leucine and tryptophan and then on another higher stringency selective medium lacking adenine , histidine , leucine and tryptophan and the plates were incubated at 30°C . Sequence analyses ( correction and alignment ) were performed using Sequencher software ( Gene Codes Corporation ) . Identification and functional annotation of SmHDAC8 interactors was performed using Blast 2GO software [23] . All animal experimentation was conducted in accordance with the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes ( ETS No 123 , revised Appendix A ) and was approved by the committee for ethics in animal experimentation of the Nord-Pas de Calais region ( Authorization No . AF/2009 ) and the Pasteur Institute of Lille ( Agreement No . A59-35009 ) . A Puerto Rican strain ( NMRI ) of S . mansoni is maintained in the laboratory using the intermediate snail host Biomphalaria glabrata and the definitive golden hamster host Mesocricetus auratus . S . mansoni adult worms were obtained by hepatic portal perfusion of hamsters infected six weeks previously [24] . Purified recombinant SmHDAC8 ( a kind gift from M . Marek and C . Romier , IGBMC , Strasbourg , France [13] ) was used to generate rat polyclonal antiserum . Male Lou Rats were injected i . p . with 50 μg of SmHDAC8 with alum adjuvant in a total volume of 500 μL three times at two-week intervals . The rats were bled two weeks after the final injection . The monospecificity of the rat antiserum was controlled after SmHDAC8 immunoprecipitation from S . mansoni protein extract ( see section below ) and western blotting ( S1 Fig ) . Three independent experiments were performed as follows . Adult worms ( 50 couples ) were suspended in 500 μL of lysis buffer ( 20 mM Tris HCl pH 7 . 4 , 50 mM NaCl , 5 mM EDTA , 1% Triton and protease inhibitors ) , crushed with a Dounce homogenizer and sonicated ten times for 30 s ( maximum power , Bioruptorplus , Diagenode ) . After centrifugation , at 10 , 000 g for 10 min at 4°C , immunoprecipitation of SmHDAC8 was performed using the Pierce Crosslink Immunoprecipitation Kit ( Thermo Scientific ) according to the manufacturer’s instructions . Briefly , the protein lysate ( 500 μL ) was pre-cleared by incubation with 20 μL of IgG from rat serum crosslinked to protein-L Agarose beads ( Thermo Scientific ) for 2 h at 4°C on a rotator . Then , pre-cleared lysate was collected after centrifugation , at 1 , 000 g for 1 min at 4°C , and incubated overnight at 4°C on a rotator , with 1 μL of anti-SmHDAC8 antibodies or 1 μL of IgG from rat serum as a control , bound to protein-L Agarose beads . Protein samples were denatured at 100°C in 5% SDS , 5% β-mercaptoethanol , 1 mM EDTA , 10% glycerol , and 10 mM Tris pH 8 buffer for 3 min , and subsequently fractionated on a 10% acrylamide SDS-PAGE gel . Electrophoretic migration was stopped when the protein sample had entered 1 cm into the separating gel . The gel was labelled briefly with Coomassie Blue , and five bands , containing the whole sample , were cut out . Digestion of proteins in the gel slices was performed as previously described [25] . Separation of the protein digests was carried out using an UltiMate 3000 RSLCnano System ( Thermo Fisher Scientific ) . Peptides were automatically fractionated onto a commercial C18 reversed phase column ( 75 μm × 150 mm , 2 μm particle , PepMap100 RSLC column , Thermo Fisher Scientific , temperature 35°C ) . Trapping was performed during 4 min at 5 μL/min , with solvent A ( 98% H2O , 2% ACN ( acetonitrile ) and 0 . 1% FA ( formic acid ) ) . Elution was carried out using two solvents A ( 0 . 1% FA in water ) and B ( 0 . 1% FA in ACN ) at a flow rate of 300 nL/min . Gradient separation was 3 min at 5% B , 37 min from 5% B to 30% B , 5 min to 80% B , and maintained for 5 min . The column was equilibrated for 10 min with 5% buffer B prior to the next sample analysis . Peptides eluted from the C18 column were analyzed by Q-Exactive instruments ( Thermo Fisher Scientific ) using an electrospray voltage of 1 . 9 kV , and a capillary temperature of 275 °C . Full MS scans were acquired in the Orbitrap mass analyzer over the m/z 300–1200 range with a resolution of 35 , 000 ( m/z 200 ) and a target value of 5 . 00E + 05 . The ten most intense peaks with charge state between 2 and 4 were fragmented in the HCD collision cell with normalized collision energy of 27% , and tandem mass spectra were acquired in the Orbitrap mass analyzer with resolution 17 , 500 at m/z 200 and a target value of 1 . 00E+05 . The ion selection threshold was 5 . 0E+04 counts , and the maximum allowed ion accumulation times were 250 ms for full MS scans and 100 ms for tandem mass spectrum . Dynamic exclusion was set to 30 s . Raw data collected during nanoLC-MS/MS analyses were processed and converted into * . mgf peak list format with Proteome Discoverer 1 . 4 ( Thermo Fisher Scientific ) . MS/MS data were interpreted using search engine Mascot ( version 2 . 4 . 0 , Matrix Science , London , UK ) installed on a local server . Searches were performed with a tolerance on mass measurement of 0 . 2 Da for precursor and 0 . 2 Da for fragment ions , against a composite target decoy database ( 25 , 970 total entries ) built with the S . mansoni Uniprot database ( taxonomy id 6183 , 12 , 861 entries ) fused with the sequences of recombinant trypsin and a list of classical contaminants ( 124 entries ) . Up to one trypsin missed cleavage was allowed . For each sample , peptides were filtered out according to the cut-off set for protein hits with one or more peptides longer than nine residues , an ion score >30 , an identity score >6 , leading to a protein false positive rate of 0 . 8% .
The aim of our study was to identify SmHDAC8 partners in order to better apprehend its biological role and function . We used here two strategies to characterize the SmHDAC8 interactome . The yeast two-hybrid ( Y2H ) screening using SmHDAC8 revealed a large number of positive clones , of which we sequenced 137 after the prescreening step . After manual correction and assembly using Sequencher , we characterized 49 different sequences , which were then identified and functionally annotated using Blast 2GO software ( S1 Table ) . Three independent co-immunoprecipitation ( Co-IP ) experiments were performed , using an anti-SmHDAC8 antibody ( named IP1 , IP2 , and IP3 ) . As a control , we performed Co-IP with a rat IgG antibody alone in each experiment . Mass spectrometry ( MS ) of the Co-IP proteins identified 1 , 500 different proteins ( S2 Table ) . A significant degree of variation between the proteins identified in each experiment was noted , for which several reasons can be invoked . The three parasite protein extracts were each obtained from a pool of S . mansoni adult worms of both sexes and not from homogeneous cell cultures . Variations in protein expression between worm batches could induce differences between the three Co-IP/MS experiments . Moreover , S . mansoni is a complex multicellular parasite and protein quantities can vary between the different cellular types within a given individual as well as between different worms . Therefore , to take account of this relative variability between each extract , we chose to pool the results obtain for the three Co-IP/MS experiments IP1 , IP2 and IP3 . The possibility that the observed variability may have been due to non-specific interactions of our anti-SmHDAC8 antibody with other proteins can be discounted . A single band corresponding to the molecular weight of SmHDAC8 was detected on western blots of the immunoprecipitated material ( S1 Fig ) and the only HDAC detected by MS in the immunoprecipitates was SmHDAC8 ( S2 Table ) . Of the 1 , 500 proteins for which peptides were detected we selected only those that fulfilled three criteria: ( i ) at least three peptides in the Co-IP experiment , ( ii ) with no more than two peptides in the control and ( iii ) with a spectral count ratio between Co-IP SmHDAC8 and control of greater than 3 . After , this selection step we obtained 160 different proteins that were considered good candidates as SmHDAC8 partners ( S2 Table ) . Among the proteins identified by these two approaches , four are common between Y2H and Co-IP/MS: the Proliferation-associated protein 2G4 , 38kDa ( PA2G4 , n°G4LXR6 ) , Cathepsin-B1 ( SmCB1 , n°Q8MNY2 ) , putative NADH-ubiquinone oxidoreductase ( n°G4VK53 ) and microsomal glutathione S-transferase 3 ( GST-3 , n°G4VH65 ) ( Fig 1 ) . Among these proteins , two illustrate the previously characterized roles in HDAC-dependent processes and/or the potential involvement of acetylation . ( i ) PA2G4 is highly conserved in eukaryotes . The human member of this family , ErbB3 binding protein 1 ( Ebp1 ) was identified as a putative downstream member of an ErbB3-regulated signal transduction pathway [26] . More particularly , the C-terminal region ( 300–372 ) of Ebp1 , which is important for transcriptional repression , was shown to bind HDAC2 and inhibitors of HDACs significantly reduced Ebp1-mediated transcriptional repression [27 , 28] . Like its human counterpart , the interaction between PA2G4 and SmHDAC8 is mediated by its C-terminal moiety , corresponding to the fragment cloned in the Y2H screen . ( ii ) SmCB1 is an essential gut-associated peptidase that digests host blood proteins as a source of nutriments ( note that we also detected , but only with the Co-IP/MS experiments , cathepsin D and L2 which are also part of the gut peptidase network ) and is also a drug target because enzyme inhibition induced severity phenotypes in the parasite [29] . Moreover , SmCB1 is a promising vaccine candidate because its administration elicits protection against S . mansoni challenge infection in mice and hamsters [30 , 31] . A direct link with HDACs has never been reported . Nevertheless , acetylome analysis from Schistosoma japonicum reveals that Cathepsin-B ( n°Q7Z1I6 ) is acetylated on K241 and sequence alignments between SjCB1 and SmCB1 show that it is conserved . The lack of convergence between Y2H screening and Co-IP/MS is perhaps not entirely surprising since they represent very different methods for interactome studies . Both strategies have weaknesses . In Y2H screening , direct interactions between two proteins are detected , but in some cases , these may be non-specific due to the juxtaposition of proteins or fragments that are never in contact within the cell . In the case of Co-IP/MS , some proteins cannot be identified because they are present in low quantities in the parasite extract and some proteins identified may be members of immunoprecipitated complexes and are not direct partners . The results obtained can therefore be considered as complementary and , taken together , provide an overall picture of the cellular processes in which SmHDAC8 is involved . For some of the partner proteins identified only by Y2H and not by Co-IP/MS screening we carried out independent experiments to verify the interaction with SmHDAC8 . For instance , the binding of SmCtBP , SmMBD2 , tensin and actin-1 proteins to SmHDAC8 was verified by candidate-specific Y2H experiments . As expected , SmHDAC8 indeed interacted with SmCtBP and SmMBD2 , as well as the other two proteins ( Fig 2 ) . These results are in agreement with available data for the human orthologues . ( i ) The human C terminal binding protein ( CtBP ) family members appear to mediate transcriptional repression in a histone deacetylase ( HDAC ) -dependent manner [32] . Some human class I HDACs , HDAC 1 , 2 and 3 [33–35] and class II HDACs ( HDAC4 and 5 ) [36] are present in the CtBP1 nuclear protein complex , but the possible involvement of hHDAC8 was not investigated . ( ii ) The methyl-CpG binding domain protein 2 ( MBD2 ) binds to methylated DNA and represses transcription through the recruitment of NuRD co-repressor complex [37] . The MBD2-NuRD complex contains a histone deacetylase core composed of HDAC1/2 , RbAp46/48 and MTA2 [38–41] . The MBD2 protein possesses , among others , an intrinsically disordered region ( IDR ) able to recruit RbAp48 , HDAC2 and MTA2 [42] . More particularly , it is possible that HDACs bind directly to the MBD2 IDR , because the human HDAC interactome study reveals a specific interaction between MBD2 and HDAC1/2 [21] . The Database of Protein Disorders ( DisProt , www . disprot . org ) [43] predicts an IDR domain located between the MBD and the C-terminal coiled-coil domains of SmMBD2 , which may be involved in SmHDAC8 binding . Focusing on the biological processes predicted with the blast2GO software for each protein identified , among the 49 partners identified with the Y2H screening ( S1 Table ) , seven were of unknown function , although one encoded a peptide including an EGF-like domain and an IgG-like domain . One further sequence corresponded to a hitherto unannotated gene . The remaining proteins are involved in 21 different biological processes ( S1 Table ) , and those represented by the largest numbers of different sequences are cytoskeleton organization , transcription regulation , metabolism , transport , cell cycle regulation , DNA repair and chromatin remodeling ( Fig 3 ) . The 160 proteins identified by Co-IP/MS ( S2 Table ) are involved in 33 different biological processes , and those represented by the most different sequences are metabolic process , cytoskeleton organization , proteasome , proteolysis , translation regulation , transport , protein dephosphorylation and stress response ( Fig 3 ) . These results show that some of these processes ( 22% ) are common between our different experiments , and in particular metabolism and cytoskeleton organization ( Fig 3 ) . Moreover , they suggest that SmHDAC8 is a versatile deacetylase involved in both cytosolic ( cytoskeleton organization , ribosome , proteolysis or proteasome ) and nuclear ( transcription regulation , DNA binding and repair , chromatin remodeling or cohesin complex ) processes . This contrasts with the current knowledge of partners and substrates proteins of human HDAC8 ( hHDAC8 ) that emphasizes nuclear functions . The global interactome study [21] that identified 15 partner proteins and the recent systematic studies of hHDAC8 substrates [19 , 20] mainly identified nuclear partner proteins involved in the cohesin complex , transcriptional regulation and chromatin remodeling . Some extranuclear proteins were identified as partners , such as the Ca2+-dependent phospholipid binding protein Copine III [21] or substrates , such as the elongation factor 1 EF1α1 [20] but these were in the minority . Here , the involvement of SmHDAC8 in extranuclear functions is illustrated by the direct interaction between SmHDAC8 and proteins involved in cytoskeleton organization like tensin , actin-1 , actin 5c and Rho1 GTPase ( S1 and S2 Tables ) . As previously mentioned , the binding of tensin and actin-1 proteins to SmHDAC8 was verified by candidate-specific Y2H experiments . As expected , SmHDAC8 indeed interacted with SmActin-1 and SmTensin ( Fig 2 ) . Interestingly , Waltregny et al . have shown that hHDAC8 interacts with smooth muscle alpha actin ( but not beta-actin ) to regulate cell contractility [44 , 45] . Proteomic analyses have shown that all three human actin isoforms ( alpha , beta and gamma ) are acetylated [22 , 46] . Moreover , the actin-associated proteins cortactin [47] and anillin [20] are hHDAC8 substrates . Several regulatory proteins of actin polymerization in human are also acetylated ( gelsolin , CapZ , profilin and the Arp2/3 complex ) [22] and we found some orthologues in our SmHDAC8 Co-IP/MS analyses ( gelsolin , septin and a subunit of the Arp2/3 complex , S2 Table ) . Actin dynamics are also controlled by small GTPases of the Rho family , like RhoA , notably in the formation of focal adhesions and stress fibers [48–50] . RhoA was not identified by Joshi et al . [21] as interacting with hHDAC8 and does not seem to be acetylated in human [22] , although it was found to be acetylated in Schistosoma japonicum [51] . Here , our Y2H screen shows the direct interaction between SmHDAC8 and SmRho1 , the orthologue of human RhoA . Hence , it is possible that binding of SmHDAC8 to SmRho1 could participate in the control of the SmRho1 pathway . Among all the hHDAC8 substrates and partners identified [19–21] , only Sec proteins , involved in endoplasmic reticulum protein secretory pathways were also found to interact with SmHDAC8 , although these are different and phylogenetically distinct proteins: Sec16A in human and Sec1 and Sec61β in S . mansoni ( S2 Table ) . Strikingly , we found no orthologues of the major hHDAC8 substrates/partners already characterized among the protein partners of SmHDAC8 . This is exemplified by members of the cohesin complex , which is notably responsible for the correct separation of sister chromatids into two daughter cells during mitosis [52] . The tripartite cohesin ring is formed by two SMC ( structural maintenance of chromosome ) proteins SMC1A and SMC3 , and the α-kleisin subfamily protein RAD21 [53–55] . Both SMC3 and SMC1A were found in the hHDAC8 interactome [21] , as is SA2 ( an accessory protein that binds to hRAD21 ) . In their study , hRAD21 was also detected , but at a non-significant level . Unexpectedly , we did not identify any of the cohesin complex members present in S . mansoni using our Co-IP/MS strategy contrary to Joshi et al . One possible explanation is that they used cells overexpressing EGFP-tagged hHDACs that generated a higher quantity of hHDAC8 protein than is present under physiological conditions . Correct MS identification depends strongly on the quantity of immunoprecipitated protein , and unfortunately , we are unable to increase SmHDAC8 expression in S . mansoni . However , and interestingly , our Y2H screen shows a direct interaction of SmHDAC8 with SmRAD21 . In order to confirm the Y2H screening result we carried out a candidate-specific Y2H experiment with the SmRAD21 clone and , as expected , SmHDAC8 interacts with SmRAD21 , suggesting that SmHDAC8 forms an integral part of the cohesin complex and may have a central role in cell division , transcriptional regulation and DNA repair . In conclusion , the protein partners of SmHDAC8 identified by our Y2H library screen and Co-IP/MS experiments are all orthologues of proteins not previously identified as substrates or partners of human HDAC8 . While we did not necessarily expect to detect substrates using these methodologies , it is striking that none of the partner proteins detected by Joshi et al . were among the SmHDAC8 partners . We propose that the principal reason for this is that the methodology used in the previous study , with tagged HDACs ( as bait to pull down protein complexes ) expressed in a T-lymphoblast cell line , may address a limited subset of potential partners . In contrast , our Co-IP/MS study was performed on adult worm protein extracts and S . mansoni is a complex parasite encompassing a variety of cell types . These Co-IP/MS approaches do not necessarily identify the proteins to which the HDACs physically bind because some proteins identified can be part of immunoprecipitated complexes . This is illustrated by the cohesin complex members identified by Joshi et al . ( SMC1A , SMC3 , and SA2 ) and that do not include RAD21 , which only appeared as a non-significant binder . This is the reason why we also performed an Y2H screen using SmHDAC8 in order to identify proteins that bound directly to SmHDAC8 . Our Y2H shows that SmHDAC8 physically binds to SmRAD21 and we suggest that the other cohesin complex components are pulled down by this interaction . Against this , it can be argued that the Y2H screen can identify non-specific interactions due to the juxtaposition of proteins or fragments that are never in contact within the cell . Again , this is why we have followed up the results of the initial screen with more detailed investigations of the interactions of two of the most interesting candidates , SmRAD21 and SmRho1 , allowing us to confirm that both are bona fide partners of SmHDAC8 . Despite the fact that we did not identify the same protein partners as Joshi et al . , it would be unwise to assume that all those identified in the Y2H screen and in the Co-IP/MS experiment are specific for the schistosome enzyme . It is probable that a number of their human orthologues will interact with hHDAC8 when investigated individually . Nevertheless , given the structural differences between the schistosome and human enzymes , and particularly the unstructured loops at the surface of SmHDAC8 , encoded by insertions in the catalytic domain sequence , it can be assumed that specific partners of the latter are present . Some of the proteins we identified have no human orthologues and have unknown functions . Further work will determine which of the partner proteins are schistosome-specific , how they interact with SmHDAC8 and the role of these interactions within the parasite .
|
Using a target-based strategy to develop new drugs for the treatment of schistosomiasis we had earlier identified Schistosoma mansoni histone deacetylase 8 ( SmHDAC8 ) as essential for parasite development and survival in the mammalian host . Selective inhibitors of this enzyme show promise as lead compounds for drug development . However , the biological role of SmHDAC8 has not been established . We identified the potential partner proteins and the processes in which it is involved by combining two methods: yeast two-hybrid screening and co-immunoprecipitation with mass spectrometry . These approaches yielded complementary sets of potential partner proteins that are actors in about 40 different cellular processes . These include known roles for the human counterpart of SmHDAC8 , like interactions with the cohesin complex and the cytoskeleton , as well as novel interactions such as with proteasomes and ribosomal proteins . Our results emphasize the implication of SmHDAC8 in both nuclear and cytosolic processes and the versatility of this enzyme , suggesting why it is vital to the parasite and a promising drug target .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] |
[
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"protein",
"interactions",
"helminths",
"dna-binding",
"proteins",
"animals",
"fungi",
"model",
"organisms",
"immunoprecipitation",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"co-immunoprecipitation",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"proteins",
"histones",
"molecular",
"biology",
"precipitation",
"techniques",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"yeast",
"biochemistry",
"eukaryota",
"library",
"screening",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"organisms",
"two-hybrid",
"screening"
] |
2017
|
Analysis of the interactome of Schistosoma mansoni histone deacetylase 8
|
Human papillomaviruses ( HPVs ) are epithelial tropic viruses that link their productive life cycles to the differentiation of infected host keratinocytes . A subset of the over 200 HPV types , referred to as high-risk , are the causative agents of most anogenital malignancies . HPVs infect cells in the basal layer , but restrict viral genome amplification , late gene expression , and capsid assembly to highly differentiated cells that are active in the cell cycle . In this study , we demonstrate that HPV proteins regulate the expression and activities of a critical cellular transcription factor , KLF4 , through post-transcriptional and post-translational mechanisms . Our studies show that KLF4 regulates differentiation as well as cell cycle progression , and binds to sequences in the upstream regulatory region ( URR ) to regulate viral transcription in cooperation with Blimp1 . KLF4 levels are increased in HPV-positive cells through a post-transcriptional mechanism involving E7-mediated suppression of cellular miR-145 , as well as at the post-translational level by E6–directed inhibition of its sumoylation and phosphorylation . The alterations in KLF4 levels and functions results in activation and suppression of a subset of KLF4 target genes , including TCHHL1 , VIM , ACTN1 , and POT1 , that is distinct from that seen in normal keratinocytes . Knockdown of KLF4 with shRNAs in cells that maintain HPV episomes blocked genome amplification and abolished late gene expression upon differentiation . While KLF4 is indispensable for the proliferation and differentiation of normal keratinocytes , it is necessary only for differentiation-associated functions of HPV-positive keratinocytes . Increases in KLF4 levels alone do not appear to be sufficient to explain the effects on proliferation and differentiation of HPV-positive cells indicating that additional modifications are important . KLF4 has also been shown to be a critical regulator of lytic Epstein Barr virus ( EBV ) replication underscoring the importance of this cellular transcription factor in the life cycles of multiple human cancer viruses .
The life cycle of human papillomaviruses is dependent upon host cell replication , differentiation and cellular gene expression [1 , 2] . HPVs infect stratified squamous epithelia through small wounds that expose basal cells to entry . Upon entry , viral genomes are maintained as low copy nuclear episomes and replicate in synchrony with cellular chromosomes [2 , 3] . Following replication of infected basal cells , HPV DNAs are partitioned equally to the resultant two daughter cells . While one daughter cell remains in the basal layer , the other leaves the basal layer and begins to differentiate leading to productive viral replication , late gene expression , and virion assembly in suprabasal layers [1 , 2 , 4 , 5] . These processes are regulated by the concerted action of both viral and cellular transcription factors . These factors act either directly by binding to viral sequences in the early or late promoter regions or indirectly by modulating expression of host genes that provide critical functions for the differentiation-dependent HPV life cycle [2 , 3 , 6–10] . In undifferentiated cells , the early viral promoter ( p97 in HPV 31 and 16 ) initiates transcription upstream of the E6 open reading frame ( ORF ) and directs expression of the E6 and E7 oncoproteins as well as the E1 and E2 replication factors [11–13] . E2 also acts as a repressor that auto regulates its own expression from the early promoter as part of a copy control mechanism [14–16] . Upon differentiation , the majority of viral transcription shifts to the late promoter located in the middle of the E7 ORF that directs high-level expression of E1 , E2 , E1^E4 , and E5 along with the late capsid proteins , L1 and L2 [11 , 12 , 17 , 18] . While many cellular factors regulating early viral expression in undifferentiated cells , such as Ap-1 , TEF-1 , Sp-1 , have been identified , the mechanisms and proteins that regulate late viral functions are still largely uncharacterized [17 , 19–24] . In addition to cellular transcription factors , microRNAs ( miRNAs ) also regulate viral and cellular gene expression . While HPVs do not encode their own miRNAs , they modulate the expression of a variety of cellular miRNAs [25–28] . One HPV regulated cellular miRNA is miR-145 which has been shown to be a negative regulator of the HPV31 life cycle [26] . Suppression of miR-145 expression in suprabasal epithelial cells by HPV proteins is necessary for differentiation-dependent viral DNA amplification and late gene expression . miRNAs have multiple targets in cells and miR-145 is one of the only miRNAs that has target sequences in the E1 and E2 open reading frames of HPV-31 with similar elements present in most HPV types . miR-145 also regulates the expression of several host genes including KLF4 [26] , which is a major downstream effector of the p63 pathway [29] . KLF4 is a transcription factor that is one of the four Yamanaka pluripotency factors along with c-Myc , Sox2 , and Oct4 , which are capable of transforming somatic cells into induced pluripotent stem cells ( iPS ) [30 , 31] . KLF4 is a member of the Kruppel-like family of transcription factors that regulate proliferation , differentiation as well as stemness in embryonic stem cells [32–36] . While KLF4's ability to regulate pluripotency has been demonstrated , how this is mediated is not yet well defined . In stratified normal epithelia , KLF4 is also a regulator of late differentiation markers such as loricrin , filaggrin as well as components of the epithelial granular layer that contribute to the development of the cornified envelope [29 , 37 , 38] . Genetically engineered KLF4 knock out mice develop an immature skin with a fragile cornified envelope and dysfunctional fat layer incapable of providing barrier function leading to death within 12 hours after birth [37] . Given the diverse roles of KLF4 in maintaining basal cell characteristics while also regulating differentiation , we investigated which of its activities was important for the HPV lifecycle . Our studies demonstrate that KLF4 is critical for regulating late viral events upon differentiation . KLF4 levels are increased in HPV-positive cells through post-transcriptional mechanisms involving down regulation of miR-145 along with suppression of post-translational modifications that negatively affect its activity . These changes result in altered expression of KLF4 target genes in HPV-positive cells that are distinct from those seen in normal keratinocytes but necessary for regulating the differentiation-dependent HPV life cycle .
In undifferentiated normal foreskin keratinocytes ( HFKs ) , the levels of KLF4 are low and increase substantially upon differentiation . A comparison of KLF4 levels in HFKs and HKF-31gen ( HFKs from the same host background that had been stably transfected with re-circularized HPV-31 genomes ) demonstrated higher levels in undifferentiated HFK-31gen cells and this further increased upon differentiation ( Fig 1A ) , which is consistent with previous observations [26] . Similar increases in KLF4 levels were also seen in matched sets of normal and HFK-16gen ( HPV-16 stably transfected HFKs ) ( Fig 1B ) . These same matched sets of cells were also grown as organotypic raft cultures and analyzed by immunohistochemistry for KLF4 levels and distribution . KLF4 was found to be expressed at higher levels in both HFK-31gen and HFK-16gen rafts compared to matched HFK controls ( Fig 1C and 1D ) . In both HFK-31gen and HFK-16gen rafts , KLF4 levels were highest in suprabasal layers as compared to basal cells , which showed minimal staining . In HFK rafts , KLF4 staining was less distinct and more uniformly distributed . To investigate if KLF4 had any role in regulating the HPV life cycle , lentiviruses expressing shRNAs against KLF4 were used to transiently infect CIN-612 cells , which are derived from a cervical biopsy and stably maintain HPV-31 episomes without expressing drug resistance markers[39] . For this analysis , we tested a series of KLF4 shRNAs and identified shRNAs that efficiently reduced KLF4 protein levels and pooled three of these shRNAs for further analysis . KLF4 levels were reduced following infection with pooled shKLF4 lentiviruses compared to infection with mock and shGFP controls ( Fig 2A . i ) . Cells were then examined by Southern blot analysis for stable viral replication in monolayer cultures , and for viral DNA amplification following differentiation in methylcellulose . Cells in which KLF4 levels were reduced exhibited minimal change in episome levels in undifferentiated cells and episomes failed to amplify upon differentiation . In contrast , cells transduced with shGFP lentiviruses displayed viral DNA amplification upon differentiation similar to non-transduced ( mock ) controls ( Fig 2A . ii ) . Knockdown of KLF4 also resulted in a severe impairment in late viral transcript levels as measured by northern blot analysis ( Fig 2A . iii ) . To rule out the possibility that off target effects are responsible for inhibiting viral DNA amplification , we infected CIN-612 cells individually with the three different shRNAs that target different regions of the KLF4 gene . Following infection , we observed that each of these shRNAs individually blocked HPV amplification indicating that the effects are specific for KLF4 ( S1 Fig ) . We also checked the mRNA levels of other KLF family members including KLF-3 , 5 , 14 , and other members by RNA sequencing , which remained unaltered upon pooled KLF4 shRNAs mediated silencing . To investigate the long-term effects of KLF4 knockdown , HFK-31gen cells were transduced with shRNA lentiviruses and expanded in culture containing puromycin . Reductions in KLF4 levels in these stable cell lines were confirmed by western blot analysis ( Fig 2B . i ) Cells in which KLF4 was knocked down grew at rates similar to controls , exhibited a modest reduction in the levels of episomes in undifferentiated cells and failed to amplify upon differentiation ( Fig 2B . ii ) . These results demonstrate that KLF4 provides critical functions for the HPV life cycle , primarily in differentiated cells . KLF4 is a transcription factor that binds to CACCC consensus sequences to regulate gene expression [29 , 34 , 38 , 41] . The HPV-31 Upstream Regulatory Region ( URR ) contains the viral origin of replication as well as binding sites for transcription factors that regulate viral gene expression . The HPV-31 URR also contains two KLF4 binding sites designated as R1 and R2 ( regions 1 and 2 ) . Using Chromatin Immunoprecipitation ( ChIP ) assays , we first determined that KLF4 binds to both R1 and R2 of HPV-31 URR to comparable levels in both undifferentiated and differentiated conditions and which are consistently higher than IgG controls ( Fig 3B ) . We also showed that KLF4 did not bind to GAPDH and 18srDNA sequences ( S2 Fig ) proving that KLF4 binding to the URR regions is specific . To determine if these two KLF4 binding sites in the HPV-31 URR have any functional significance for the HPV life cycle , individual point mutations were introduced into the viral genome using site directed mutagenesis and verified by whole genome sequencing . Stable cell lines were generated by transfection with wildtype , region 1 URR mutant ( R1M ) , and region 2 URR mutant ( R2M ) recircularized HPV 31 genomes , and expanded . Mutations in either region of the URR led to significantly reduced levels of viral DNA amplification as shown by Southern blot analysis ( Fig 3Ci ) along with impaired viral late gene expression as indicated by northern blot analysis ( Fig 3Cii ) upon differentiation . These results suggest that KLF4 acts a positive regulator of viral gene expression . To further confirm a role of KLF4 in activating viral expression , KLF4 expression plasmids were co-transfected with either URR-Luciferase or Lpro-luciferase ( URR containing Late promoter ) plasmids into 293T cells . KLF4 activated both URR-and Lpro-luciferase activities at low levels of transfected expression vector ( 0 . 2μg ) and the activity further increased at higher levels ( 0 . 6 μg ) of KLF4 expression vector ( 3D ) . While overexpression of KLF4 in transient assays can activate both early and late reporters , in viral infections KLF4 are low in undifferentiated cells and high in differentiated layers suggesting KLF4’s primary effect may be to regulate the late viral promoter upon differentiation . KLF4 regulates differentiation as well as proliferative capability in basal/stem-like cells [33 , 37 , 42–45] . We investigated if these two pathways were targeted in HPV-positive cells by screening for changes in cell cycle regulatory genes such as the cyclins , along with differentiation-specific markers such as loricrin . Genetically matched HFKs and HFK-31gen cells were infected with shKLF4 lentiviruses and total protein lysates were examined for levels of KLF4 , cyclins A and B1 , and loricrin by western blot analysis . These assays showed that KLF4 levels were reduced to comparable levels in both HFK-31gen cells and HFKs following transduction with lentiviruses expressing KLF4 shRNAs ( Fig 4A ) . By maintaining high levels of cyclins A and B1 , HPV-positive cells remain active in the cell cycle upon differentiation to allow for viral DNA amplification . Upon differentiation of HPV-positive cells , the levels of cyclin A were higher in comparison to those seen in HFKs . Knockdown of KLF4 with shRNAs significantly reduced cyclin A levels in differentiated HFK-31gen cells while only a modest reduction as seen in HFKs . Consistent with previous observations [40] , cyclin B1 levels were maintained at high levels in HFK-31gen cells upon differentiation but not in HFKs , and these levels were reduced when KLF4 was knocked down with shRNAs . Furthermore , the levels of the differentiation-specific protein loricrin were comparable in HFKs and HFK-31gen cells , but knockdown of KLF4 had a greater effect in reducing loricrin in HFK-31gen cells ( Fig 4A ) . These experiments indicate that KLF4 regulates genes involved in cell cycle control and differentiation in HPV-positive cells , and that these functions of KLF4 are most significant in HPV-positive cells as compared to HFKs . KLF4 is a transcription factor that regulates the expression of a number of cellular genes [29 , 34 , 38] . To investigate if this regulation was altered in HPV-positive cells , we performed global RNA-seq analyses on HFKs and HFK-31gen cells that had been transduced with control ( shGFP ) and shKLF4 lentiviruses . The use of KLF4 knockdowns for this assay was critical as it allowed us to identify which genes are transcriptional targets of KLF4 . In this analysis , cells were infected with lentiviruses expressing shRNAs and after 48 hours were cultured in methylcellulose for an additional 48 hours to induce differentiation . RNA-seq analyses were then performed on mRNAs from both undifferentiated and differentiated cells . We first confirmed that KLF4 mRNA levels were reduced in cells transduced with KLF4 shRNA expressing lentiviruses , and found reductions to similar levels in both HFKs and HFK-31gen cells compared to shGFP controls ( Fig 4B i ) . As a control we examined the levels of the related KLF5 mRNAs and saw no difference in mRNA levels between cell types after KLF4 silencing as compared to shGFP controls ( Fig 4B i ) . To identify genes regulated by KLF4 we applied a threshold criteria of at least 1 . 5 fold change following KLF4 knockdown , along with a minimum of 100 reads to eliminate genes expressed at low levels . The complete set of KLF4 target genes in HFKs and HFK-31gen cells are presented in S1 Table . Genes that were differentially expressed between these cells types upon KLF4 silencing were categorized based on their functions and represented as fold activation or suppression in HFK-31gen cells over HFKs ( Fig 4B ) . Our studies indicate that many previously identified KLF4 genes were either enhanced in activation or suppression by 2 to 5 fold in HPV-positive cells . Included among previously known KLF4 responsive genes whose expression was significantly altered in HPV-positive cells were genes associated with differentiation , factors involved in cell–cell/cell-matrix adhesion , and those associated with cornified layer formation ( Fig 4B ii , S4 Fig ) . The changes in KLF4 target gene expression between HFK and HFK-31gen control cells ( shGFP ) are shown in S3 Fig . Differentiation associated genes such as trichohyalin , filaggrin , keratin 5 , keratin 14 are all increased two to five fold in HPV-positive cells , while genes involved in cell adhesion , such as laminin alpha 3 , laminin gamma 2 , desmocolin 1 , vimentin , and collagen17 alpha 1 , are decreased to 2 to 5 fold greater compared to HFKs . In addition to previously characterized targets of KLF4 , we identified a number of uncharacterized KLF4 targets including genes associated with differentiation , cell adhesion , telomere maintenance , and DNA damage ( Fig 4B iii , S4 Fig ) . Importantly , we identified a series of novel genes ( highlighted in red in Fig 4B iii ) that are uniquely regulated by KLF4 in HPV-positive HFK-31gen cells but not in HFKs . This latter group includes trichohalin-like1 , whose expression is enhanced by over three fold , as well as genes such as vimentin , Laminins ( alpha3 , beta3 , gamma2 ) , Actinin1 , Protection of telomeres 1 , and Telomere maintenance 2 , which are suppressed by KLF4 up to 2 to 3 fold in HPV-positive cells . A subset of genes that were activated rather than suppressed by KLF4 in HFK-31gen cells as compared to HFKs is listed in S5 Fig . Finally , we identified late viral transcripts , particularly those encoding L1 , to be KLF4 targets . KLF4 knockdown had a minimal effect on early transcripts , while late transcripts encoding L1 , E4 , and E5 were significantly reduced ( Fig 4B iv ) . One additional target of KLF4 identified by our analysis in HPV-positive cells is Blimp1 . Blimp1 was recently identified as a factor that cooperates with KLF4 in EBV infected keratinocytes to regulate viral gene expression , as well as lytic replication in suprabasal layers [46] . Our RNA-seq analysis indicated that Blimp1 mRNA levels were reduced upon silencing of KLF4 in both HFKs and HFK-31gen cells , with approximately two fold greater reduction in HFK-31gen cells compared to HFKs ( Fig 4B iii ) . To confirm that Blimp1 is a KLF4 target , we examined protein lysates from KLF4-silenced HFKs and HFK-31gen cells for Blimp1 expression . Silencing KLF4 significantly reduced Blimp1 protein levels in HFK-31gen cells compared to HFKs ( Fig 5A ) reflecting RNA-seq data . We next used co-immunoprecipitation to determine if KLF4 forms protein complexes with Blimp1 . KLF4 was found to bind to Blimp1 in both differentiated HFKs and HFK-31gen cells , with significantly higher levels in the latter cells ( Fig 5B ) . It was next important to determine if Blimp1 cooperates with KLF4 in the activation of HPV promoters . For this analysis , HPV late promoter-luciferase reporters were co-transfected together with KLF4 and Blimp1 expression vectors in various ratios and screened for levels of luciferase expression . These late promoter luciferase reporters contain nucleotides 7045 through 891 of HPV31 genome , which includes the major start site for the late promoter ( p742 ) as well as the complete URR . While Blimp1 expression alone had a modest effect on the activation , when co-transfected with KLF4 , increased activation of luciferase expression in a concentration-dependent manner was seen ( Fig 5C ) . We conclude that Blimp1 is a KLF4 target that forms protein complexes with KLF4 to additively activate HPV promoters . To verify if KLF4-Blimp1 association is important for the activation of HPV late gene expression , we conducted ChIP assays using HPV31 keratinocytes in which KLF4 was stably depleted with lentiviral shRNAs to examine Blimp1 binding to the KLF4 binding site R2 in the URR . Both KLF4 and Blimp1 bound to R2 in both undifferentiated and differentiated conditions . Upon stable knockdown of KLF4 with shRNAs , Blimp1 binding to R2 was significantly reduced both in undifferentiated and differentiated conditions ( Fig 5D ) . These results indicate that KLF4 is required for Blimp1’s ability to bind efficiently to the HPV31 promoter . The data described above indicate that KLF4 has enhanced transcriptional activation and suppression abilities in HPV-positive cells compared to normal keratinocytes , suggesting it may provide different functions in these cells . We next investigated the effects of silencing KLF4 on cell growth and differentiation capabilities in HFKs and HFK-31gen cells . Cells were infected with shRNA lentiviruses , and 48 hours after transduction seeded onto collagen plugs and grown as organotypic raft cultures . Reductions in KLF4 protein levels were comparable in both sets of cells . Transient silencing of KLF4 in HFKs abolished the ability of cells to form stratified cultures in organotypic rafts , whereas KLF4-depleted HFK-31gen cells formed rafts with stratified layers but with morphologically altered cornified envelopes ( Fig 6A ) . Similar results were seen in three independent experiments . We surmised that the inability of KLF4-depleted HFKs to form rafts might be due to a loss in stem cell proliferative capacity . To test this hypothesis , HFKs and HFK-31gen cells were assessed for their colony forming abilities using a holoclone assay [47] . In this assay , keratinocytes are seeded sparsely ( 100–500 cells ) in 100 mm dishes , thereby forcing them to undergo multiple cell divisions . In such stringent conditions , only cells with extensive proliferation capacity such as stem cells and early stage transit-amplifying cells can form viable colonies , while late stage transit-amplifying cells produce abortive colonies due to their limited proliferative capacity . KLF4-depleted HFK-31gen cells formed colonies comparable to control cells , while KLF4-depleted HFKs completely abolished colony-forming abilities indicating a loss in stem cell-like proliferative activity ( Fig 6B ) . In a similar manner , it was not possible to generate stable cell lines of HFKs following infection with lentiviruses expressing KLF4 shRNAs , whereas both HFK-31gen and HFK-16gen cells readily formed lines with stable KLF4 depletion ( Fig 6C , S6 Fig ) . HFK-KLF4 knockdown cells continued to proliferate for up to one passage following transduction but failed to expand after passaging . In contrast , HFK-31gen KLF4 knockdowns could be readily passaged and grew at rates comparable to parental cells . Reductions in KLF4 levels were confirmed in the HPV-positive cells after multiple passages by western blot analysis ( Fig 6C . i ) . As observed in transient silencing experiments , KLF4-silenced HFK-31gen cells formed stratified cultures in organotypic rafts , but with morphologically altered cornified layers ( Fig 6C . ii ) . Raft sections were further analyzed for KLF4 , loricrin , and E1^E4 levels by immunohistochemistry . Loricrin and E1^E4 were used as a read out for the changes in cornified envelope composition and viral late gene expression respectively . KLF4 levels were substantially reduced in shKLF4 rafts compared to control rafts , and both loricrin and E1^E4 were not detected in KLF4-silenced rafts ( Fig 6D ) . Similar effects were seen when KLF4 levels were silenced in HFK-16gen cells ( S6A Fig ) , with morphologically altered cornified layers in organotypic raft cultures ( S6B Fig ) . These results indicate that KLF4 may have similar functions in modulating cell cycle and differentiation capabilities of multiple high-risk HPV types . Furthermore , our studies indicate that KLF4 regulates the expression of genes in undifferentiated HFKs that are critical for proliferation , while similar effects are not seen in HPV-positive cells . The finding that KLF4 has different functions in HFKs and HPV-positive cells led us to investigate a potential mechanism , beyond changes in KLF4 expression levels , that could contribute to these effects . KLF4 undergoes post-translational modifications such as phosphorylation , sumoylation , and acetylation , which in turn determine its binding partners and ability to activate or suppress gene expression [43 , 48–51] . KLF4 is phosphorylated at serine 245 and this results in suppression of its transcriptional activities . We therefore investigated if there were any differences in phosphorylation between HFKs and HPV-positive cells as determined by western blot analysis . In undifferentiated HFK-31gen cells , the levels of phospho-Ser-245 KLF4 levels were substantially reduced as compared to matched HFKs . Furthermore , upon differentiation , p-KLF4 levels were also lower in HFK-31gen cells than HFKs ( Fig 7A ) . Next , we used immunofluorescence analysis of HFKs and HFK-31gen cells grown on coverslips in either normal or high calcium media for 72 hours to screen for effects on the levels of p-KLF4 . Less intense staining for p-KLF4 was observed in HFK-31gen cells as compared to HFKs in either undifferentiated or differentiated conditions ( Fig 7B ) . Calcium-induced differentiation was used for these analyses as this method does not distort cell morphologies , as is seen with methylcellulose-induced differentiation . Finally , similar distributions and levels of p-KLF4 were detected in organotypic raft cultures of these same cells ( Fig 7C ) . Interestingly , the band we identified by western analysis as phospho-Ser-245 KLF4 migrated at approximately 75 kDa , which is approximately 20kDa higher than the unmodified KLF4 band . While phosphorylation alone does not induce such large shifts in mobility , phosphorylation coupled with sumoylation could induce such a shift . Phosphorylation is often a prerequisite for sumoylation and KLF4 has been reported to be sumoylated under some conditions . To provide support that this band was actually a phosphorylated form of KLF4 , lysates from HFKs and HFK-31gen cells grown in both undifferentiated and high calcium conditions were run in duplicate on a single polyacrylamide gel . After electrophoresis , the gel was transferred to a PVDF membrane and the membrane was cut into two halves , which were either incubated with lambda phosphatase in buffer or buffer alone , for one hour . The membranes were then blocked and processed as usual for western analysis . Treatment with lambda phosphatase specifically reduced the levels of the 75kDa band , without altering the surrounding non-specific bands , demonstrating that this band is a phosphorylated protein ( Fig 7D ) . To further demonstrate that the 75kDa band is specific to KLF4 , protein lysates of HFKs transiently infected with lentiviruses targeting KLF4 were screened by western blot with antibodies against unmodified KLF4 and phospho-ser-245 KLF4 . KLF4-specific shRNAs reduced levels of both unmodified KLF4 and the observed 75kDa phospho-band ( Fig 7E ) in comparison to control samples . These experiments verified that KLF4 is phosphorylated at Serine 245 and runs at approximately 75kDa . We next investigated if the 75kDa band is a sumoylated form of KLF4 . Protein lysates from matched HFKs and HFK-31gen cells grown in either low or high calcium media were immunoprecipitated with total KLF4 antibody and then screened for the presence of Sumo-1 by western analysis . KLF4 pull down samples yielded a 75kDa band , which was absent in IgG pull down samples . In addition , the intensity of the band was lower in HFK-31gen differentiated samples than HFKs , consistent with changes observed in the phosphorylation experiments ( Fig 8A ) . To validate that the 75kDa band is sumoylated , we performed immunoprecipitation with either mock or sumo protease- treated protein lysates of differentiated HFKs . Immunoprecipitation of mock-treated samples with KLF4 antibodies resulted in the appearance of the 75kDa band , but the levels of this band were significantly reduced in sumo protease-treated KLF4 immunoprecipitated samples ( Fig 8B ) . These experiments demonstrate that the 75kDa KLF4 band is both sumoylated and phosphorylated , which together explain the observed 20kDa shift in weight . We next wanted to confirm that the sumoylated band is indeed KLF4 and not any other KLF4 binding protein undergoing sumoylation . For this analysis , we performed immunoprecipitation assays using a mouse secondary KLF4 antibody to pull down protein complexes and a rabbit secondary KLF4 antibody for the western blot analysis . KLF4-KLF4 immunoprecipitation experiments demonstrated an enrichment of unmodified 55 kDa KLF4 along with its isoforms compared to igG control and more importantly showed an enrichment of a 75kDa KLF4 band ( Fig 8C ) . These results indicate that the levels of phosphorylated and sumoylated KLF4 are significantly reduced in HPV-positive cells . We next investigated which viral proteins are responsible for the increased levels of KLF4 and suppression of its post-translational modifications . For this analysis , we used retroviruses expressing either HPV-31 E6 or E7 to infect HFKs and to isolate stable cell lines . Lysates from both undifferentiated and differentiated cells were then screened by western analysis for total levels of KLF4 as well as phospho-Ser 245 KLF4 . Total KLF4 levels were elevated in both E6-and E7-expressing keratinocytes under undifferentiated or differentiated conditions , with the largest effect mediated by E7 ( Fig 9A ) . In contrast , the levels of phospho-Ser245 KLF4 levels were significantly reduced in E6-expressing cells in either undifferentiated or differentiated conditions ( Fig 9 ) . p53 levels were used as a surrogate marker for E6 and E7 expression , as p53 levels were significantly reduced in E6-expressing cells but increased in E7-cells [52] . We then wanted to check which one of E6’s functions is essential for the phosphorylation of KLF4 . For this analysis , we generated stable cell lines of HFKs expressing previously characterized E6 mutants including E6 G134V ( lack CBP/p300 binding ) [53] , E6 I128T ( cannot degrade p53 ) , and E6 L37S ( cannot bind as well as degrade p53 ) [54] using retroviral vectors and checked for phospho-ser-245 KLF4 levels by western blot analysis . While wildtype E6 reduced the levels of phospho-ser-245 KLF4 , none of the E6 mutants exhibited any alterations in phospho-ser-245 KLF4 levels ( Fig 9B ) indicating that E6’s ability to bind to p53 and p300 as well as degradation of p53 is essential for the reduction of phosphorylation of KLF4 . To understand the mechanisms responsible for E7 mediated increases in KLF4 levels , we investigated a possible role for NFκB . Previously , we showed that E7 decreases miR-145 levels [26] and additional studies demonstrated that E7 reduces NFκB activity [55] . We observed a corresponding decrease in NFκB activity in HPV-positive cells ( S7A Fig ) and determined that the promoter of miR-145 contains two p65 ( NFκB active subunit ) binding sites . Using transient reporter assays , we further demonstrated that increasing levels of p65 expression vector activated miR-145 promoter activity in a dose-dependent manner ( S7B Fig ) . This indicates that E7 may contribute to increased levels of KLF4 through NFκB inactivation , while E6 is primarily responsible for suppression of KLF4 phosphorylation .
High-risk human papillomaviruses infect keratinocytes and modulate both the expression and activity of host factors to control their differentiation-dependent productive life cycles [7 , 8 , 26 , 56 , 57] . Our studies identify KLF4 as a critical host transcription factor that regulates the HPV life cycle by controlling viral gene expression , cellular differentiation , and cell cycle capabilities in suprabasal epithelial layers . KLF4 is one of four Yamanaka pluripotency factors , which along with Oct4 , SOX2 and c-Myc are able to reprogram somatic cells into pluripotent stem cells by targeting their proliferation and differentiation capabilities . In HPV-positive cells , E7 and E6 proteins regulate KLF4 levels and activity , respectively , through post-transcriptional and post-translational mechanisms to induce a number of activities distinct from those seen in normal epithelia . E7 controls KLF4 levels post-transcriptionally by suppressing the expression of a cellular microRNA , miR-145 , which targets KLF4 transcripts , leading to increased levels of KLF4 proteins . The E6 protein acts through post-translational mechanisms to regulate KLF4 phosphorylation and sumoylation , which negatively affect KLF4 functions . These virally induced changes result in differential activation or suppression of previously characterized transcriptional targets of KLF4 as well as expression of novel genes , all of which contribute to cell proliferation , stratification and differentiation during the HPV life cycle . The HPV life cycle is closely associated with differentiation of the infected host keratinocyte . Following infection of undifferentiated stem-like basal cells , viral genomes are established as low copy episomes that replicate coordinately with cellular chromosomes in S phase [2] . Upon differentiation , late viral events such as viral DNA amplification , late gene expression , and virion packaging are induced in differentiated suprabasal cells following transition into S/G2[58] . Since KLF4 is important for regulating stem-cell like proliferative abilities along with controlling differentiation , we investigated what role , if any , KLF4 played in either of these processes during the HPV life cycle . In HPV-positive stratified epithelia , KLF4 is primarily expressed in suprabasal layers with only low levels in basal cells . Our studies demonstrate that KLF4 regulates late viral events in two ways in suprabasal cells . One function of KLF4 is to directly regulate expression of HPV late genes . Our studies show that KLF4 binds to two sites in the HPV-31 URR and that mutation of these sites reduces differentiation-dependent late gene expression as well as significantly impairs genome amplification . Similarly , silencing KLF4 with shRNAs impairs both viral late gene expression and viral DNA amplification . RNA seq analyses confirmed reductions in viral transcripts encoding L1 , L2 , E4 , and E5 in HPV-positive cells in which KLF4 was knocked down with shRNAs , with minimal effects on early viral gene transcripts . In stable knockdown cultures , we observed a modest reduction in the levels of viral episomes , a significant impairment in genome amplification upon differentiation , and reduced late gene expression . Together , these data demonstrate a critical role for KLF4 in regulating late viral functions . Interestingly , one novel target of KLF4 is Blimp1 and it forms protein complexes with KLF4 to cooperatively regulate late viral transcription . The ability of KLF4 to act coordinately with Blimp1 to activate viral promoters has recently been shown to control EBV lytic replication[46] . Our studies along with those of Nawandar et al . demonstrate that KLF4 is a critical regulator of the differentiation-dependent life cycles of two oncogenic viruses , EBV and HPV . Whether KLF4 is post-transcriptionally regulated by EBV proteins , and if KLF4 controls Blimp1 expression in EBV-infected cells remains unclear . In addition to a direct effect in regulating viral gene expression , RNA-seq analyses demonstrated that KLF4 positively regulates the expression of genes that control differentiation in the suprabasal cornified , granular , and spinous cell layers of HPV-positive keratinocytes . At the same time , KLF4 negatively regulates the expression of basal cell markers including cell-matrix adhesion genes , keratinocyte growth factor receptors , and telomere maintenance proteins . Not only were KLF4 functions enhanced in HPV-keratinocytes , but also a number of KLF4 targets unique to HPV-positive cells were identified including vimentin , integrin beta4 , laminins , protection of telomeres1 , and Rad51D . Our studies show that in normal keratinocytes , KLF4 is required for the proliferative ability of basal cells and for the expression of differentiation genes following growth in raft cultures . Transient knockdown of KLF4 in normal keratinocytes reduced expression of both proliferation as well as differentiation associated genes , leading to a failure to grow beyond a single passage . Furthermore , these cells did not form colonies in holoclone assays nor grow in raft cultures . RNA-seq analysis demonstrated that KLF4 positively regulates expression of the cell-matrix adhesion gene Actinin1 , which is critical for proliferation of basal cells as well as differentiation . In addition , we observed that KLF4 positively regulates expression of cytokeratin 14 , which heterodimerizes with cytokeratin 5 to form the cytoskeleton of basal epithelial cells[59] . These changes may be responsible for the impaired proliferative ability of HFK stable KLF4 knockdowns , or additional effects may be required . Regardless , these functions are not critical in HPV-positive cells , which continued to propagate even after stable KLF4 knockdown . This difference could be the result of altered KLF4 and target gene expression in HPV- positive cells , or due to the actions of E6 and E7 in stimulating proliferation . In HPV-positive cells , knockdown of KLF4 had minimal effects on the proliferative ability of undifferentiated cells , and stable knockdowns were readily established . The levels of Actinin1 and cytokeratin 14 are suppressed in HPV-positive cells unlike normal keratinocytes , which may contribute to their ability to survive and proliferate in long-term stable assays . In HPV-positive cells , KLF4 knockdown had its most significant effect upon differentiation: cells in which KLF4 was knocked down retained the ability to stratify in raft cultures but were significantly altered in the expression of differentiation specific genes and cell cycle regulators . The levels of loricrin and filaggrin , two differentiation genes associated with late viral functions , were significantly reduced . KLF4 silencing also severely affected the expression of many late cornified envelope ( LCE ) genes , leading to the formation of morphologically altered cornified layers in raft cultures . In addition , HPV-positive suprabasal cells were no longer active in the cell cycle , as evidenced by an absence of cyclin A or B1 proteins . While the levels of cyclin A and B1 proteins are reduced in KLF4 knockdowns , transcript levels are unchanged , indicating they are not direct transcription targets of KLF4 . This indicates that KLF4 acts to regulate cyclin levels at the post-transcriptional level . While KLF4 silenced HPV keratinocytes formed stratified cultures in rafts , late viral functions were significantly impaired as evidenced by the loss of E1^E4 proteins . These analyses demonstrate that KLF4 has different critical functions in normal and HPV-positive keratinocytes . Our studies show that HPV proteins regulate KLF4 levels and activities through post-transcriptional and post-translational changes . KLF4 levels are increased in HPV- positive cells , in part due to decreased levels of a cellular miRNA , miR-145 , which we have previously shown to have recognition sites in the 3’UTR of KLF4 mRNA[26] . In addition , our previous studies suggest that E7 acts through NFκB to suppress miR-145 levels to increase KLF4 protein levels . Additional changes in KLF4 functions are likely the result of suppression of post-translational modifications such as sumoylation and phosphorylation . Sumoylation of KLF4 negatively regulates its activity and is dependent upon phosphorylation . KLF4 phosphorylation also negatively regulates KLF4 function facilitating its ubiquitination and subsequent enhanced degradation [43] . As discussed above , E7 suppresses miR-145 expression while E6 suppresses KLF4 phosphorylation and sumoylation . Erk1 and Erk2 kinases , which phosphorylate KLF4 to suppress its ability to induce proliferation and self-renewal in embryonic stem cells , may be potential targets of E6 . Consistent with our observations , E6 has been shown increase the turnover of UBC9 leading to global reductions in sumoylation[60] , providing a potential mechanism explaining its effect on KLF4 . In addition to regulating protein stability , post-translational modifications can also regulate interactions with protein binding partners , cellular localization , and transactivation ability of factors . Suppression of KLF4 sumoylation or phosphorylation could induce the binding of different factors to KLF4 , leading to activation of novel targets seen in HPV-positive cells . In HPV-positive cells , KLF4 has a number of distinct functions that cannot be explained by increased levels alone . In particular , knockdown of KLF4 in normal keratinocytes blocks stratification while HPV-positive cells readily grow in raft cultures that have morphologically altered suprabasal layers . Furthermore , KLF4 knockdown in HPV-positive cells does not affect growth of basal cells in holoclone assays while this is inhibited in KLF4 knockdowns of HFKs . Finally , while many transcriptional targets of KLF4 are shared between HFKs and HPV-positive cells , there are a number of genes that are only repressed by KLF4 in HPV-positive cells while other genes that are activated by KLF4 in HFKs are repressed in HPV-positive cells . These differences cannot be easily explained by increases in KLF4 levels alone and implicate post-translational mechanisms as also being important . Human cancer viruses regulate their productive life cycles by modulating the activities of cellular factors . Our work shows that high-risk human papillomaviruses alter the activities and expression of a critical cellular transcription factor , KLF4 , through post-transcriptional and post-translational mechanisms , and thereby regulate differentiation-dependent late viral events . KLF4 has been shown to be a critical regulator of lytic replication of another oncogenic human cancer virus , EBV , but whether EBV alters KLF4 in a similar fashion is unclear . Furthermore , we show that the KLF4 transcriptome is altered in HPV- positive cells but whether similar changes occur in EBV cells remains to be determined . In summary , our work provides novel insight into mechanisms by which HPVs regulate host transcription factors to promote viral amplification and infection , and identifies a common pathway shared among human cancer viruses .
The human keratinocytes used in this study were obtained from discarded foreskin circumcisions from anonymous donors by the Keratinocyte Core in the Northwestern University Skin Disease Research Center ( SDRC ) and are not classified as human subjects research . These specimens were not specifically collected for this study and lack all identifiers . Human Foreskin Keratinocytes ( HFKs ) were isolated from the foreskin tissues of neonates obtained from anonymous donors by the Keratinocyte Core in the Northwestern University Skin Disease Research Center ( SDRC ) and grown as described previously[8] . Recircularized HPV-31 genomes were transfected into HFKs along with antibiotic resistance plasmid pSV2 Neo and stably selected with G418 to obtain HFK-31gen cells . HPV-16 Cre genome containing plasmids were transfected along with Cre and pSV2 Neo plasmids into HFKs and stably selected with G418 to obtain HFK-16gen cells . CIN-612 cells are HPV31-positive cervical cells , which were obtained from biopsy of an early stage cervical cancer patient and originally expanded by the Laimins laboratory ( 39 ) . HFK-E6 and HFK-E7 cell lines were derived by retroviral transduction of HPV31 E6 and E7 into HFKs , followed by selection with G418 . All the experiments were conducted with at least three genetically matched HFKs and derived cell lines . Keratinocytes were seeded onto collagen plugs containing fibroblasts placed on metal grids over growth media , which creates air-liquid interface and grown as organotypic raft cultures for 13 days as described before[61 , 62] . Pre-validated lentiviral shRNAs were purchased from Open Biosystems . Five different shRNAs were individually tested for silencing and pooled shRNAs were used for silencing experiments . Forty-eight hours post transduction; cells were induced to differentiate for further 48 hours in 1 . 5% methylcellulose . Stably silenced cell lines were made by selecting with puromycin after transduction . Reduction in protein levels was confirmed by western analysis . Total DNA was isolated from cells as described before [8] . DNA was electrophoresed in 0 . 8% agarose gel and transferred to membranes . Membranes were blocked with salmon sperm DNA containing blocking buffer , and probed with p32-HPV31 probe . After a series of washes with various stringency buffers , membranes were analyzed to autoradiography [63] . Total RNA was isolated from cells using STAT60 reagent according to manufacturer’s instructions ( Tel-Test Inc . ) . RNA was electrophoresed in 0 . 8% agarose gel with formaldehyde and transferred to membranes . Membranes were blocked , and probed with p32-HPV31 probe . After a series of washes with various stringency buffers , membranes were examined by autoradiography [63] . Cells were grown on treated coverslips with regular E-media for undifferentiated conditions . For differentiation , cells were pretreated with low calcium media for 24hours followed with high calcium for 72 hours . Cells were then fixed with 4% paraformaldehyde , permeabilized with Triton-X100 , blocked with normal goat serum and probed overnight for p-KLF4 antibody . After 3 washes with PBS , Alexaflour rabbit secondary antibody was added for one hour and washed thrice in PBS . After a brief incubation with DAPI , coverslips were washed in PBS multiple times before mounting onto glass slides with gelvatol for analysis . Raft sections were de-paraffinized overnight at 55°C followed by a series of washes in xylene ( 3 ) , ethanol ( 3 ) , 70% ethanol and permeabilized with triton x-100 . Antigen retrieval was performed in citrate buffer at 95°C for 20 mins . Sections were blocked with normal goat serum ( NGS ) and probed for 1:200 primary antibodies ( KLF4 CST 4038S , pKLF4 Thermo PA5-13081 , Loricrin SC-133757 , and E1^E4 , a kind gift from Sally Roberts ) in NGS overnight . After three washes in PBS , Alexaflour secondary antibodies in NGS were added for 1 hour , followed by three washes in PBS . After a brief incubation with DAPI , slides were washed in PBS before mounting coverslips with gelvatol . Primers for mutations were designed using the Agilent primer design program . Mutagenesis PCRs were performed according to the manufacturer’s instructions ( Agilient 200523 ) . The primer sequences are as follows . URR-R1M-Forward: GTGTTGTGTATGTTGTCCTTATATACACACTATTAGTAACATACTATTACTATTTTA . Reverse: TAAAATAGTAATAGTATGTTACTAATAGTGTGTATATAAGGACAACATACACAACAC . URR-R2M-Forward: CCATAGTAAAAGTTGTACACACGGTCCGTTTTTTGCAACTA . Reverse: TAGTTGCAAAAAACGGACCGTGTGTACAACTTTTACTATGG . ChIP assays were performed as described previously[64] . Primers: KLF4-URR-R1 , For: ATGTGTATGTGCTTGTGCTG , Rev: TGACTATTGGGAGGAGCAGG . KLF4-URR-R2 , For: ACTTGTTCCTACTTGTTCCTGC , Rev: GCATCAGCATAGTTGTACTAGC . Cell lysates were harvested using RIPA buffer with SDS containing protease , phosphatase , and desumoylation inhibitors . Two hundred micrograms of protein sample was immunoprecipitated with protein A/G agarose beads ( SC-2003 ) and KLF4 antibody ( SC-393462 ) overnight . Beads were spun down at 7500 rpm for 2 minutes at 4°C , followed by three cycles of washes with RIPA buffer and spins . Beads were then boiled at 95°C for 5 minutes with Lamelli sample buffer , cooled down , and spun in microcentrifuge . The supernatants were loaded into 8% polyacrylamide gels and western blotting analyses were performed with the following antibodies: Sumo-1 ( CST 4930S ) , and Blimp1 ( CST 9115S ) . Keratinocytes were seeded sparsely ( 100–500 cells ) with mitomycin treated NIH-3T3 cells in 100 mm dishes and were grown in E-media with twice the amount of Epidermal Growth Factor ( 10ng/mL ) for 3 weeks . The colonies were stained with crystal violet in methanol for 30 minutes , followed by series of water washes . RNA-sequencing was performed using Illumina HiSeq 2500 NGS platform . Sequencing data were used as input to CRI Illumina RNA-seq pipeline for quality control assessment of raw sequencing data , reads mapping , post-alignment QC , expression quantification , and DEGs identification . The quality of raw sequencing reads was assessed using FastQC v0 . 11 . 2 , and the post-alignment QC was evaluated with RSeQC v2 . 3 . 9[65] and Picard tools v1 . 117 . Reads were mapped to 1 ) . UCSC human genome ( hg19 ) obtained from GATK resource bundle v2 . 8 using TopHat v2 . 0 . 13 [66] guided with UCSC gene annotation model ( hg19 ) obtained from Illumina iGenomes , and 2 ) HPV31 virus genome with GenBank accession number J04353 . 1 . Gene transcripts were assembled and quantified on human and HPV31 genome separately using Cufflinks v2 . 2 . 1 [67] with the hg19 UCSC gene model annotation and HPV31 RefSeq gene annotation as a guide respectively for transcript assembly and bias detection/correction . Sample-based assemblies were merged together using Cuffmerge wrapped in Cufflinks v2 . 2 . 1 before quantification of transcripts using Cufflinks wrapped method Cuffnorm and count-based method featureCounts[68] . DEGs were identified between 4 various group comparisons using Cuffdiff wrapped in Cufflinks v2 . 2 . 1 . During the entire analysis , R ( R Core Team , 2014 ) were used to assist in the exploration and summarization of the analysis results . KLF4 expression vector was purchased from Addgene . Blimp1 expression vector was a kind gift from Shannon Kenney . pBR322-HPV31gen , pUC-HPV16gen , URR-pro , pLxSn 31-E6 , pLxSn 31-E7 , and Lpro-luc were described before [18 , 69 , 70] .
|
Viruses that induce persistent infections often alter the expression and activities of cellular transcription factors to regulate their productive life cycles . Human papillomaviruses ( HPVs ) are epithelial tropic viruses that link their productive life cycles to the differentiation of infected host keratinocytes . Our studies show that KLF-4 , originally characterized as a pluripotency factor , binds HPV-31 promoters activating viral transcription as well as modulates host cell differentiation and cell cycle progression . KLF4 levels and activity are enhanced in HPV-positive cells by E6 and E7 mediated post-transcriptional and post-translational mechanisms resulting in altered target gene expression and biological functions from that seen in normal keratinocytes . Importantly , silencing KLF4 hinders viral genome amplification and late gene expression . Along with its recently identified role in Epstein Barr Virus reactivation during differentiation , our studies demonstrate the importance of KLF4 in the life cycles of multiple human cancer viruses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"phosphorylation",
"urology",
"keratinocytes",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"cell",
"differentiation",
"epithelial",
"cells",
"viruses",
"developmental",
"biology",
"immunoprecipitation",
"dna",
"viruses",
"sexually",
"transmitted",
"diseases",
"hpv-31",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"papillomaviruses",
"human",
"papillomavirus",
"infection",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"life",
"cycles",
"biological",
"tissue",
"precipitation",
"techniques",
"biochemistry",
"cell",
"biology",
"anatomy",
"post-translational",
"modification",
"viral",
"pathogens",
"genitourinary",
"infections",
"genetics",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"viral",
"diseases",
"organisms"
] |
2016
|
Post-Transcriptional Regulation of KLF4 by High-Risk Human Papillomaviruses Is Necessary for the Differentiation-Dependent Viral Life Cycle
|
Ebola Virus Disease ( EVD ) is a condition with high fatality . Though the disease is deadly , taking precautions to reduce contact with infected people and their secretions can prevent cross- infection . In the 2014 EVD outbreak , socio-cultural factors were identified to be responsible for the spread of the disease in the three most affected countries in West Africa . In this light , we undertook this study to identify socio-cultural factors that may influence the prevention and containment of EVD in Ghana and ways to address such practices . We conducted a descriptive qualitative study in five regions in Ghana . Twenty-five focus group discussions ( 5 in each region ) with community members ( 4 in each region ) and nurses ( 1 in each region ) were conducted . In addition , forty ( 40 ) in-depth interviews were conducted with various stakeholders and opinion leaders; eight in each region . All interviews were recorded using a digital voice recorder and transcribed . With the aid of Nvivo 10 for windows , we analyzed the data using framework analysis . We found that socio-cultural practices , such as care of the body of dead and burial practices , widowhood rites and anointing children with water used to rinse the dead , were common . These practices require individuals coming into direct contact with either the dead or items used to take care of the dead . Social norms also require frequent handshakes in all social gatherings such as funeral , and religious congregations . We also found that self-medication ( using herbs and orthodox medications ) was a common practice . People use both biomedical and non-orthodox health outlets either simultaneously or in sequence in times of ill-health . The study concludes that high risk socio-cultural practices were common among Ghanaians and generally perceived as indispensable . These high risk practices may hinder containment efforts in the event of an outbreak . Community leaders should be engaged in any social mobilization to modify these practices as part of preparation efforts .
Ebola Virus Disease ( EVD ) is a condition with high infectivity and case fatality rate . The case fatality rate of the 2014 EVD epidemic was initially estimated as 70% at the beginning of the epidemic [1 , 2] . However , the case fatality declined with the establishment of treatment centres and provision of logistics to care for infected individuals . This epidemic which started in a rural community in Guinea was declared a public health emergency of international concern by the World Health Organization as the outbreak spread across countries . Statistics show that the number of confirmed , probable and suspected cases kept increasing even with several interventions to contain the epidemic . The number of cases and deaths reported across all the eight previously affected countries ( Guinea , Liberia , Mali , Sierra Leone , Nigeria , Senegal , Spain and the United States of America ) were 28 , 639 and 11 , 316 respectively [3] , with an estimated case fatality rate ( based on these figures ) of 40% . Though EVD is deadly , measures such as washing of hands with soap and water , using hand sanitizers , and avoiding contacts with people who are infected with the condition and their body fluids can reduce cross-infection and help contain spread of the disease . This is because the virus is present in the bodily fluids of infected individuals and all items contaminated with such bodily fluids are potential sources of transmission [4–6] . During EVD outbreaks , health workers and family members who take care of EVD patients , mourners and undertakers who come into physical contact with the corpse as part of rites of passage ( burial ceremonies ) have a greater risk of infection through direct mode of transmission [6] . One can also get infected with EVD through indirect means such as handling contaminated clothes , materials and equipment . Therefore , individuals who handle or physically touch these items are also at high-risk of infection . As part of containment efforts , some of these high-risk practices including washing and handling of the corpse were banned in the past during EVD outbreaks [7 , 8] . Attending funerals have also been reported to facilitate transmission to distant areas . This is because funeral attendants get infected and carry the infection to their communities and this can turn small outbreak into a major epidemic [9] . Many studies have identified economic and sociocultural factors as key hindrances to timely identification of cases and implementing control efforts in the affected regions [10–12] . Adjusting such cultural practices in the wake of an EVD outbreak is therefore critical in controlling transmission and communities that have not transformed from these high risk practices have been reported to face challenges containing an outbreak [13] . In response to the West African EVD epidemic , the Government of Ghana established an inter-ministerial committee which was chaired by the Minister of Health to lead EVD preparation and containment plan . Through their leadership and collaboration with development partners , a national preparedness and response plan was developed in August 2014 . The plan itemized Ghana’s preparation in five key areas: planning and coordination of all EVD-related activities; passive and active surveillance , situation monitoring and assessment across the country; management of suspected and confirmed cases; social mobilization and risk communication; and provision of logistics , security and financial resources [14] . In line with the EVD response plan , Ghana Health Service also intensified social mobilization and risk communication across the country . Besides , national , regional and district EVD response teams were constituted along the decentralized governance system in Ghana to implement the plan . Ghana Health Service also established EVD treatment centres across the country , organized training for frontline health workers and provided Personal Protective Equipment to health facilities [14] . Screening of international travellers at entry and exit points in Ghana was also started in Ghana . Although Ghana did not record a confirmed EVD case , the country is at risk through cross-border travel of people from affected countries in West Africa ( including EVD survivors ) , or through de novo animal-human transmission . This research was therefore designed to explore cultural and social practices that may influence the prevention and containment of EVD and how to address these factors in Ghana .
The Institutional Review Board of Noguchi Memorial Institute for Medical Research ( NMIMR ) in Ghana reviewed and approved the proposal to conduct this study . The reference number for this protocol was NMIMR-IRB CPN 057/14-15 . Informed written consent were solicited and obtained from all study participants before the study and data collected were anonymised . We adopted a qualitative study design and method of data collection . This strategy was employed because the purpose of our study was to gain deeper understanding of the relevant socio-cultural factors that may hinder EVD containment from both community members and health workers . We combined two qualitative study designs; phenomenology and narrative . Phenomenology approach to qualitative research is used to study participants’ perceptions , feelings , and lived experiences of a particular phenomenon in the community and how these knowledges can affect the person viewpoints concerning a specific situation [15] . Narrative research on the other hand allows study participants to describe their experiences in the community [16] . We conducted this research in the Republic of Ghana which is located on West Africa's Gulf of Guinea . Ghana shares boundary with three neighbouring countries; Côte d'Ivoire , Togo and Burkina Faso to the west , east , and north respectively . Located at the southern border of Ghana is the Atlantic Ocean and the Gulf of Guinea . The population of Ghana is estimated as 24 , 658 , 823 with yearly growth rate of 2 . 4% [17] . The nation is divided into ten decentralized administrative regions . English is the official language of Ghana but there are about 75 local languages . Despite the fact that we have 10 regions in Ghana , this study was conducted in five regions with a combined population of 15 , 764 , 171 . These regions include: Western , Greater Accra , Volta , Ashanti and Northern . The two most populous regions in Ghana are the Greater Accra region ( GAR ) and Ashanti region ( AR ) with populations of 4 , 010 , 054 and 4 , 780 , 380 respectively . The population densities of Greater Accra region is 1 , 236 per square km whilst that of Ashanti region is 196 per square km [17] . The Volta ( VR ) and Western ( WR ) regions have a population of 2 , 118 , 255 and 2 , 376 , 021 respectively [17] . Furthermore , the Northern Region ( NR ) has a population of 2 , 479 , 461 [17] . We considered population density , rural-urban factors and entry and exit points in Ghana as the criteria in selecting the five regions . Animal-to-human and human-to-human transmissions of EVD is faster in areas of high population density , therefore the two regions ( GAR and AR ) with high population densities were selected for the study [18] . The Greater Accra region which is the national capital of Ghana is also home to the main international airport and seaport . The Volta and Western regions share boundary with Togo and Cote D’ Ivoire respectively . The Western region also has seaport and therefore serves as point of entry for sea travellers . The Northern Region is the biggest region located in the northern Ghana and also shares boundary with neighboring Cote D’Ivoire . Fig 1 shows the regions that were selected for the study and their boundaries with neighbouring countries which serve as entry/exit points to/from Ghana . We designed semi-structured IDIs , and FGDs guides for the data collection . The guides were translated from English to the local language before the field work began . Back-to-back translation strategy was used to ensure that the versions in English were the same as those in the local languages . The FGDs and IDIs guides covered areas such as funeral practices , religious practices and other social activities in the community that can facilitate EVD transmission . We recruited 15 graduate research assistants , trained them after which they were deployed to the regions ( 3 to each region ) to collect the data between 10th February and 3rd March , 2015 . We included both males and females aged ≥ 18 years in line Ghana’s constitution for informed consent [19] . Purposive sampling was used to select participants who could provide in-depth information [20] on socio-cultural issues in the community . Research assistants entered each community through community leaders . The purpose of the study was explained to the community leaders who assisted the research assistants to select people from different sections of the community to constitute a group for the discussions . In selecting opinions leaders and chiefs , we took into consideration the ethnic diversity of the various communities . This was done to ensure that the data collected reflected views from various ethnic groups in the society . Members of EVD response teams were also purposively selected as they were supposed to lead preparation efforts in various regions and districts . Port health officials and immigration officers were also selected because they were trained to provide screening service at entry and exit points in Ghana . The participants were between 20–86 years and were either Christians , Moslems or adhered to African Traditional Religion . Some participants had no formal education whilst other had attained tertiary education . Whereas majority of the participants were married , a number of them were single . We sought permission from participants to record the FGDs and IDIs using digital voice recorders . We also took field notes which were turned into data document for analysis . The interviews that were conducted in local languages were translated to English by two independent language experts during transcription . The transcripts were again reviewed by listening to the voice recordings and comparing its content with the corresponding transcript . All the transcripts were imported into Nvivo 10 for windows for analysis . We used framework method of qualitative data analysis in this study [21] . This method of data analysis involved the use of five iterative steps: familiarization , identifying a thematic framework , indexing , charting , mapping and interpretation . The actual data analysis process involves moving between these steps until the data analysis is completed [21] . The study team first read the transcripts to understand the key themes from the data which were used to develop a codebook to guide the data analysis . This codebook was discussed and accepted by the research team . Guided by this codebook , the data was then coded first as free nodes and later transformed to tree nodes . We classified the data based on the source of the data and linked these sources to their attributes such as region , sex of respondent , and place of interview and nodes in Nvivo . We used the explore and query functions in Nvivo to provide a descriptive accounts of the data as required in framework analysis [22] . We then reviewed these descriptive accounts with existing literature to provide an explanation and interpretation of the data . Based on the data , we sketched frameworks which were grounded on the data to explain strategies to address socio-cultural practices that may hinder EVD containment in Ghana .
In this study , it emerged that funerals are considered very important in all cultures in Ghana . As such celebrated with elaborate ceremonies and rituals from death to performing final funeral rites . These ceremonies and rituals vary between cultures and geographical areas across the country . Three main sub-themes emerged regarding ceremonies and rites related to funerals were identified . These include; care of the body of the deceased , burial practices and funeral practices . These subthemes and the common practices across the regions is shown in Table 1 . Regarding care of the body , the study found that in communities in Ghana , a person who is believed to have died a natural death has to be bathed at home . This practice ( bathing ) is considered an honour and respect for the dead by family members and community . Bathing is only forbidden in situations where the person is believed to have died from an unnatural cause such as suicide or any other unnatural causes depending on the community . The study showed that bathing is often done without any form of protection except where the body is given to a funeral home for preparation towards burial . Any attempt to protect oneself at home during bathing is often viewed as a sign of disrespect for the dead . This practice was of great concern to members of the EVD response team and generally perceived as one of the cultural practices that may hinder prevention and control of EVD in Ghana as illustrated below: Filing pass the deceased body or staging the body on a palanquin was also identified as a common funeral practice in Ghana . In some instances , individuals in an attempt to show their extreme grief tend to lie on the body of the deceased during such funeral rites as indicated by this response in FGDs: Furthermore , some people use handkerchiefs and other materials to drive away flies that settle on the corpse . This occurs when the corpse is staged for long , in which case , it starts to decompose thereby attracting flies . Respondents also believed the dead also express their grief by weeping whilst being mourned . Therefore , wiping of the tears of the corpse was required in some instances as illustrated by respondents . Closely related to his practice is throwing of money at the dead as part of funeral rites . These monies sometimes come into direct contact with bodily secretions should the corpse starts to decay . Such monies are often collected by the dirge singers and the undertakers as reward for their various roles in the funeral activities . Others also share cigarette or drinks with deceased to give opportunity to the dead to share in a habit he/she was engaged in whilst alive as illustrated: Widowhood rites also emerged as a high-risk funeral practice among southern Ghanaian . In this practices , the widow is made to drink the water used to rinse the dead husband to show her innocence in the death of the husband . This practice is often activated when there are suspicions that the widow could be responsible for the death of her husband . The study also showed that sometimes children are bathed with water used to rinse the body of dead as a way of fortifying children . This ritual is undertaken when the deceased is believed to possess special spiritual powers that should be transferred to grandchildren . Bathing this water is also believed to offer spiritual protection against evil spirits in the community especially among Northern Ghanaians . The study generally revealed that majority of these high-risk funeral practices were part of the culture of the people and therefore very difficult to modify . However , some respondents in northern Ghana were of the view that if a particular death is believed to be due to Ebola , people may be unwilling to undertake such high-risk funeral and burial practices for fear of contracting the disease . Some also believe the use of some concoction could offer protection and prevent anybody who come in contact with the deceased’s body from getting infected as illustrated: Nonetheless in the other regions , bathing the dead body was perceived to be mandatory and could therefore not be overlooked for fear of the reprisal effects from the dead . They were of the opinion that if the bathing is not carried out then the soul of the person does not go to rest in peace but will wander about and sometimes confront those who took the decision to deny the deceased this rite of passage: Handshaking emerged as one of the everyday communal social norms that may influence Ebola prevention and containment . The study showed that handshaking are deeply embedded in all social gatherings such as funerals , religious congregations , and festivals . In fact , it is a daily occurrence when friends , work colleagues or acquaintance meet . To respondents , this was mandatory in some social settings and had deep roots in the culture and behavioral etiquettes of the people as illustrated by a respondent . Another community practice that can promote the spread of Ebola in the community is the practices of sharing drinks in drinking bars . It is a common practice that one person would share his/her drink with another person using the same calabash . Closely related to this is the use of communal glasses by operators of drinking bars . Such glasses are often not washed well before they are used to serve another customer as illustrated by the following responses: The study further revealed that during Eucharistic celebration by Christians , common cups are often used by members of the congregation . The communal use of cups is a high-risk practice for Ebola since the cups may come into contact with the saliva of users . Gender related social factors also emerged in both FGDs and IDIs across the country . These gender-related issues may increase vulnerability of either the male or female gender depending on whether the person belong to a matrilineal or a patrilineal descent . First and foremost , the study revealed that in all cultures , women were the prime caretakers with the responsibility of taking care of the sick at home and this included bathing , feeding and administering drugs . Hence , in the event that there should be an Ebola outbreak in Ghana , females will be more likely to get infected than their male counterparts . Men only offer to assist when the condition becomes critical which requires moving the sick person to another source of care: In the Ashanti Region , and other Akan speaking regions in Ghana , inheritance is matrilineal and this also affects domestic care giving . This type of inheritance makes it obligatory for women to take care of sick people in their descent . On the other hand , men in society are often in-charge of taking care of dead bodies . This also makes men more susceptible to Ebola should the death be as a result of EVD: Decision-making process also largely favoured men in the communities as men often make key decisions at the household level as showed by these responses: The study showed that some Ghanaians use multiple health care outlets in their communities including self-medication at home , buying drugs from license drugs store , and health facilities or a traditional healers . However , taking medications ( both herbal and orthodox ) at home or buying medicines from a licensed drug stores as first aid was reported to often precede reporting to any health facility . In FGDs , it was unanimously agreed that drugs are first purchased from the drug stores for treatment at home and if the sick person condition showed no improvement after some days , then the individual is sent to either the hospital or traditional healer . Others will take some herbal preparations first and then report to the hospital if their condition does not improve . According to respondents , several factors determine the type of health care that is sought . Quick access to health care emerged as one of the factors that influence the place a sick person seeks health care . Another consideration in seeking health care is the type of health condition . In the opinion of some community members , some conditions are not “sickness meant for the hospital” . These conditions are often described as spiritual or supernatural in origin and therefore unsuitable for hospital care . One of such conditions was mentioned as “yɔgu” , a Dagomba word in Northern Ghana which according to FGD discussants is characteristically similar to Ebola . The condition “yɔgu” translated to mean “deep forest” is a disease that people get from deep forest through either eating the touching an infected animal or eating the infected meat . From the study most of the socio-cultural practices were perceived to be deeply rooted in tradition . Nonetheless , respondents were of the view that some of them could be modified in the event of Ebola outbreak through collaborations with chiefs and religious leaders in the community . People who adhere to the Islamic faith , generally believed that involving their religious leaders is the only practical ways of changing high-risk funeral and burial rites . Bathing the dead was believed to be part of their religious doctrine and could only be changed or modified by the Islamic leaders who understand the teachings of the Quran better .
Practice of bathing the body of the deceased is a major risk factor in EVD prevention and control as identified in this study . The people who bath the corpse are often not protected and several people may be involved in the process . This therefore puts all the people involved at risk of getting infected should it be a case of Ebola . In situations where bathing is performed by the undertakers of a particular clan , it implies that the infection could be transmitted to other clans in the community . Practices such as bathing the corpse , anointing people with water used to rinse the corpse , and staging the body for funeral rites and burial practices , are high-risk activities and have been reported to have exacerbated human-to-human transmission in Liberia and Sierra Leone [23] . In Guinea , 60% of all EVD cases had been linked to traditional burial practices [24] . In this study , almost all respondents acknowledged the high-risk inherent in these funeral and burial practices . Nevertheless , there was little readiness to change these practices as it would be seen as an affront to their culture . Despite that impression , with the fear and general perception of the high fatality of EVD among communities , it may be possible to induce a change if an EVD case should be confirmed in the nation . This however will require the involvement of community leaders coupled with close monitoring . Lesson learnt from the history of cholera in Ghana should guide this effort . It has been documented that two of the worst areas in Ghana that have been affected with cholera are Akplabanya and Nyanyano located in Ada and Winneba districts of Ghana respectively . These outbreaks were widely speculated to have resulted from the bringing in of relatives of Ghanaians who had died from cholera in Togo and Guinea for burial in these two fishing communities thereby infecting others [25] . In Sierra Leone , as many as 350 Ebola deaths were attributed to the funeral of a renowned traditional healer who got infected when treating EVD patient [24] . Involving community leaders’ especially religious leaders and strict monitoring will be required to ensure that people adhere to directives on safe funeral practices . Ghana should consider developing a safe funeral and burial practices documents which will consider inputs from various stakeholders . Lessons from Liberia has showed that funerary as well as caregiving practices could either be suspended or altered by community members when they are fully engaged [26] . Local community engagement is reported to be crucial in reducing transmission rates [27–29] . The study showed that hand shaking is highly cherished in Ghanaian settings . The popularity of the handshake continue to grow and is seen as a sign of love , affection and concern in all social gatherings . Therefore , an individual stand the risk of being perceived to be rude where he/she refuses to accept a hand shake or offer one when it is required . Since , handshake is another medium of transmission of EVD , it will be important for information , education and communication messages to provide alternative greeting strategies that will limit the more physical contact inherent in the traditional handshakes . In a recent study , it was reported that the so-called high five and fist bump compared with a traditional handshake substantially reduce the transmission of infectious disease between individuals [30] . This should therefore be emphasized in health education strategies to help reduce physical contacts through handshakes . In the absence of alternatives , people are compelled to stick to the traditional ways of handshaking which increases the risk of transmission of infections such as EVD . The use of communal cups and calabashes were mentioned as common and is a high risk social practice . This was not only reported as a common practice in drinking bars but also in the Eucharistic practice in churches . During the Eucharistic practice , common communion cups are used by the congregation . A classical study has showed that during the use of such communion glasses in church , several micro-organisms are deposited on the glasses [31] . This therefore requires serious attention in EVD prevention and control strategies . For the use of communal glasses and cups in drinking bars , EVD-related health education should highlight the potential risk in this practice . When people are sensitize to demand disposable cups , the bar operators will be compelled to make available disposable cups . In times of emergencies like EVD outbreak , a directive can also be given by the appropriate authorities to ban the use of communal cups and drinking glasses in bars . However , regarding the use of communion glasses , church leaders can be sensitize to stop the use of communal drinking glasses in their religious practices . The study showed that socio-cultural norms from society makes both men and women vulnerable to one form of high-risk behavior or the other . Socio-cultural norms require that females take care of sick at home and at the hospital . Taking care of the sick involves feeding and washing of the sick person . These activities may require coming into contact with bodily fluids of the sick person thereby increasing the risk of contracting the EVD . This social norm therefore increases the vulnerability of women than menfolk . Studies in the three epicenter countries ( Liberia , Guinea and Sierra Leone ) have revealed that EVD deaths is higher among women than men with 50 . 8% of the cases being women as of 7th January , 2015 [32] . In Guinea , number of EVD cases among males and females were 1 , 309 and 1 , 410 respectively . Also in Sierra Leone , more females ( 4 , 151cases ) were infected than males ( 3 , 891 cases ) [32] . Similarly , in previous ebola outbreaks in Congo , Gabon and Uganda , women were reported to more affected [32] . In the same vein , the study showed that some socio-cultural norm also make males responsible for certain activities in the community which are considered high-risk activities . Though some women may be called upon to care for the dead , this activity is mostly undertaken by men hence , if the deceased was a case of EVD , it will increase the risk of men getting infected . The study further showed that decision making tend to favour men as they are responsible for making important decisions such as determining the health seeking behaviour of their household . It is therefore important for such gender related issues to be considered in any preparation towards containing EVD . Fig 2 shows how socio-cultural practices can increase an individual’s risk of infection . The study showed that multiple treatment outlets are often sought either simultaneously or in sequence . From all indications , the first point of call for a sick person is not the hospital or herbalist/ritualist but care at home through the use of either drugs obtained from drugstores , or passed on from other relatives , with some people taking herbal preparations . An earlier study in Nigeria reports that about a quarter ( 26% ) of the study population believed local and traditional medicines ( herbs and concoctions ) could provide a cure for EVD [33] . Given the nature of EVD , this approach will be an impediment in efforts to contain it should there be an outbreak . This is because sick people will be kept at home and only sent to hospital or herbalist/ritualist after their conditions have become worse and/or they have infected other members of their household . Since these first aid treatments are sometimes bought from drug stores in the community , in the event of an Ebola outbreak in Ghana , involving caretakers of drug stores may be essential since they are the first point of call for sick people in the community . They should be equipped with the knowledge and skills required to screen people for EVD so that they can screen their customers appropriately . The training of frontline health workers should therefore extend to members of chemical license sellers associations in communities because of their potential role in identifying cases of EVD . Nonetheless , majority of respondents in the study stated that if a sick person is suspected to be carrying EVD , the hospital will be the first point of call for health care . Despite this , the mere fact that signs and symptoms of EVD are akin to endemic conditions like malaria , diarrhea , typhoid fever ( enteric fever ) and pneumonia , means that people might misconstrue the early clinical presentation of EVD for these conditions . Therefore , home care will still persist until such a time that haemorrhagic signs manifest as this appears to be the only definite indication that suggests an EVD disease to community members . Haemorrhagic manifestation are later signs of EVD and do not occur immediately [4 , 5] , therefore the possibility of infecting household members and close associates before health care is sought at a biomedical health facility could be high . The summary of the health seeking behavior of Ghanaians and the ways to address delays are highlighted on Fig 3 . Involving the various health outlets in EVD prevention and control will ensure that people that report to such outlets are screened appropriately to avoid delay which will invariably increase the risk of infecting household and community members . Interviews conducted in local language were translated to English by independent language experts and these translation were verified by another language expert . Nevertheless , there is still the possibility of words losing their original meaning during the translation procedure . To circumvent this weakness , we placed emphasis on well-entrenched themes during coding rather than the specific phrases and word used by participants . Where it was impossible to find a suitable English word or phrase for a local term , we maintained the local word and placed the direct translation in English in parenthesis . We were also only able to conduct the study in five out of the ten regions in Ghana . This therefore has implications in generalizing the findings of the study to the regions that were not covered in the study . Nonetheless , we ensured that regions that were selected fairly represented the main ethnic groups in the country . The study showed across Ghana , high risk socio-cultural practices were common and generally perceived to have deep root in culture , therefore making such practices indispensable . Socio-cultural factors may hinder containment efforts in the event of an outbreak . Community leaders should be engaged in any social mobilization to modify these practices as part of preparation efforts . Social mobilization through community leaders and culturally appropriate health education are required .
|
The 2014 Ebola Virus Disease outbreak emerged as the most devastating outbreaks in recent history . This outbreak spread across continents with West Africa remaining the epicenter . Although , the three most affected countries have been declared Ebola free , the recent re-emergence of cases in Sierra Leone require that countries in West Africa do not relent in their EVD containment efforts . Burial and funeral practices are high-risk practices that can facilitate the spread of ebola and hinder containment efforts . These practices vary across countries and regions . In this study , we found that despite the social and contextual difference across regions in Ghana , common high-risk practices exist and perceived to be indispensable . We also found that little attention were given to these practices in risk communication and community engagements . These practices required either touching the dead or items used in taking care of the dead . Widowhood rites require a widow drinking water used to bath the deceased husband to show her innocence in the husband’s death . Findings of this study attest to the need for improved dialogue with community leaders to provide alternatives to these high-risk socio-cultural practices to be able to contain an Ebola outbreak . Countries with similar socio-cultural practices could also use the findings of this study to guide behavioural change communication strategies .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"amorphous",
"solids",
"medicine",
"and",
"health",
"sciences",
"health",
"services",
"research",
"sociology",
"geographical",
"locations",
"tropical",
"diseases",
"social",
"sciences",
"ebola",
"hemorrhagic",
"fever",
"anthropology",
"neuroscience",
"glass",
"herbs",
"health",
"care",
"cognitive",
"psychology",
"drug",
"licensing",
"materials",
"science",
"neglected",
"tropical",
"diseases",
"pharmacology",
"cultural",
"anthropology",
"plants",
"africa",
"materials",
"by",
"structure",
"religion",
"language",
"viral",
"hemorrhagic",
"fevers",
"infectious",
"diseases",
"people",
"and",
"places",
"psychology",
"ghana",
"drug",
"research",
"and",
"development",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"physical",
"sciences",
"cognitive",
"science",
"organisms"
] |
2016
|
Preparing towards Preventing and Containing an Ebola Virus Disease Outbreak: What Socio-cultural Practices May Affect Containment Efforts in Ghana?
|
Microglia are resident immune cells that play critical roles in maintaining the normal physiology of the central nervous system ( CNS ) . Remarkably , microglia have an intrinsic capacity to repopulate themselves after acute ablation . However , the underlying mechanisms that drive such restoration remain elusive . Here , we characterized microglial repopulation both spatially and temporally following removal via treatment with the colony stimulating factor 1 receptor ( CSF1R ) inhibitor PLX5622 . We show that microglia were replenished via self-renewal , with no contribution from nonmicroglial lineages , including Nestin+ progenitors and the circulating myeloid population . Interestingly , spatial analyses with dual-color labeling revealed that newborn microglia recolonized the parenchyma by forming distinctive clusters that maintained stable territorial boundaries over time , indicating the proximal expansive nature of adult microgliogenesis and the stability of microglia tiling . Temporal transcriptome profiling at different repopulation stages revealed that adult newborn microglia gradually regain steady-state maturity from an immature state that is reminiscent of the neonatal stage and follow a series of maturation programs , including nuclear factor kappa-light-chain-enhancer of activated B cells ( NF-κB ) activation , interferon immune activation , and apoptosis . Importantly , we show that the restoration of microglial homeostatic density requires NF-κB signaling as well as apoptotic egress of excessive cells . In summary , our study reports key events that take place from microgliogenesis to homeostasis reestablishment .
Microglia are resident parenchymal macrophages in the central nervous system ( CNS ) . Along with the perivascular , meningeal , and choroid plexus macrophages , these cells govern the innate immunity of the CNS [1] . Apart from their primary function in immune surveillance , microglia also perform a multitude of essential CNS tasks: supplying neurotrophic factors , pruning unwanted synapses , and promoting programmed cell death [2] . Consequently , microglia have been implicated in neurodegenerative and neuropsychiatric diseases [2 , 3] . In particular , loss of microglial homeostatic control in the form of microgliosis is considered one of the key pathological hallmarks of neurodegenerative diseases . During embryonic development , myeloid progenitor cells leave the blood island and travel from the yolk sac to colonize the developing fetal brain , which later become microglia [4–6] . In the adult brain , microglia are maintained locally through self-division if the blood brain barrier is kept intact [7] . However , how microglia repopulate the brain following microglial ablation has been intensely debated in the field . It has been shown by multiple groups that the empty microglial niche can be readily repopulated in a matter of days after ablation of more than 90% of cells using colony stimulating factor 1 receptor ( CSF1R ) inhibitor [8 , 9] . In this setting , Elmore and colleagues described a highly proliferative Nestin+ nonmicroglial population , which preceded the appearance of microglial markers , leading to the conclusion that a hidden CNS Nestin+ progenitor pool can rapidly replenish microglia deficiency [8] . Other microglial ablation studies have shown that microglia can be regenerated almost entirely via self-renewal [9 , 10] . In addition , Nestin+ progenitors transitioning into adult microglia have not been detected in microglia generated under steady-state conditions [11] . Microglia are territorial sentinels in the CNS . Resting-state microglia evenly tile the parenchyma , forming an intricate cellular grid , with microglial processes constantly surveilling local territories [12] . This tiling pattern is largely determined during seeding of the primitive yolk sac progenitors in the embryo [5] . However , adult microglia have the capacity to migrate over long distances through the optic nerve during microglial repopulation [13] , making it challenging to determine whether maintenance of microglial tiling is static or dynamic . Here , we applied transient 5-Ethynyl-2′-deoxyuridine ( EdU ) labeling , lineage tracing , as well as bone marrow–transplant experiments during microglial repopulation to resolve the debate over the origin of repopulating cells . Our results indicate that the repopulated microglia are exclusively derived locally from self-renewed microglia rather than a Nestin+ progenitor or circulating myeloid cells . Using a dual labeling system and spatial analyses , we investigated spatial tiling and found that microglial colonies are very stable in the CNS once they are formed . In addition , to further elucidate the mechanisms underlying microglia homeostasis restoration during repopulation , we performed an RNA-seq–profiling experiment from isolated microglia at various stages of repopulation and showed that the early-stage , adult-born microglia are reminiscent of the neonatal counterpart and undergo a series of transcriptional changes to eventually regain a mature phenotype . In particular , components of immune signaling pathways , particularly nuclear factor kappa-light-chain-enhancer of activated B cells ( NF-κB ) and interferon , were highly elevated , accompanied by a high rate of apoptosis . Conditional knockout of I-Kappa-B-Kinase Beta ( IKKβ ) , a kinase required for NF-κB activation , impaired microglial repopulation in vivo . In summary , our results provide new insights into microglial homeostatic control , demonstrating that the cells have an intrinsic regenerative capacity , activate developmental programs during regeneration , and undergo regulated apoptosis to regain homeostatic density .
Microglia are resident immune cells that territorially tile the entire CNS . Microglial homeostasis is tightly controlled under physiological conditions [11] . To investigate the mechanisms underlying microglial homeostatic restoration , we depleted microglia by feeding C56/BL6J mice a diet containing the CSF1R inhibitor PLX5622 ( PLX ) . Administration of PLX for 2 weeks reduced the number of ionized calcium binding adaptor molecule 1 ( Iba1+ ) microglia over 90% . After switching to a regular diet for 7 or 14 days , the microglial population was readily restored and exceeded steady-state level ( Fig 1A and 1B ) . Similar microglial repopulation was observed following diphtheria toxin–mediated selective ablation ( S1A Fig ) . In this model , microglial density was partially restored after 1 day of repopulation and surpassed normal density after 7 days of repopulation ( S1B and S1C Fig ) . These results indicate that repopulation is an inherent property of microglia irrespective of the method of ablation . To trace the source of repopulating microglia , we injected EdU , a thymidine analog that can be incorporated into newly synthesized DNA , during the first 4 days of microglial repopulation ( Fig 1C ) . Approximately 60% of EdU+ cells were also Iba1+ ( Fig 1D and 1E and S2 Fig ) . Among those cells , we observed pairs of EdU+ microglia that appeared to be undergoing mitosis ( Fig 1F ) , suggesting that microglial repopulation involves self-proliferation . In line with activation of cell cycle re-entry , live imaging of brain slices from microglia reporter mice ( CX3CR1eGFP/+ ) , which express eGFP under the myeloid promoter CX3CR1 , revealed that microglia exhibited increased movement on repopulation day 6 ( S3 Fig , S1 Video , and S2 Video ) . However , roughly 10%–30% EdU+ cells did not express the microglial marker Iba1 ( Fig 1D ) . To identify the EdU+Iba1− cells , we performed staining with Nestin and an array of other common CNS markers ( Fig 1G–1L ) . In contrast to what was previously reported [8] , Nestin was not expressed in the EdU+Iba1− cells ( Fig 1J , open arrow , and Fig 1M ) but was present in roughly 30% EdU+Iba1+ cells ( Fig 1J , solid arrow , and Fig 1N ) , indicating that Nestin may be a transient marker of young microglia . In addition , we also found that a small subset of the EdU+Iba1− cells expressed doublecortin ( DCX ) or oligodendrocyte transcription factor 2 ( Olig2 ) ( Fig 1I , Fig 1L and Fig 1M ) , markers of neurogenesis and oligodendrocytes , respectively . To directly test whether circulating monocytes contribute to microglial repopulation after PLX-mediated depletion , we performed bone marrow transplantation ( BMT ) using wild-type C57BL/6J mice as recipients and ACTB-eGFP mice as donors . The ACTB-eGFP transgenic mice constitutively express eGFP under the control of a chicken beta-actin promoter and cytomegalovirus enhancer [14] , enabling the tracing of reconstituted circulating monocytes in the chimeric mice by the presence of eGFP ( Fig 2A ) . Since irradiation of the head can cause artificial monocyte infiltration of the brain [15] , mice were fitted with a lead helmet to shield their heads; protection was evidenced by the maintenance of pigmented head hair following irradiation ( Fig 2B ) . Following BMT , mice were subjected to microglial depletion using the PLX diet for 2 weeks and then switched to normal diet for repopulation . On average , chimeric mice showed 70% myeloid reconstitution , as determined by the presence of eGFP in the CD11b+CD45+ population from blood or splenocyte suspension ( S4 Fig ) . However , immunofluorescence staining of coronal brain slices showed very few green fluorescent protein ( GFP ) + cells in the parenchyma , either after 14 days or 2 months of repopulation ( Fig 2C ) . Overall , only a small fraction ( <0 . 5% ) of the repopulated microglia were GFP+ ( Fig 2D ) . Most of the GFP+Iba1+ myeloid cells were found in the choroid plexus ( Fig 2E ) , which is expected because choroid plexus macrophages are continuously replenished by peripheral monocytes [1] . Consistent with previous studies [9 , 10] , our findings show that bone marrow–derived hematopoietic myeloid cells are not a major source for the repopulated microglial pool . We and others have observed that Nestin , a marker of neuroprogenitor cells , is transiently expressed in early-stage newborn microglia [8 , 10] . To directly test if Nestin+ progenitors can also contribute to the repopulated microglial pool , as proposed previously [8] , we performed fate-mapping experiments using Nestin-CreERT2 [16] and the STOPflox-RFP reporter to label Nestin+ progenitor cells with red fluorescent protein ( RFP ) ( Fig 3A ) . Upon tamoxifen treatment , removal of the STOP cassette by the CreERT2 recombinase triggers RFP expression under a constitutive CAG promoter , thus ensuring any descendent cell divided from the labeled microglia will also be RFP+ . Neuronal precursor cells were labeled with RFP in the subgranular zone ( SGZ ) , as expected ( Fig 3B and 3C ) . After 3 weeks of PLX treatment , followed by 2 weeks of microglial repopulation ( Fig 3D ) , the number of RFP+ cells remained unchanged ( Fig 3E ) , and none of the repopulated microglia expressed RFP ( Fig 3F ) . Thus , Nestin+ progenitors do not contribute to the repopulated microglia pool . Given that a small proportion of EdU+Iba1− express Olig2+ during microglia repopulation ( Fig 1L ) , we also tested whether oligodendrocyte precursor cells ( OPCs ) could give rise to microglia . STOPflox-RFP reporter mice were crossed with mice that express tamoxifen-inducible Cre recombinase under either the platelet derived growth factor receptor alpha ( PDGFra ) promoter [17] or the neural/glial antigen 2 ( NG2 ) promoter [18] . While PLX treatment did not affect the number of PDGFra+ progenitor cells labeled with RFP , none of the RFP+ cells became repopulated microglia ( S5A–S5D Fig ) . Similar results were seen in the NG2-CreERT2 mice ( S5E–S5H Fig ) . Taken together , these lineage-tracing experiments indicate that the origin of adult microgliogenesis is distinct from that of neurogenesis or oligodendrogenesis . To further investigate the origin of the repopulated microglia , we performed fate-mapping analysis using a combination of the STOPflox-RFP reporter [19] and the tamoxifen inducible myeloid-specific driver CX3CR1-CreERT2 [20] to genetically label adult microglia ( Fig 3A ) . Treating mice for 10 days with 2 mg tamoxifen/day resulted in approximately 85% efficient microglial labeling ( Fig 3G and 3H ) , similar to other studies [20 , 21] . If microglia are repopulated from a nonmicroglial source , this labeling rate would be expected to drop following ablation . To test this , we subjected labeled mice to 2 rounds of microglial depletion and repopulation . We found that the percentage of RFP expression in repopulated microglia remained constant at approximately 85% for both repopulation rounds ( Fig 3G and 3H ) . Thus , the repopulated microglia are almost exclusively derived from the remaining CX3CR1+ microglia , most likely via self-renewal . Notably , the second round of depletion did not exhaust the regenerative capacity of microglia , further supporting the self-renewal property of this population ( Fig 3I ) . Results from the CX3CR1+ lineage tracing suggests that microglia self-renewal alone is sufficient to repopulate the entire parenchyma after acute ablation , as also found by Bruttger and colleagues [10] as well as Huang and colleagues [9] . In order to investigate the tiling behavior of microglia during repopulation , we applied Microfetti labeling [22] using the Brainbow2 . 1 reporter [23] and CX3CR1-CreERT2 . Microfetti utilizes stochastic labeling of microglia with 4 fluorescent proteins: GFP , yellow fluorescent protein ( YFP ) , cyan fluorescent protein ( CFP ) , and RFP . Due to fixative-induced inactivation of native fluorescent signals , we used an anti-GFP antibody to detect GFP/CFP/YFP and an anti-RFP antibody to detect RFP ( Fig 4A ) . After 2 weeks of PLX treatment , the CX3CR1-CreERT2/STOPflox-Brainbow mice were switched to a normal diet for repopulation over an extended period ( up to 1 month ) to examine the long-term clonal behavior of newborn microglia ( Fig 4B ) . At a low dose of tamoxifen , uniform and sparse labeling of microglia with the Brainbow reporter was observed via immunostaining of GFP and RFP ( Fig 4C ) . On average , 6 . 9% microglia were labeled with RFP , while 9 . 9% were labeled with GFP . PLX-mediated depletion removed over 90% of Iba1+ microglia ( S6A and S6B Fig ) and retained a few remaining RFP+ or GFP+ cells at day 0 ( Fig 4C ) . After 7 days of repopulation , clonal clusters of both RFP+ and GFP+ microglia were observed ( Fig 4C ) , indicating that adult microgliogenesis occurs via clonal expansion . To study the properties of the clusters and their dynamics over time , we assigned nearest neighbor distance ( NND ) values to individual cells based on their distance away from the nearest neighbor . In this analysis , cells inside a tightly formed cluster will have low NND values , whereas dispersed cells will have high NND values . The NND-encoded heatmap offers an overview of the clustering patterns and shows that labeled cells form patches of clusters after repopulation , as expected from self-renewal ( Fig 4D and 4F ) . The density of NNDs near the 50-μm interval increased during repopulation , indicative of colony formation ( Fig 4E and 4G ) . Because unlabeled Iba1+ microglia in naive mice have an NND value of approximately 50 μm ( S6C Fig ) , we used NND of less than 50 μm to quantify changes in clustering patterns compared to controls . Using this criterion , there was an increase in the percentage of RFP+ cells with NND values less than 50 μm at 7 days and 1 month following ablation compared to the Ctrl group ( Fig 4H ) . Interestingly , there was no difference in clustering pattern between 7 days and 1 month , suggesting that once the clusters are formed , they remain stable . Similar results were obtained by analyzing GFP+ cells ( Fig 4F and 4G ) , although NND analysis did not reach statistical significance ( Fig 4I ) , most likely because 3 fluorophores ( CFP/YFP/GFP ) were combined in this group , resulting in denser labeling in the controls . To further confirm the results obtained from the NND analysis , we next used Ripley’s K-function analysis and its derivative H-function to survey the spatial points in a defined field with varying radii . Consistent with our NND analysis , the H-function analysis showed that the 7 day and 1 month groups shared similar clustering patterns that were quite different from the Ctrl group for both RFP+ and GFP+ cell populations ( Fig 4J and 4K ) . Furthermore , quantification of the radius of H-function at the maximum score [24] showed that the cluster sizes were indistinguishable between the 7 day and 1 month groups ( Fig 4L ) , suggesting that the size of both RFP+ and GFP+ clusters remain relatively constant over time . We next investigated whether distinctive clusters could merge with their neighbor through cell migration and spread . Using 2D kernel density maps generated by applying kernel smoothing to actual microglia spatial points , we examined the cluster interaction between RFP+ and GFP+ cells . In order to distinguish cluster boundaries , we used top 10% density from the kernel density map as a threshold . This method allows us to artificially outline cluster boundaries for both RFP+ and GFP+ cells at 7 days ( Fig 4M ) and 1 month of repopulation ( Fig 4N ) . We then measured the relative overlapping area between RFP+ and GFP+ clusters: if a cluster merged into neighboring clusters , the overlapping area would increase accordingly . The analysis showed a consistent level of overlapping area between the 7 day and 1 month groups ( Fig 4O ) , suggesting that the clusters’ territories are stable with minimal migratory diffusion during the repopulation process . Together , these data indicate adult newborn microglia redistribute spatially , form distinct clusters , and maintain stable tiling pattern over long periods of time . We next performed RNA-seq to profile adult newborn microglial transcriptome at different stages of repopulation after PLX treatment ( S1A Table and S2 Data ) . After 2 weeks of PLX treatment , repopulated microglia were collected at 4 days ( 4 D ) , 14 days ( 14 D ) , and 1 month ( 1 Mo ) and compared to immature microglia isolated from postnatal day 4 ( P4 ) pups ( Fig 5A ) [25 , 26] . Microglia were isolated by the magnetic-activated cell sorting ( MACS ) method using magnetic beads conjugated with CD11b antibody and showed no significant presence of other major CNS cell types based on their specific marker expression ( S7A Fig ) . We confirmed the RNA-seq data via quantitative PCR ( qPCR ) based on a few differentially expressed genes ( S7B–S7D Fig ) . Dimensional reduction with principal component analysis ( PCA ) showed biological replicates from each treatment group were very closely related , as indicated by the clustering ( Fig 5B ) . There was a clear separation of the adult microglia ( Ctrl ) and immature microglia ( P4 ) clusters as expected ( Fig 5B ) . Interestingly , early-stage newborn microglia ( 4 D ) samples clustered between the Ctrl and P4 microglia clusters , while medium- ( 14 D ) and late-stage ( 1 Mo ) microglia clustered closer to control adult microglia ( Ctrl ) , suggesting that newborn microglia undergo progressive restoration toward a steady-state microglial transcriptional profile . Next , we applied k-means clustering to arbitrarily categorize all genes into 4 distinct clusters . Among all the repopulating microglia groups , the earliest stage ( 4 D ) showed the most transcriptional difference from Ctrl ( Fig 5C ) . Interestingly , genes in cluster 3 and 4 shared expression directionality between 4 D repopulated microglia and P4 neonatal microglia ( Fig 5C ) . This suggests that the newly regenerated microglia might have partially reverted back to an immature developmental state . To assess the similarity of microglia isolated from different repopulation stages , we employed a Poisson distance matrix based on gene expression profiles . Interestingly , as repopulation time goes by , newborn microglia ( 4 D , 14 D , and 1Mo ) showed decreasing similarity with P4 neonatal microglia but increasing similarity with steady-state microglia ( Ctrl ) ( S8A Fig ) , suggesting that newborn microglia gradually regains maturity . In particular , we found 4 D microglia shared 34% similarity with P4 neonatal microglia and 55% similarity with Ctrl microglia ( S8A Fig ) . This suggests that 4 D microglia adopt a unique immature transcriptome signature that partially overlaps with that of neonatal microglia . We next performed differentially expressed ( DE ) gene analysis . Roughly half of the DE genes found in 4 D microglia were also present in P4 microglia ( Fig 5D and 5E ) . Among all of the up-regulated genes shared between 4 D and P4 microglia , many are involved in cell cycle regulation ( S8B and S8C Fig ) , including Mcm5 and Cdk1 ( Fig 5F and 5G ) . Importantly , Nestin was also elevated at this early stage ( Fig 5H ) but returned to control levels after 14 days , further validating our earlier observation that Nestin was an immature marker expressed in newborn microglia rather than a bona fide microglial progenitor marker . On the other hand , among all of the down-regulated genes shared between 4 D and P4 microglia , a large fraction is involved in Mitogen Activated Protein Kinase ( MAPK ) pathway and transforming growth factor β ( TGF-β ) signaling ( S8D and S8E Fig ) . TGF-β signaling has been shown to be critically important for microglial development and homeostatic maintenance [26] , and TGF-β signaling-related genes such as Pmepa1 and Smad7 were found to be down-regulated in 4 D microglia ( Fig 5I and 5J ) . In addition , previously established mature microglia genes such as MafB and Selplg [25] and P2ry12 and Tmem119 [25–27] were also down-regulated in 4 D microglia but restored after 14 days ( Fig 5K–5N ) . These results indicate that newborn microglia regain maturity over time , and this was validated in morphological analysis and immunostaining of P2ry12 and Tmem119 protein expression in separate experiments ( S9 Fig ) . Interestingly , disease-associated microglial genes such as Trem2 [28] and Grn [29] were also differentially expressed in 4 D microglia ( Fig 5O and 5P ) . Collectively , the results gathered from the RNA-seq analyses suggest that adult microgliogenesis involves maturation steps that partially recapitulate development . To further dissect the underlying programs associated with microglial maturation , we classified 4 different gene sets based on their distinctive differential expression patterns during each of the repopulation stages ( Fig 6A and 6B; log2 ratio of ±1 , false discovery rate [FDR] < 0 . 05 ) . Fast-returning genes were differentially expressed only in 4 D microglia but returned to control levels in 14 D microglia ( S1B Table ) . Medium-returning genes were differentially expressed in 4 D and 14 D microglia but not in 1 Mo microglia ( S1C Table ) . Slow-returning genes were differentially expressed in all repopulation stages ( S1D Table ) . In addition , we also uncovered a fourth set of genes that were only differentially expressed in 14 D microglia , which we refer to as delayed-response genes ( S1E Table ) . We applied gene set enrichment analysis ( GSEA ) with molecular signatures database ( MSigDB ) on these 4 gene sets [30 , 31] . Comparison of the fast-returning genes with the hallmark gene data set in MSigDB revealed enrichment of genes involved in cell division such as E2F target genes and genes involved in the G2M checkpoint or mitotic spindle ( Fig 6C ) . Indeed , cell cycle–related genes such as Ccnb2 , Cdc20 , and Cdc25b were highly expressed in 4 D microglia but returned to homeostatic levels in 14 D microglia ( Fig 6D–6F ) . This result further supports our earlier observation that re-entry to the cell cycle is an early step in microglial repopulation ( Fig 1F ) . Gene enrichment analyses of fast-returning genes revealed a high degree of overlap with NF-κB pathway-related genes ( 41 out of 200 genes ) such as Traf1 , Il1a , and Il1b ( Fig 6G–6I ) . Both Il1a and Il1b are also associated with cell death . Interestingly , apoptosis-related genes ( 23 out of 161 genes ) were also among the most enriched genes , including Bik ( Fig 6J ) . Next , we examined the medium-return genes . Gene enrichment analysis showed strong representation of genes involved in interferon ( IFN ) pathways such as Stat1 , Mx1 , and Oas2 ( Fig 6K–6N ) . The enrichment of IFN genes indicates inflammatory activation during early and middle phases of microglial repopulation that was resolved after 1 month . The gene enrichment analysis for the slow-return gene set did not have statistically significant overlapping pathways . However , gene ontology analysis focused on cellular components revealed 34% of slow-return genes are cell surface related ( Fig 6O ) . Among the most up-regulated genes was Axl ( Fig 6P ) , an important receptor involved in microglial phagocytic clearance of dead cells [32] . Other down-regulated genes include Mmp15 ( Fig 6Q ) , a metalloproteinase involved in extracellular matrix remodeling [33] , and Sdc4 ( Fig 6R ) , which is an important cell surface adhesion molecule involved in cell activation and migration [34] . The delayed-response gene set revealed modest enrichment of GSEA hallmark genes ( Fig 6S ) . Interestingly , immune-related genes across different inflammatory functions were highly over-represented . Ccl5 , also known as RANTES , is an important chemotactic cytokine that modulates T-cell migration [35] and was found to be highly expressed ( Fig 6T ) . Rgs16 , which antagonizes inflammatory activation in monocytes [36] , was significantly down-regulated ( Fig 6U ) . Gbp6 , an IFN inducible guanosine triphosphatase , was found to be highly expressed ( Fig 6V ) . Altogether , pathway enrichment from each gene set revealed a molecular snapshot of the stages of microglial homeostatic restoration . Among the fast-returning genes , an intricate network of NF-κB pathway-associated genes was identified ( Fig 7A ) . To understand the biological relationship between NF-κB signaling and early-phase microglial repopulation , we conditionally knocked out both copies of loxP-flanked I-Kappa-B-Kinase Beta ( IKKβ ) in microglia using CX3CR1-CreERT2/IKKβF/F mice ( Fig 7B ) . This approach has been shown to suppress microglial activation mediated by the NF-κB pathway [37] . Conditional IKKβ knockout mice were treated for 2 weeks with PLX diet followed by 4 days of repopulation . Interestingly , IKKβ deletion showed a modest but statistically significant impairment in microglial repopulation density compared to Cre− control ( Fig 7C and 7D ) . These results suggest that NF-κB signaling enhances the early phase of microglia repopulation . Apoptosis-related genes were also highly represented among the fast-returning genes found in our RNA-seq data ( Fig 6C and Fig 8A ) . To validate the involvement of cell death , we measured cell death by TUNEL assay at different stages of microglia repopulation . The number of TUNEL-positive cells was dramatically increased in the parenchyma , peaking at 4 days of repopulation ( Fig 8B and 8C ) . This rate of death was significantly higher than at the end of PLX administration ( 0 D ) , suggesting that it was not directly caused by PLX treatment ( Fig 8B and 8C ) . We also observed phagocytic cups , a special morphological feature that is indicative of phagocytosis [22 , 38] , around the TUNEL+ cells ( Fig 8D and 8E ) . Further quantification in 4 D microglia show that approximately 40% TUNEL+ cells were in phagocytic cups ( Fig 8F ) , while only 0 . 6% of microglia displayed this structure ( Fig 8G ) , suggesting a high-clearance rate of dead cells in the repopulated brain and that the repopulated microglia are functionally intact . We next compared the percentage of TUNEL+ cells in 4 D microglia ( EdU+Iba1+ ) versus resident microglia ( EdU−Iba1+ ) . We found that the TUNEL+ signal was more frequently associated with EdU+Iba1+ cells than EdU−Iba1+ cells ( Fig 8H and 8I ) , suggesting that newborn microglia were more likely to undergo cell death . To rule out the possibility that the increased cell death might be an artifact of cell toxicity caused by EdU labeling , we compared the number of TUNEL+ cells in samples with or without EdU injection at 4 D and 14 D , which showed that EdU labeling did not skew the level of cell death ( S10 Fig ) . Thus , active cell death is closely associated with microglial proliferation , consistent with findings by Askew and colleagues [11] . Interestingly , we and others have consistently observed an overproliferative phenotype that resembles microgliosis during the first week of repopulation ( Fig 1B ) [8–10] . We hypothesized that the resolution of microgliosis is achieved by egress of extra newborn cells . To measure the turnover rate of the repopulated microglia , we applied a pulse-chase approach using EdU labeling during the first 4 days of repopulation to track newborn microglia and extended the recovery time to 4 months ( Fig 9A ) . As expected , the number of EdU+Iba1+ microglia decayed over time ( Fig 9B , 9C and 9D ) . By fitting the data with a nonlinear 1-phase decay model , we estimated the half-life of repopulated microglia to be 111 . 3 days ( Fig 9E ) . A much shorter lifespan was also determined for EdU+Iba1− cells ( T1/2 = 21 . 44 days ) . The nonlinear 1-phase decay model assumes no additional proliferation , which was confirmed by the absence of proliferation marker Ki-67 after day 4 ( S11 Fig ) . We found that the turnover rate of the EdU+Iba1+ microglia was well correlated with egress of excessive microglia ( Fig 9F ) . Therefore , based on the decay rate of newborn microglia , microgliosis in the repopulated brain required up to 2 months to fully resolve ( Fig 9G ) . Microgliosis resolution in the repopulated brain is also likely to be facilitated by microglia self-engulfment , as we observed examples of microglia forming phagocytic cups around dead microglia in both 14 D and 4 Mo brains ( S12 Fig and S3–S6 Videos ) . Thus , restoration of microglial density seemed to involve steady turnover of newborn cells .
Our current study employed a chemical microglial ablation approach to query the mechanistic details of microglial homeostatic regulation . Remarkably , microglia seem to have inherent memory of their steady-state signature , as repopulated microglia eventually became almost identical to those in the resting state . The data presented here piece together a series of events that take place during microglial homeostatic restoration in adulthood ( Fig 10 ) . First , the depleted microglial pool expands through self-renewal that partially requires NF-κB signaling . Second , microglia recolonize the parenchyma from proximal clonal expansion and maintain stable spatial clusters . Third , newborn microglia re-establish maturity in a process that involves mitosis , apoptosis , interferon pathway activation , and surface molecule re-expression . Finally , homeostatic density is reached after egress of excessive newborn cells . The existence of CNS-resident microglial precursors cells has been controversial [8] . In line with other studies [7 , 9 , 10 , 39] , our results demonstrate that the microglial pool can be replenished solely by self-renewal , with no detectable contribution from either resident CNS progenitors or peripheral circulating precursors . In the BMT chimera experiment , we observed most of the GFP+Iba1+ cells in the choroid plexus ( Fig 2E ) , in line with recent findings that choroid plexus macrophages represent a distinctive parenchyma myeloid population that constantly receives cellular exchange with circulating monocytes [1] . Consistent with a previous report by Elmore and colleagues [8] , we found a group of nonmicroglial cells that are highly proliferative after release from CSF1 inhibition ( Fig 1D and S2 Fig ) , indicated by EdU labeling . Testing against a panel of common CNS markers , these cells appeared negative for Iba1 , GFAP , NeuN , and S100β ( Fig 1M ) . Although a small subset of them appeared to be DCX+ or Olig2+ , the vast majority of cells may represent an uncharacterized cell type that remains to be further investigated . Nevertheless , our lineage-tracing experiments using CX3CR1-CreERT2 unambiguously demonstrate that this population does not represent microglial precursors . Spatial characteristics of microgliogenesis were largely inaccessible in the past . Stochastic labeling in microglia using the Microfetti approach now opens new possibilities to extract spatial information during microglial tiling [22] . Using this approach , Tay and colleagues reported that microglia underwent clonal expansion in response to facial nerve injury [22] . In agreement with their finding , we found microglia also proliferate through clonal expansion following ablation . We also observed that not all seeding cells initiated clonal expansions , as evidenced by few scattered cells positioned between clusters ( Fig 4C , 4D and 4F ) . This might reflect the heterogenous nature of microglia , arguing for the potential existence of a subpopulation of microglia with differential “stemness . ” To examine the stability of microglial tiling , we followed the retiled parenchyma over the course of 1 month . We found that microglia only migrate during clonal expansion and that tiling boundaries remained stable up to 1 month , suggesting that once microglial expansion is completed , the tiling pattern appeared to be static . Originating from yolk sac progenitors , microglia follow a series of transcriptional programs to reach maturity [25] . Recent studies have shown that adult microglial maturity is highly influenced by inflammatory signals [25 , 40] . Exposure to immune-activating agents during pregnancy has also been shown to shift microglia to a more mature stage [25] . In contrast , adult microglia deprived of any immune stimuli ( using a germ-free environment ) display an immature morphology , immature transcriptional signatures , and a dampened response to lipopolysaccharides ( LPS ) challenge [25 , 40] . These results suggest that microglial maturity is not a simple continuum of their birth age but rather is shaped by external factors such as immune activation . In support of a role for immune signaling in maturation , we found that genes associated with the NF-κB pathway were highly enriched during the early phase of microglial repopulation ( Fig 6C ) . Although the effects of microglia depletion under disease conditions have been investigated [41–43] , how NF-κB activation in repopulating microglia would impact neuropathology is unknown and needs to be addressed in future studies . Inhibiting the NF-κB pathway via conditional knockout of IKKβ resulted in partially impaired microglial repopulation ( Fig 7 ) . Furthermore , we and others have found that adult microglial renewal is closely associated with cell death [11] . More studies are needed to clarify the biological significance for cell death during the maturation of newborn microglia . Microglia have previously been considered a stable population [44] , but this view has been recently challenged [11] . By quantifying the steady-state microglial proliferation rate , it was estimated that it takes 96 days to renew the entire rodent microglial pool [11] . However , a recent study by Tay and colleagues used a similar approach and concluded that microglia are extremely stable , with a steady-state turnover rate of over 41 months in the cortex , 15 months in the hippocampus , and 8 months in the olfactory bulb [22] . Although this study examined microglia in an unperturbed state , there are several caveats to its interpretation . First , assumptions regarding microglial cell cycle duration were made . Second , the labeling efficiency associated with nucleotide analogs such as EdU or Bromodeoxyuridine/5-bromo-2'-deoxyuridine ( BrdU ) might not capture all proliferating microglia , resulting in an underestimation of the turnover rate . Third , sampling power was greatly limited since only a handful of cells could be captured at a given time . To overcome these limitations , we used a pulse-chase strategy with EdU labeling to directly track the decay of repopulated microglia and showed that they have a half-life of 111 . 3 days . ( Fig 9E ) . In other words , adult-born microglia have a lifespan of 7 . 5 months , such that a rodent brain could renew its entire microglial pool roughly 5 times in its lifetime . However , since our measurement started from repopulation day 14 when microglial homeostasis was not yet fully settled , 7 . 5 months may still be an underestimate of the actual microglial longevity . Indeed , a recent study by Füger and colleagues recorded microglial longevity via single-cell imaging and reported a lifespan of approximately 15 months [45] . In humans , microglial longevity is estimated be several decades long [46] . Thus , the egress of microglia that occurs following repopulation can be extremely slow and take months to fully resolve . In a way , this emphasizes the persistent nature of microgliosis in the context of sterile neuroinflammation . Therefore , strategies that can accelerate microglial turnover might be an overlooked avenue to treat neurological diseases for which microgliosis is the driving force . The tremendous stability associated with microglial density over the lifetime in a rodent brain is quite remarkable [11] . Even after acute ablation , microglia are able to return to the same homeostatic density . The mechanisms by which microglia are able to remember their homeostatic density is still unclear . One possible way could be due to contact inhibition . After microglial ablation , residual microglia isolated in space would lose contact inhibition exerted by neighboring microglia , thus allowing them to freely multiply . The observed overproliferation ( Fig 1A ) [8–10] suggests that the inherent regulatory mechanism responsible for controlling density is switched off during microglial repopulation . Coincidentally , some surface molecules were still expressed at lower levels 1 month after repopulation including Syndecan-4 , an important regulator of contact inhibition during cancer metastasis [34] . These genes are of great interest for future investigations ( S1D Table ) . Nevertheless , the exact molecular machinery that sets the default homeostatic density remains to be characterized .
All animal work was performed according to the approved guidelines from the University of California , San Francisco , Institutional Animal Care and Use Committee ID number ( AN173162 ) . Mice with ad libitum access to food and water were housed in a pathogen-free barrier facility with 12-hour light on/off cycle . C57BL/6J mice were used as wild-type controls . Equal numbers of male and female mice were used for all experiments except for the RNA-seq experiment , which used only female mice . Different mouse lines used in the study can be found in S2A Table . Diet containing 1 , 200 mg/kg PLX5622 ( Plexxikon Inc . , Berkeley ) was given to mice as the sole food source for 2–3 weeks to deplete microglia . Control diet with the same base formula but without the compound was given to the control group . Tamoxifen ( Sigma-Aldrich , T5648 ) was prepared in corn oil before use at 20 mg/mL . To efficiently induce Cre recombination , mice were given tamoxifen via intraperitoneal ( IP ) injection at 2 mg per day for 10 days . For sparse labeling in CX3CR1-CreERT2/Brainbow mice , tamoxifen was given via daily IP injection for 4 days . Mice receiving tamoxifen injections were housed for at least an additional 21 days after the last injection and before use . For labeling newborn cells , EdU ( Santa Cruz , sc-284628 ) was prepared in sterile PBS at 20 mg/mL . EdU solution was warmed to 55°C before use to dissolve any chemical precipitation . Mice were injected at 80 mg/kg via IP diphtheria toxin ( Sigma-Aldrich , D0564 ) and was resuspended in sterile PBS at 20 ug/mL working solution for injection . Diphtheria toxin was delivered to animals via IP injection at 1 ug per day for 3 days . Bone marrow was isolated from the tibias and femurs of 3–6 months old ACTB-eGFP transgenic mice . The bone marrow was triturated using an 18-gauge needle and passed through a 70-μm nylon mesh cell strainer to make a single-cell suspension . Erythrocytes were lysed with ACK lysis buffer ( 150 mM NH4Cl , 10 mM KHCO , and 0 . 1 mM Na2EDTA ) , washed with PBS , and suspended in PBS with 0 . 1% BSA at 3 . 5 × 107 cells/mL . Recipient mice that were 4 to 6 months old were given 2 equal doses ( 2 × 600 rads , 3 hours apart ) of irradiation with a cesium source irradiator . To protect the brain from radiation , the head of the mouse was shielded by a lead plate during irradiation . Irradiated mice were reconstituted with 7 × 107 cells of donor bone marrow via tail vein injection . Repopulation efficiency was determined by counting the percentage of GFP-positive myeloid population ( CD45+CD11b+ ) by flow cytometry . Blood and spleen samples were collected in EDTA . Erythrocytes were lysed in FCK lysis buffer for 5 min , and sample was pelleted at 500 g . Leukocytes were resuspended in FACS buffer ( PBS , 0 . 5% BSA , 5% FBA , and 0 . 1% NaN3 ) and incubated for 15 min with anti-CD16/CD32 monoclonal antibodies ( 1:200 , BD PharMingen ) to block Fc receptors . Cells were then stained with APC-labeled anti-CD11b ( 1:100 , Biolegend ) and PE-labeled anti-CD45 ( 1:100 , Biolegend ) for 30 min on ice , followed by fixation with 2% PFA . Fluorescence intensity was measured using a FACS Calibur ( BD Biosciences ) flow cytometer . Data were analyzed in FlowJo ( V10 . 0 . 7 ) . Adult microglia isolation was performed using MACS , as previously described [47] . Briefly , mice were anesthetized with avertin and transcardially thoroughly perfused with PBS to remove circulating blood cells in the CNS . Each dissected brain was chilled on ice and then minced in enzymatic digestion buffer containing 0 . 2% Collagenase Type 3 ( Worthington , LS004182 ) and 3 U/mL Dispase ( Worthington , LS02104 ) . Minced brain tissue was then incubated at 37°C for 45 min . The enzymatic digestion was stopped with inactivation buffer containing 2 . 5 mM EDTA ( Thermofisher , 15575020 ) and 1% fetal bovine serum ( Invitrogen , 10082147 ) . The digested brain tissue was then triturated in a serological pipette several times before passing through a 70-μm filter . The homogenate was then depleted of myelin using myelin removal beads ( Miltenyi Biotec , 130-096-733 ) and magnetic LD column ( Miltenyi Biotec , 130-042-901 ) . The elute was enriched for microglia with CD11b magnetic beads ( Miltenyi Biotec , 130-049-601 ) and MS column ( Miltenyi Biotec , 130-042-201 ) . Mice were anesthetized with avertin and transcardially perfused with PBS . Whole brains were drop-fixed in 4% paraformaldehyde prepared in PBS for 48 hours before switching to 30% sucrose for at least another 48 hours before cutting on a sliding microtome ( Leica , SM2010R ) . Brain sections were prepared in a stereological manner . Sequential coronal planes in 30-μm thickness were collected and preserved in cyroprotectant containing 30% glycerol , 30% ethoxyethanol , and 40% PBS . Brain tissue samples were stored in −20°C before use . All staining experiments were performed on slices that were collected at the similar coronal planes . One or 2 coronal sections per mouse were used for each staining . Free-floating sections were washed in PBS and then permeabilized in PBST buffer ( 0 . 5% Triton X-100 diluted in PBS ) , followed by blocking in 3% normal donkey serum ( NDS ) at room temperature for 1 hour . Primary antibodies were diluted in PBST containing 3% NDS and incubated with tissues at 4°C overnight . Primary antibodies used in the study can be found in S2B Table . Secondary antibodies were then prepared the same way as primary antibodies and incubated with tissue at room temperature for 1 hour . All secondary antibodies were obtained from Jackson ImmunoResearch . Secondary antibodies used in the study can be found in S2B Table . After secondary antibody staining , DAPI nuclear stain was conducted during the washing step as needed . Tissues were then mounted on glass slides for further processing or applied with anti-fade mounting media ( Vector Laboratories , H-1000 ) for imaging . For detecting EdU+ cells in brain sections , Click-iT EdU-imaging kits ( ThermoFisher Scientific , C10337 , C10339 ) were used following the manufacturer’s instructions before the immunofluorescence staining procedures . For terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling ( TUNEL ) , the DeadEnd Fluorometric TUNEL System ( Promega , G3250 ) was used with minor alterations to the manufacturer’s recommendations . We adapted a method developed by Deng and colleagues [48] . In brief , after the immunofluorescence staining was finished , free-floating brain sections were mounted on a charged glass slide ( Fisher , 12-550-15 ) and dried at 55°C for 5 mins . Tissues were outlined with a hydrophobic pen ( Vector Laboratories , H-4000 ) . Mounted glass slides were then incubated in 0 . 5% Triton/PBS at 85°C for 20 min . After cooling to room temperature , the equilibration buffer from the kit was applied directly to the slide for 5 min , followed by application of TUNEL reaction solution and incubation at 37°C for 1 hour . Regular epifluorescence images were acquired on a Keyence BZ-9000 inverted epifluorescence microscope equipped with an RGB and monochrome camera ( Keyence , Osaka , Japan ) . Either the entire coronal hemibrain slice or particular brain regions were scanned using 10x magnification and stitched in Keyence BZ-X Analyzer software ( V1 . 3 . 0 . 3 ) . Confocal microscopy was performed with a Zeiss LSM880 inverted scanning confocal microscope ( Carl Zeiss Microscopy , Thornwood , New York ) equipped with 2 PMT detectors , a high-sensitivity GaAsP detector , and a 32-GaAsP Airyscan super-resolution detector and run by Zeiss Zen imaging software . Confocal fluorescence images were acquired with 5–10 focal planes at 1–2-μm intervals . Representative images are shown using Z-max intensity projections . All image analyses were performed in FIJI V1 . 50i [49] . Image analyses codes used are available on GitHub ( https://github . com/lihong1github/Image-analysis ) . In brief , TIFF images were processed with adaptive threshold function to generate binary cell masks ( https://sites . google . com/site/qingzongtseng/adaptivethreshold ) . Then , the “Analyze Particles” function was used for cell counting . Cell density ( per mm2 ) was calculated by normalizing cell number to the size of the analyzed area . For analyzing cells with multiple markers , the region of interest ( ROI ) was first generated from the base channel . Channels containing other markers of interest were thresholded to generate binary images . Each ROI generated from the base channel was then overlaid on to the binary images to calculate percent area of the marker of interest . In general , >80% overlapping area is used to select ROIs that are positive for a second marker . To analyze microglia 3D morphology , confocal z-stacks with 8 focal planes of 2-μm interval were used . Three 425 . 1 × 425 . 1-μm2 connecting fields of view spanning the hippocampal CA1 to CA2 were captured and stitched in ZEN Blue software ( Zeiss ) . The stitched tile images were analyzed in IMARIS software ( V9 . 0 . 2 , Bitplane ) . Microglia processes were analyzed using IMARIS filament function . Microglia soma size were measured using IMARIS surface function . Heterozygous CX3CR1eGFP/+ mice of 3–3 . 5 months old were treated for 2 weeks with a PLX-containing diet and then switched to a regular diet to allow microglial repopulation for 6 days . Acute brain slices were prepared , as previously described [47] . Mice were anesthetized with isoflurane and perfused with 20 mL of ice-cold , carbogen-saturated ( 95% O2 , 5% CO2 ) NMDG-HEPES artificial cerebrospinal fluid ( ACSF ) solution containing 93 mM NMDG , 2 . 5 mM KCl , 1 . 4 mM NaH2PO4 , 30 mM NaHCO3 , 20 mM HEPES , 25 mM D-glucose , 5 mM ascorbic acid , 2 mM thiourea , 3 mM sodium pyruvate , 12 mM N-acetyl-L-cysteine , 10 mM MgSO4 , and 0 . 5 mM CaCl2 . The ACSF was adjusted to pH 7 . 4 before use . All reagents were obtained from Sigma . Brains were washed in NMDG-HEPES . Coronal slices ( 300-μm thick ) were prepared using a vibratome ( Microm , Walldorf , Germany ) at 4°C . For imaging , hemibrain slices were transferred to a 35-mm glass bottom dish ( MatTek ) and secured under a slice anchor ( Warner instruments ) . Imaging sessions were conducted within 4 hours after decapitation and preparation of the slice in order to reduce expression of activation markers by microglia . All imaging was done in non-NMDG ACSF solution containing ( in mM ) NaCl 92 , KCl 2 . 5 , NaH2PO4 1 . 4 , NaHCO3 30 , HEPES 20 , D-glucose 25 , Ascorbic acid 5 , Thiourea 2 , Sodium pyruvate 3 , N-acetyl-L-cysteine 12 , MgSO4 2 , CaCl2 2 , and pH 7 . 4 ( all from Sigma ) . Carbogen-saturated ACSF flowed continuously over the slice using a perfusion pump system ( Cole-Parmer ) at a rate of 3 mL/min . Images were captured with a Zeiss Z1 Observer inverted epifluorescence microscope ( Carl Zeiss Microscopy , Thornwood , New York ) with an ORCA-Flash 4 . 0 sCMOS camera ( Hamamatsu Photonics , Shizuoka , Japan ) and a Zeiss Axiocam MRm monochrome camera run by Zeiss Zen imaging software . The microscope is equipped with a motorized stage and temperature-controlled incubation system . Acquisition of cortical microglia was performed at a range of 50 to 100 μm from the surface of the slice to prevent capture of activated microglia , and z-stack images ( 1-μm step-size ) were taken every 60 seconds for 15 to 30 min at 488-nm excitation and 510-nm emission wavelengths using a 40x objective . The motility of microglia processes was analyzed using the ImageJ plugin MTrackJ [50] to calculate the velocity of process retraction and extension per imaged cell . Hemibrain slices from CX3CR1-CreERT2/Brainbow mice were stained with anti-GFP and anti-RFP antibody . The entire coronal section was scanned using a Keyence BZ-9000 inverted epifluorescence microscope . RFP+ or GFP+ cells were segmented , and the XY coordinates were extracted using centeroids function in FIJI . NNDs were calculated with ImageJ plugin NND ( https://icme . hpc . msstate . edu/mediawiki/index . php/Nearest_Neighbor_Distances_Calculation_with_ImageJ ) . NND of the labeled cells was extracted from an entire hemibrain coronal slice with area of 25 . 5 ± 1 . 03 mm2 . Spatial point pattern clustering analysis with Ripley’s K-function was performed , as previously described [51] . The K-function is expressed as Eq 1 , in which n is the number of total cells , r is the varying radius , NPi ( r ) is the regional density of the ith cell at radius r , and λ is overall cell density . The overall degree of “Clusterness” is estimated as a cumulative function of the radius r . A negative score in H-function indicates patterns of dispersion , while a positive score indicates patterns of clustering . Theoretical complete spatial randomness ( CSR ) modeled by Poisson distribution in K ( r ) equals πr2 . H-function , as shown in Eq 2 , is transformed from the Ripley’s K-function for improved data visualization . In H-function , CSR equals zero . H ( r ) function was computed using Kest function in R package Spatstat [52] . H ( r ) values from each mouse were calculated separately . Average H ( r ) were then computed for each treatment group . Smoothing function using a moving average with interval of [r − 15 μm , r + 15 μm] was applied . Domain size was estimated from the corresponding radius at maximum of H ( r ) [24] . To define cluster boundaries , 2D kernel density estimates were calculated by using density function in Spatstat package . Spatial points of RFP+ or GFP+ cells were smoothed with 100 μm2 Gaussian kernel to generate density estimates . Regions containing the highest top 10% kernel density were chosen to generate cluster boundaries , as this parameter best matches expected cell density . Overall size shift of the RFP+/GFP+ cluster overlapping region were quantified as percent of RFP and GFP overlapping area relative to the combined nonoverlapping RFP and GFP area . RNA was extracted using Direct-zol RNA micro-prep kit ( Zymo Research , R2061 ) following the manufacturer’s instructions . Roughly 100 , 000 microglia cells were lysed in 300 μl of Direct-zol reagent . Genomic DNA digestion was performed on column using DNaseI ( Zymo Research , E1010 ) . Extracted RNA ( approximately 500 ng ) was then converted to cDNA using iScript cDNA synthesis kit ( Bio-Rad , 1708890 ) and diluted 10-fold with nuclease free water as template DNA . qPCR was performed using SYBR Green PCR Master Mix ( Applied Biosystems , 4309155 ) . Data were acquired using 7900HT real-time system equipped with a 384-well thermal block ( Applied Biosystems , Forster City , United States of America ) . Raw Ct values were obtained using the Sequence Detection Systems software ( Applied Biosystems , version 2 . 4 ) . Relative gene expression was calculated based on the delta–delta Ct method using qbase+ software ( Biogazelle , version 3 . 0 ) . GAPDH was used as reference gene . Primer sequences can be found in S2C Table . Female wild-type C57BL6 mice of 5–6 months were used . To identify differential gene expression during microglial repopulation , mice were treated for 2 weeks with PLX-containing diet and then switched to regular diet allowing microglial repopulation for various durations before being sacrificed for microglia isolation . These repopulation duration time points were set at 4 days ( 4 D ) , 14 days ( 14 D ) , and 1 month ( 1 Mo ) . P4 microglia were isolated from 4-day-old postnatal mice as an immature microglia control . Microglia isolated from each single mouse were used as an individual sample for downstream steps . A total of 3 pups ( P4 ) , 4 untreated adult mice ( Ctrl ) , and 4 repopulated mice from each designed time point were used . Total RNA from fresh isolated microglia was extracted using Direct-zol RNA microprep kit ( Zymo Research , R2061 ) . cDNA library generation and RNA-seq service was performed by Novogene ( Novogene Co . , Ltd , Sacramento , California ) . RNA quality was examined by Bioanalyzer 2100 ( Agilent Genomics ) . RNA samples with RIN value greater than 8 were used for cDNA library . Oligo ( dT ) beads were used to enrich for mRNA . After chemical fragmentation , a cDNA library was generated using NEBNext Ultra RNA Library Prep Kit for Illumina ( New England Biolabs , E7530S ) . Quality of the cDNA library was assessed using Qubit assay for preliminary concentration , Bioanalyzer 2100 for insert size , and qPCR for effective library size . QC passed cDNA library samples were then sequenced with the HiSeq 4000 system ( Illumina ) at PE150 . On average , less than 0 . 01% error rate was detected , and over 95% effective rate was observed in all sequencing results . Raw read ends containing low-quality reading or adapter sequence were trimmed prior to downstream analysis . RNA-seq read mapping was performed using the STAR program [53] . Gencode mouse genome GRCm38 was used as reference ( release M16 , 2017 ) . On average , roughly 80% of reads were uniquely mapped to the reference genome . The read count table was generated with the RSEM program [54] . Differential gene expression was calculated with R package edgeR [55] and limma [56] . Genes which showed less than 1 count per million ( CPM ) in at least 3 samples were filtered out from further analysis . Normalization was performed with the weighted trimmed mean of M-values ( TMM ) method [57] . RNA-seq data were then transformed for linear model fitting with voom and lmFit functions inside the limma package . Finally , empirical Bayes statistics were applied to correct variance of genes with low expression in the data set . FDR was calculated by the Benjamini–Hochberg method [58] . DE genes were defined as 2-fold change with FDR less than 0 . 05 in comparison to unperturbed adult microglia ( Ctrl ) . Gene network analyses were performed with GSEA with molecular signatures database ( MSigDB ) [30 , 31] . Network visualization was made in Cyotscape ( version 3 . 6 . 0 ) with access to STRING database [59] . All experiments were performed with a minimum of at least 3 biological replicates . All data were averaged to individual animals . Mean values from each animal were used for computing statistical differences . Error bar in plots indicates standard error of the mean ( SEM ) . Statistical analyses were performed in Graphpad prism 7 . 0c ( Graphpad , San Diego , California ) and R ( F Foundation for Statistical Computing , Vienna , Austria ) . Data visualization were achieved with R package ggplot2 [60] . Data normality was assessed using the Shapiro–Wilk normality test . F test was used to assess homoscedasticity prior to unpaired t test , and Brown–Forsythe test was used to homoscedasticity prior to ANOVA . For data with normal distribution and equal variance , unpaired t test was used to compare 2 groups . One-way ANOVA was used to compare data with more than 2 groups . Dunnett's multiple comparisons test was used to compare difference between designated groups . Two-way ANOVA was used for groups with genotypes and treatment as factors . Sidak's multiple comparisons test was used to compare statistical difference between genotypes . For data that failed to pass the normality test , Mann–Whitney test ( for 2 groups ) and Kruskal–Wallis test ( >2 groups ) were applied . For data with unequal variance , unpaired t test with Welch's correction were applied . P value and FDR are summarized as ns ( P > 0 . 05 ) , * ( P ≤ 0 . 05 ) , ** ( P ≤ 0 . 01 ) , *** ( P ≤ 0 . 001 ) , and **** ( P ≤ 0 . 0001 ) .
|
Microglia are the brain’s primary immune sentinels . After being seeded in the central nervous system ( CNS ) during embryonic development , microglia form an evenly distributed grid-like population in the adult brain . In the adult , microglia have the remarkable ability to regenerate and restore their population following acute insults , but what drives this homeostatic process is unclear . Here , we addressed this question by studying regeneration following chemical ablation of microglia . In this model , we addressed the origin of repopulating cells , their spatial redistribution characteristics , and the temporal spectrum of gene-expression patterns at different stages of repopulation . We found that the restoration of microglial homeostasis during repopulation is achieved through self-renewal , proximity clonal expansion , and activation of maturation programs .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"nervous",
"system",
"cell",
"processes",
"diet",
"microglial",
"cells",
"physiological",
"processes",
"developmental",
"biology",
"nutrition",
"homeostasis",
"microscopy",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"gene",
"expression",
"scanning",
"electron",
"microscopy",
"glial",
"cells",
"cell",
"biology",
"anatomy",
"central",
"nervous",
"system",
"physiology",
"apoptosis",
"genetics",
"electron",
"microscopy",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neonates"
] |
2019
|
Proximal recolonization by self-renewing microglia re-establishes microglial homeostasis in the adult mouse brain
|
Theoretical methods for predicting CD8+ T-cell epitopes are an important tool in vaccine design and for enhancing our understanding of the cellular immune system . The most popular methods currently available produce binding affinity predictions across a range of MHC molecules . In comparing results between these MHC molecules , it is common practice to apply a normalization procedure known as rescaling , to correct for possible discrepancies between the allelic predictors . Using two of the most popular prediction software packages , NetCTL and NetMHC , we tested the hypothesis that rescaling removes genuine biological variation from the predicted affinities when comparing predictions across a number of MHC molecules . We found that removing the condition of rescaling improved the prediction software's performance both qualitatively , in terms of ranking epitopes , and quantitatively , in the accuracy of their binding affinity predictions . We suggest that there is biologically significant variation among class 1 MHC molecules and find that retention of this variation leads to significantly more accurate epitope prediction .
In order to make the prediction values comparable between each MHC molecule , it is recommended that the MHC-peptide binding affinity scores are rescaled [28]; this is explicitly implemented in NetCTL . The method of rescaling involves obtaining the predicted binding affinities of 500 , 000 random natural peptides for each MHC allelic predictor . From these affinities , a rescale value is calculated , defined as the binding affinity that is the threshold for the top 1% of total binding affinities . The rescaled affinity is then defined as the predicted affinity score divided by this rescale value [3] . Hence , from this calculation , all alleles are predicted to bind the same number of high-affinity peptides . One pragmatic reason for rescaling is to correct for any discrepancies between the allelic predictors that resulted from inconsistent training data ( e . g . data that came from different sources ) , by assuming that all alleles should bind the same number of epitopes ( C . Keşmir , pers . comm . ) . Additionally , there are biological arguments for believing that different alleles should bind similar numbers of epitopes . It has been postulated that the opposing constraints of effective pathogen recognition but tolerance of self would result in a very narrow range of optimal promiscuity for viable MHC class I molecules . A narrow range of promiscuity would also be predicted as a direct outcome of effective tapasin-dependent peptide optimization in the endoplasmic reticulum [29] , [30] , [31] . However , we will present evidence in this paper that in correcting for differences between the allelic predictors , information is being lost that reflects true biological variation between MHC molecules and , by extension , differences in their ability to bind to peptide sequences . We show that , for both qualitative and quantitative measures of binding , rescaling impairs rather than improves allelic predictor performance . This is of importance for vaccine design and to understand the nature of the CTL response . In particular , crucial between-allele variations in binding affinity and preference which may contribute to differences in the outcome of infection are likely to be obscured by rescaling .
In order to test the effect of rescaling on epitope prediction accuracy , we used two web-based prediction methods , NetCTL v1 . 2 [3] and NetMHC v3 . 0 [25] , [26] , [21] . NetCTL is an integrated method that uses information pertaining to TAP and protein cleavage in its predictions , together with MHC binding . The output is combined by rescaling the MHC binding result and adding this to the weighted scores for TAP and protein cleavage . NetCTL has allelic predictors for 12 different class I alleles that are chosen to be representative of each of 12 supertypes; hence it has 12 different rescaling factors . NetMHC v3 . 0 simply predicts MHC-peptide binding , using ANNs to predict binding affinities for 43 MHC molecules . In order to test the effect of rescaling , it was necessary to produce rescale values for each of the 43 allelic predictors . This was performed as in NetCTL; 500 , 000 unique random nonamers were obtained from the proteome of Mycobacterium tuberculosis , their binding affinity was predicted and the rescale value ( top percentile ) was found for each allelic predictor . We also performed this calculation with 500 , 000 random natural peptides to test for the possibility of error from bias in amino acid usage in Mycobacterium tuberculosis . There was no significant difference in the rescale values obtained using these two different sources ( supplementary material , figure S4 ) . In summary , we tested two sets of rescaling values: those obtained from NetCTL v1 . 2 and those that we calculated using NetMHC v3 . 0 . Epitope datasets were constructed from sources detailed below . In each case , the prediction methods were tested by their ability to detect these epitopes amongst the full set of overlapping nonamers derived from the proteins that contained the epitopes . The full set of nonamers will contain a small number of known epitopes and the remainder will be ‘non-epitopes’ . Of course , this set of non-epitopes could include epitopes that have not been experimentally verified . However , the majority ( see Introduction ) would be non-binders with the corresponding MHC molecule . Added to this , the labelling of epitopes as ‘non-epitopes’ impact on both rescaled and non-rescaled calculations equally . Previous research has also shown that this property of the ‘non-epitope’ set did not produce significantly different results [24] . Each respective set of experimentally defined epitopes was denoted the positive dataset and the set of non-binding ( or unknown ) peptides was denoted the negative dataset . The SYF1 dataset is a supertype dataset derived from SYFPEITHI [20] and is identical to that used in the original paper for NetCTL [3] . Each epitope in SYF1 was experimentally verified to bind to one of 10 MHC class I supertypes [32] . The resulting dataset consisted of 148 epitope-supertype pairs . The corresponding negative dataset was obtained by concatenating the SwissProt entry proteins from which each of the epitopes was derived . The length of the concatenated protein sequence was 78 , 259 amino acids . The ROC curve ( see below for explanation ) was generated using a negative set of ( ( 78 , 259*10 ) −148 ) = 782 , 442 nonamers and a positive set of 148 nonamers . The positive set of SYF1 is available in the supplementary material ( dataset S2 ) . Experimentally defined epitopes in HIV-1 were extracted from the HIV Molecular Immunology Database [33] . In total , 1 , 618 CTL epitopes were found that were bound by human MHC molecules . However , this set was highly redundant; the epitope lengths were variable and a large number of epitopes differed only by mutations within the sequence . Also , resolution of their MHC typing varied from 2 to 4 digits . To correct for this variability , a number of changes were made to the MHC allele-epitope list . Firstly , all MHC alleles were defined to two digits . Secondly , variant epitopes binding the same allele were discarded . Finally , as the prediction software only produced binding predictions for nonamer epitopes , all epitopes that were not 9 amino acids long were removed from the list . In summary , it was possible to test 41 of the 43 allelic predictors for MHC molecules in NetMHC v3 . 0 . The positive set consisted of 661 epitopes , defined in terms of start and end positions relative to the HIV reference strain HXB2 ( supplementary dataset S1 ) and a matching MHC type to 2 digits . The input protein sequence to NetMHC contained 3 , 000 overlapping nonamers that covered the proteome from which the whole positive set of epitopes was derived . The total ‘negative set’ for the ROC analysis was ( 3 , 000 * 41 ) −661 ) = 122 , 339 nonamers , and a positive set of 661 nonamers . The positive set of Lanl661 is available in the supplementary material ( dataset S3 ) . The Lanl661 dataset was modified for testing with NetCTL . From these 661 epitopes , a total of 179 bound to the 12 alleles for which NetCTL has allelic predictors . The input sequence to NetCTL contained 3 , 000 overlapping nonamers . For this experiment , the negative set consisted of ( ( 3 , 000 * 12 ) −179 ) 35 , 821 nonamers , and a positive set of 179 nonamers . The positive set of Lanl179 is available in the supplementary material ( dataset S4 ) . ROC curves give a visual measure of the accuracy of a prediction method . The threshold at which the prediction method identifies a peptide as being an epitope varies along the length of the curve . Each point on the curve gives the fraction of true positive epitopes found as a function of the number of false positive ‘epitopes’ at that threshold . Hence , setting a strict threshold for epitope detection will result in high specificity ( correct predictions ) but low sensitivity ( missing a high proportion of true binders ) . The area under the ROC curve gives the AUC ( Area under Curve ) measurement . In order to test for significant difference between ROC curves , we conducted the bootstrapping analysis detailed in [34] . Briefly , using bootstrapping with replacement , 100 replicates were formed from each dataset and the resulting non-rescaled and rescaled whole AUC values were compared using a paired t test . Using the 2 epitope datasets , HIV216 and SYFPEITHI863 , and the same methods from [35] , we repeated 3 of the measurements described in that paper for the rescaled and non-rescaled results of NetCTL v1 . 2 . For the Rank measure , we analysed the proteins from which each epitope was derived . For each protein , we calculated the rank of the epitope amongst all overlapping 9-mers using rescaling and non-rescaling scoring methods for all alleles . We then analysed these ranks to see which method ranked the epitopes higher . For the second method , we measured the specificity of both rescaling and non-rescaling at predefined sensitivities . Finally , we measured the sensitivity among the top 5% top-scoring peptides , again for the rescaled and non-rescaled binding affinities . The training data for NetMHC v3 . 0 is available at http://mhcbindingpredictions . immuneepitope . org/ . An independent set of experimental epitope-allele binding affinities was obtained from the Immune Epitope Database and Analysis Resource ( IEDB ) by selecting all experimental data that did not originate from the laboratories of Sette et al . or Buus et al . ( the training data originated from these two sources ) .
ROC curves were used to analyse the effects of rescaling on epitope prediction . Both NetCTL v1 . 2 and NetMHC v3 . 0 were tested and 3 datasets were used ( figure 1 and table 1 ) . In each case , rescaling resulted in a significant loss of performance ( bootstrap test: p<0 . 001 ) . One possible explanation for why rescaling has a detrimental impact on prediction is that there may be a positive correlation between rescale factor and allelic predictor accuracy . To check this hypothesis we calculated the AUCs for each NetMHC v3 . 0 predictor using the Lanl661 dataset and plotted this against the corresponding rescale factor , the results of which are shown in figure 2 . This shows no evidence of a correlation between rescaling values and the AUC values ( R2 = 0 . 0068 , p = 0 . 606 ) . Consequently , it is unlikely that a correlation between rescale values and AUC values explains our findings . However , certain alleles like B0801 do have both a low rescale value and a low AUC . To double check that these poor accuracy predictors were not causing the inaccuracies in rescaled predictions we repeated our ROC curve analysis for Lanl661 without the low accuracy predictors ( those with an AUC value below 0 . 9; namely A6801 , A6802 , B3501 , B0702 , B0801 , B0802 and B4501 ) . In the remaining , reduced subset of predictors there was even less evidence for a correlation between AUC and rescale factor ( R2 = 0 . 0007 , p = 0 . 887 ) . For this subset of predictors the accuracy was still significantly better if rescaling was not applied ( figure S1; bootstrap test: p<0 . 001 ) and comparable to the ROC curve analysis using the full set of alleles ( figure 1C ) . Therefore , we believe there is no evidence to support the hypothesis that the reason rescaling is detrimental is because there is a correlation between rescale factors and AUC . We used 3 other metrics [35] to compare predictive performance with and without rescaling . Using 2 sets of experimentally-derived epitope-allele binding affinities , we also showed that the correlation between predicted and experimental affinities was weaker with rescaling than without ( supplementary figure S3 ) .
Rescaling is , in theory , a sound approach to improving epitope prediction and in particular comparability of predictions obtained using different allelic predictors . However , using a number of different measures of accuracy , in the context of two commonly used prediction methods , we have demonstrated that rescaling actually impairs rather than improves predictive performance and comparability . We suggest that rescaling predicted affinities results in a loss of information that outweighs any advantage gained in correcting for differences in training data . The first approach used ROC curve analysis and showed clear differences between rescaling and non-rescaling . The ROC curve gives a graphical representation of how well the prediction method ranks true epitopes among a set of non-binding peptides . Or to use an analogy , how efficient it is at finding the epitopic needle in a haystack of random peptides . From figure 1 , it is clear that rescaling across all allelic predictors results in a performance loss in terms of how well the method ranks its peptides by binding affinity; that is , rescaling impairs intra-allelic comparisons . This loss could be demonstrated using epitope data from a number of sources ( SYFPEITHI , the HIV Molecular Immunology Database ) and with two different methods of prediction ( the combined approach of NetCTL v1 . 2 and NetMHC v3 . 0 ) . This effect of rescaling would be detrimental to any studies screening across a number of alleles for possible epitopes ( such as [15] ) . The effect of this performance difference can be gauged from figure 1 ( A ) . In order to identify correctly 85% of the epitopes the percentage of false positives detected was 9% and 15% , for non-rescaled and rescaled methods respectively . To put this result into context , the viral protein NS1 from the H5N1 strain of Avian Influenza A consists of 221 overlapping nonamers . To screen this protein for potential epitopes , 33 epitopes would need to be experimentally checked for each MHC molecule of interest if rescaled predictions were used , as opposed to 20 for the non-rescaled predictions ( providing 85% epitope coverage was sufficient ) . Added to the significant results from the ROC curve analysis , the supplementary analysis demonstrated the positive effect of removing rescaling in terms of the correlation with experimental data ( supplementary figure S3 ) and also in terms of per-protein and sensitivity analysis ( supplementary figure S2 and tables S1 and S2 ) . Taken together , these results strongly demonstrate the improvement in accuracy of removing the condition of rescaling when comparing predictions between alleles . There has been little research on the variation in ‘stickiness’ among MHC molecules , i . e . whether some MHC class I molecules are capable of binding to a greater number of epitopes than others . The binding motifs for MHC-peptide binding vary across the range of alleles , but the assumption made for rescaling is that each molecule would bind to the same number of peptides out of a large random selection . Estimates based upon mass spectrometry suggest that over 2 , 000 peptides are associated with HLA-A2 . 1 and −B7 and it is speculated that the actual total could be over 10 , 000 per MHC molecule [36] . However , it is not known how this number varies between molecules . It has been postulated that the twin constraints of effective pathogen recognition but tolerance of self would result in a very narrow range of promiscuity for viable MHC class I molecules [29] . Contrary to this , recent research has shown that this range may be wider than initially envisaged [37] and our results suggest that there is considerable inter-allelic variation in promiscuity . This data may also be informative regarding optimization of peptide cargo in the endoplasmic reticulum ( ER ) . We would argue that peptide optimization is the biological interpretation of rescaling: alleles have similar numbers of epitopes because peptides with a lower binding affinity are replaced in the ER . We know that optimisation cannot be complete because otherwise every allele would just present one epitope: the one with highest affinity . However , it seems likely that there is a degree of optimization [30] , [31] . The observation that rescaling gives worse predictions may put a bound on how much optimisation is occurring . Allied to this , it has been observed that the release of an MHC class I molecule from the peptide-loading complex with a suboptimal peptide takes precedence over the prolonged detention of the MHC class I molecule in the complex until an optimal peptide comes along [30] . Hence , peptide optimization acts to reduce inter-allelic variation and promiscuity results from inter-allelic variation in allele-peptide affinity . However , this peptide optimization is limited by time and is not complete and hence , we note this variation in promiscuity across different alleles . In summary , we suggest that much of the observed variation between allelic predictors reflects genuine biological information which should not be discarded as experimental noise and that rescaling is based on an unjustified assumption: that all alleles bind the same number of peptides . Removing this assumption , we have demonstrated a significantly improved predictive performance . These conclusions are important both for studies that use prediction methods to understand the CTL response and for T cell epitope discovery programs where avoiding rescaling could save a large amount of experimental effort , ultimately leading to improved vaccine implementation .
|
The use of prediction software has become an important tool in increasing our knowledge of infectious disease . It allows us to predict the interaction of molecules involved in an immune response , thereby significantly shortening the lengthy process of experimental elucidation . A high proportion of this software has focused on the response of the immune system against pathogenic viruses . This approach has produced positive results towards vaccine design , results that would be delayed or unobtainable using a traditional experimental approach . The current challenge in immunological prediction software is to predict interacting molecules to a high degree of accuracy . To this end , we have analysed the best software currently available at predicting the interaction between a viral peptide and the MHC class I molecule , an interaction that is vital in the body's defence against viral infection . We have improved the accuracy of this software by challenging the assumption that different MHC class I molecules will bind to the same number of viral peptides . Our method shows a significant improvement in correctly predicting which viral peptides bind to MHC class I molecules .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"immunology/antigen",
"processing",
"and",
"recognition",
"immunology/immunity",
"to",
"infections",
"computational",
"biology"
] |
2009
|
T-Cell Epitope Prediction: Rescaling Can Mask Biological Variation between MHC Molecules
|
The homeostasis of meristems in flowering plants is maintained by cell-to-cell communication via CLE ( CLAVATA3/EMBRYO SURROUNDING REGION-related ) peptide hormones . In contrast , cell signals that regulate meristem activity remains elusive in bryophytes that maintain apical meristems in the gametophyte ( haploid ) body and undergo a gametophyte-dominant life cycle . We here show that MpCLE1 confines the proliferative activity of gametophytic meristem and affects the overall size of gametangiophores ( reproductive organs ) in Marchantia polymorpha , which is in sharp contrast with the meristem-promoting function of its ortholog TDIF/CLE41/CLE44 in Arabidopsis vascular meristems . Expression analysis suggests that MpCLE1 and its receptor gene MpTDR are expressed in distinct patterns across the apical meristem . These data suggest that local CLE peptide signaling may have had a role in regulating cell proliferation in the shoot meristem in the ancestral land plant and acts in both sporophytic and gametophytic meristems of extant plants .
Land plants have evolved unique peptide hormones to control various physiological processes including development and stress responses [1 , 2] . A notable example is CLE ( CLAVATA3/EMBRYO SURROUNDING REGION-related ) family peptides involved in various developmental contexts in flowering plants , such as stem cell maintenance in meristems , vascular development , seed formation and growth control in response to environmental cues [3 , 4] . The 12–13 amino acid CLE peptides are proteolytically processed from precursor proteins encoded by CLE genes [5–7] . CLE peptide hormones undergo post-translational modification such as proline hydroxylation and arabinosylation during maturation [8] . Mature CLE peptides bind to specific membrane receptors that transmit signals to direct cell behavior , thereby manifesting cell-to-cell communication [9 , 10] . For example , the leucine-rich repeat receptor kinase CLV1 ( CLAVATA1 ) is a receptor for the CLV3 peptide in Arabidopsis , participating in the stem cell homeostasis in the shoot apical meristem [11–13] . A phylogenetically related receptor , TDR/PXY ( TDIF RECEPTOR/PHLOEM INTERCALATED WITH XYLEM ) , mediates TDIF ( tracheary element differentiation inhibitory factor ) peptide signaling essential for stem cell maintenance in the vasculature [14] . CLV3 and TDIF peptides possess characteristic residues for exclusive interaction with their specific receptors [15] and represent two major subclasses of CLE peptide family . Comparative genomics studies have revealed that the repertoire of developmental regulatory genes is conserved among land plants even though body plans vary among different groups [16–18] . All land plants undergo alternation of generations where the both haploid ( gametophyte ) and diploid ( sporophyte ) phases develop multicellular bodies and one of the two phases is dominant depending on the plant lineage . Phylogenetically , the monophyletic , diploid-dominant vascular plants either nest within a bryophyte grade , or are sister to a clade of bryophytes that possess haploid-dominant life cycles . In both types of body plans , meristems function as the source of growth by continually providing new undifferentiated cells . This is achieved by functional zonation of meristems: one or few pluripotent stem cells act as a source of rapidly proliferating cells that often undergo specific division orientations giving rise to differentiating cells [19] . CLE peptides function in shoot , root and vascular meristems of Arabidopsis and other vascular plants , by controlling cell division and differentiation . TDIF , encoded by CLE41/44 in Arabidopsis , is involved in three aspects of vascular cell behavior: inhibition of cell differentiation , enhancement of proliferation and control of cell division orientation [5 , 20 , 21] . TDR/PXY encodes a leucine-rich repeat receptor kinase ( LRR-RK ) of which extracellular LRR domain forms a superhelical structure that binds TDIF at its inner surface [8 , 22–24] . We have previously reported that the TDIF activity in vascular development is conserved in most vascular plants [25] . However , the biological function of TDIF/CLE , as with any other peptide hormones , is poorly understood in bryophytes . An intriguing question is whether the TDIF/CLE peptides in bryophytes control the meristem activity in the gametophytic body . We here investigated the role for TDIF/CLE peptide in the liverwort Marchantia polymorpha , a model bryophyte species [26 , 27] . Like many other developmental regulatory genes , the CLE gene family is conserved among land plants [28] . The M . polymorpha genome encodes two CLE genes , MpCLE1 and MpCLE2 , belonging to two distinct subclasses ( H-type including TDIF and R-type including CLV3 , respectively ) of the CLE family based on the initial amino acid in the mature peptide hormone motif ( Fig 1A , S1 Fig ) [25 , 29] . In addition , two distinct receptors for the CLE peptides , MpTDR and MpCLV1 are encoded in the M . polymorpha genome ( Fig 1B ) [29] . In contrast , the moss Physcomitrella patens has only R-type CLE genes and CLV1-type receptors . Thus , M . polymorpha provides a model system for studying TDIF/H-type CLE signaling in bryophyte development .
To test if MpCLE1 is functionally equivalent to AtCLE41 , a TDIF-encoding gene of Arabidopsis , we generated gain-of-function alleles in Arabidopsis . The effects of constitutive TDIF expression in Arabidopsis have been previously observed [5 , 15 , 20 , 30] . In the first leaves of 14-day-old seedlings , xylem vessels are formed along the leaf vein in wild type ( S2 Fig ) but it can be fragmented when AtCLE41 is driven by the constitutive 35S promoter ( S2 Fig; 35S:AtCLE41 ) . In addition , hypocotyl stele thickening can be enhanced in 35S:AtCLE41 plants compared to wild type ( S2 Fig ) . Along with these vascular phenotypes , overall plant growth can be significantly reduced in 35S:AtCLE41 plants ( S2 Fig ) . In contrast , 35S:MpCLE1 plants did not exhibit any of these phenotypes ( S2 Fig ) despite the expression level of MpCLE1 in this line being comparable to that of AtCLE41 in the 35S:AtCLE41 line ( S2 Fig ) . We further examined CLE bioactivities by peptide treatment assays [5 , 31] . Treatment with 5 μM TDIF caused xylem fragmentation and stele thickening resembling AtCLE41 overexpression phenotypes ( S2 Fig ) . In contrast , MpCLE1 peptide treatment did not induce these effects ( S2 Fig ) . Similar results were obtained in assays with 20 μM peptide ( S2 Fig ) . Collectively , these data indicate that MpCLE1 cannot functionally replace AtCLE41/TDIF in Arabidopsis . To elucidate the discrepancy between the phylogenetic and functional relationships of MpCLE1 and AtCLE41 , we performed amino-acid swapping between the two peptides , which differ at four residues ( Fig 1C; MpCLE1 and AtCLE41 ) . We synthesized four MpCLE1 peptide variants in which one of the four residues is changed to that of the AtCLE41 peptide ( Fig 1C; MpCLE1-2E/-3V/-5S/-11S , respectively ) . In addition , hydroxyprolines were incorporated in an MpCLE1 variant ( MpCLE1-Hyp ) to mimic the natural structure of TDIF ( Fig 1C; O indicates hydroxyproline ) . Among these MpCLE1-variants , MpCLE1-3V and MpCLE1-5S enhanced stele thickening ( Fig 1D ) and suppressed xylem differentiation in the leaf vein ( Fig 1E ) . These data indicate that MpCLE1 is indeed a TDIF-type CLE gene although a minor amino-acid substitution is required to convert MpCLE1 into a functional peptide in Arabidopsis . Since the MpCLE1-type residues ( Fig 1C; N3 and A5 ) differ from all other known CLEs in land plants , acquisition of these residues in MpCLE1 may have occurred in the liverwort lineage . We next analyzed the biological function of MpCLE1 in M . polymorpha . M . polymorpha is a thalloid liverwort and the body ( thallus ) grows at the apical notches which are indeterminate apical meristems and bifurcate periodically . For the clonal propagation , disc-shaped small progenies called gemmae are formed in gemmae cups that develop at the dorsal side of the thallus near the apical notch ( Fig 2A ) . When mature gemmae were cultivated for 14 days on solid medium supplemented with MpCLE1 peptide , the overall growth of plants was slightly reduced , and thallus lobes were twisted ( S3 Fig ) . Similar and marginally stronger phenotypes were observed with MpCLE1-Hyp peptide ( S3 Fig ) . Unexpectedly , the effects of TDIF were even stronger than those of MpCLE1 peptides ( S3 Fig ) . Transgenic lines overexpressing either MpCLE1 or AtCLE41 formed small and convoluted thalli ( Fig 2A and 2B , S3 Fig ) . Quantification of the area of thalli showed significant reduction of growth in the MpCLE1 overexpression lines ( Fig 2C ) . In addition , MpCLE1 overexpression lines produced fewer gemmae cups and did not form gametangiophores even 2 months after far-red light induction ( S3 Fig ) . To verify the minimal functional domain of MpCLE1 , we produced overexpression lines with C-terminal deletions ( MpCLE11-420 and MpCLE11-417 ) . The former contains the 12 amino-acid CLE peptide motif while the latter lacks the C-terminal asparagine residue of the peptide motif , an essential residue for CLE peptide activities [5 , 23 , 24] . As expected , MpCLE11-420 overexpression but not MpCLE11-417 overexpression exhibited growth defects ( S3 Fig ) , supporting the notion that the 12 amino-acid CLE peptide is the functional domain of MpCLE1 . To elucidate the cytological function of MpCLE1 , we analyzed apical meristem anatomy in 14-day-old plants grown from gemmae . In wild-type meristems , a single apical cell produces derivatives in four planes—dorsal , ventral and two lateral—and each of these primary derivatives undergoes a stereotypical pattern of divisions producing a 'merophyte' [27 , 32] . Cell divisions within merophytes produce a pattern of cells in the mature thallus , that when viewed in longitudinal section , appear as rows of cell files emanating from the apical meristem . In longitudinal sections of wild-type plants ( Fig 2D ) , proliferative region , which can be characterized by vertical cell division planes in internal tissues , is approximately 200 μm from the tip of the thallus ( bracket in Fig 2E ) . In contrast , the proliferative region in MpCLE1 overexpression plants was reduced to less than 100 μm from the tip ( bracket in Fig 2F ) , with cell differentiation/expansion occurring at a position closer to the apex , resulting in a distortion of thallus growth . We also examined meristem anatomy in transverse sections ( Fig 3A–3C ) . Consecutive transverse sections of wild-type and MpCLE1 overexpression plants were compared at same positions relative to a basal position “0 μm” at which the two lobes flanking the meristem merged ( Fig 3C ) . In wild type , small cytoplasmically dense cells are persistently observed in all sections examined ( 0–180 μm ) while MpCLE1 overexpression plants initiate cell expansion at the 40 μm position , and larger cells are observed compared to wild type throughout the sections resulting a thickened dorsi-ventral axis ( Fig 3A and 3B ) . Collectively , these data demonstrate MpCLE1 overexpression reduces the size of the proliferative region in the apical notch . A loss-of-function MpCLE1 allele ( Mpcle1-GT85 ) was generated via gene targeting ( S4 Fig ) . Fourteen-day-old Mpcle1-GT85 plants grown from gemmae formed convoluted thalli , with the thallus periphery curled upward and the apical notches growing downward ( epinastically ) into the medium . Unlike MpCLE1 overexpression plants , overall growth was not reduced ( S4 Fig ) . To gain a better understanding of the effects of the loss of MpCLE1 , we examined the anatomy of Mpcle1-GT85 plants . In longitudinal sections of Mpcle1-GT85 plants , the proliferative region was expanded ( n = 2; 266 and 232 μm in length ) compared to 200 μm in wild type ( bracket in Fig 4A ) , although the proliferative region was not as clear as in the wild type due to the less organized cell division plane orientation . In contrast , the timing of cell differentiation in epidermal tissue on the dorsal side is not significantly altered in the mutant since the air chamber development started at similar positions from the apical cells ( Fig 4A ) . Complementation analysis was performed by introducing a 6 . 2 kb genetic fragment spanning MpCLE1 into the knock-out line , designated as Mpcle1-GT85 gMpCLE1 . Compared to the knock-out mutant , the complementation line had a proliferative region similar in size to that of wild type in both internal ( n = 2; 198 and 196 μm ) and epidermal tissues , but the orientation of cell division planes in internal tissues is still less organized than in wild type ( Fig 4B ) . Curled thalli were also observed in the complementation line ( S4 Fig ) . Thus , additional genomic regulatory components might be required to fully complement the phenotype . In consecutive transverse sections , dorsi-ventral thickening of thalli was observed in Mpcle1-GT85 plants compared to the complementation line ( Fig 4C and 4D ) . In the consecutive sections , cell expansion continued until the 320 μm position in Mpcle1-GT85 while it ceased at 200 μm in wild type and the complementation line ( S5 Fig ) . These data suggest that MpCLE1 acts to suppress proliferative activity at the apical notch . Gametangiophores , specialized reproductive structures , are extensions of the vegetative thalloid body and the proliferative activity of apical meristem . Mpcle1-GT85 plants developed larger antherodiophores ( male gametangiophores ) than wild type ( Fig 5A and 5B ) . Consistent with the increased proliferation in vegetative thalli , stalks are thicker in Mpcle1-GT85 plants compared to wild type ( 1 . 12 ± 0 . 14 mm v . s . 0 . 65 ± 0 . 14 mm , mean ± S . D . , n = 10 ) and are composed of more cells in cross sections ( Fig 5C and 5D ) . The diameter of the antheridial receptacle was also increased in both longitudinal and transverse axes in Mpcle1-GT85 ( Fig 5E and 5F , S6 Fig ) . The mutant receptacles were thicker and contained larger antheridia , which can reach 1 . 12 mm in length at the maxima ( mean ± S . D . = 0 . 80 ± 0 . 14 mm , n = 28 ) in contrast to wild-type antheridia of 0 . 70 mm in length at the maxima ( mean ± S . D . = 0 . 55 ± 0 . 08 mm , n = 17 ) in our observation ( Fig 5G–5J ) , which is consistent to Higo et al . [33] . The complementation line developed normal antheridiophores and antheridia of 0 . 66 mm in length at the maxima ( mean ± S . D . = 0 . 56 ± 0 . 08 mm , n = 19 ) ( S4 Fig ) . Gametangiophore overgrowth was also consistently observed in archegoniophores of female Mpcle1-GT85 plants obtained by cross with Tak-2 , without affecting the branching pattern of fingered-lobes ( Fig 6H and 6I ) . These observations support the notion that MpCLE1 negatively controls the proliferative activity in apical meristems . To examine the ligand-receptor relationship between TDIF and TDR in M . polymorpha , we analyzed the physiological function of the TDR ortholog , MpTDR , by generating a knock-out line via homologous recombination ( S4 Fig; Mptdr-GT400 ) . Similar to Mpcle1-GT85 , 21-day-old Mptdr-GT400 thalli were curled upward at the periphery ( Fig 6A ) . In this mutant background , introduction of an MpCLE1 overexpression transgene did not alter thallus morphology ( Fig 6B ) . The area of thalli was not significantly changed between Mptdr-GT400 and MpCLE1 overexpression in Mptdr-GT400 background ( Fig 6C ) , indicating that MpCLE1 activity is dependent on MpTDR . Consistently , development of Mptdr-GT400 thalli was insensitive to 10 μM TDIF , the most effective MpCLE1-type peptide in our assays ( Fig 6D–6G ) . The archegoniophores of Mptdr-GT400 plants were larger than wild type , resembling the female Mpcle1-GT85 phenotype ( Fig 6H–6J ) . Introduction of MpCLE1 overexpression did not significantly alter the archegoniophore morphology of Mptdr-GT400 ( Fig 6K ) . Collectively , these data indicate that MpCLE1 acts through MpTDR to restrict the proliferative activity in the meristems of M . polymorpha . To analyze the expression patterns of MpCLE1 and MpTDR , we constructed GUS-reporter lines using 5 kb of genomic sequence upstream of the MpCLE1 and MpTDR coding sequences , respectively . GUS signal for both MpCLE1 and MpTDR promoters was first detected in the apical notches in 5-day-old gemmalings ( Fig 7A and 7B ) . In 10-day-old gemmalings , proMpCLE1:GUS signal was also detected along the midrib in addition to the signals at the apical notches ( S7 Fig ) . Both proMpCLE1:GUS and proMpTDR:GUS signals were observed in the developing antheridiophores suggesting that MpCLE1 signaling is functional during the development of antheridiophores ( S7 Fig ) . In longitudinal sections , proMpCLE1:GUS signal was detected in a small area around the apical cell in 5-day-old gemmalings ( Fig 7C ) . In contrast , proMpTDR:GUS signal was detected in the dorsal part in the apical meristem ( Fig 7D ) , suggesting that MpCLE1 signaling may operate within the apical meristem . Since the expression of proMpTDR:GUS was observed in cells close to the apical cell , the expression domains of MpCLE1 and MpTDR could partially overlap . The influence of MpCLE1 signaling on the expression domains was examined by peptide treatment . In 5-day-old gemmalings grown in the presence of 10 μM MpCLE1-Hyp or TDIF peptide , proMpCLE1:GUS signals were not conspicuously affected , while proMpTDR:GUS signals appeared more intense than in the control , however , no change in expression levels was detected in fluorometric quantification assays ( S7 Fig ) . In whole-mount in situ hybridization ( WISH ) assays , MpCLE1 expression was detected at the apical notches of 7-day-old gemmalings ( S8 Fig ) . MpEF1α-as signal was detected throughout the thallus and strongly in apical notches while MpEF1α-s was not , consistent with Althoff et al . [34] ( S8 Fig ) . Collectively , these data suggest that the MpCLE1 peptide signal may move from a ventral region towards more dorsal regions , with a maximum potential response , i . e . MpTDR expression , near the apical cell .
In this work , we show that TDIF-type CLE peptide signaling restricts proliferative activity in the meristem and the overall size of reproductive organs in the liverwort M . polymorpha although development of apical cell , merophyte , or cell division markers such as labile cyclin-GUS is important to unambiguously quantify the change of proliferative activity and its location in the meristem [35] . TDIF has been known as a positive regulator of meristem activity in vascular plants , but conversely MpCLE1/TDIF acts as a negative regulator in M . polymorpha . This function resembles CLV3 peptides in flowering plants . In Arabidopsis clv3 mutants , the shoot and floral meristems are enlarged , resulting in club-shaped siliques and additional floral organs [36] . Tomato mutants and cultivars deficient in CLV3 signaling have excess floral organs and fasciated fruits [37] . In monocots , loss of CLV signaling enhances the meristem activity , which results in the increase of floral organ number in rice and kernel row number in maize [38–43] . Since the reproductive structures are determinate in these plants , the effects of enhanced proliferative activities are more conspicuous than effects on vegetative growth . Correlation of proliferative activity in the meristem and reproductive structure size was also found in Mpcle1 mutant that exhibits gametangiophore enlargement along with the enhanced proliferative activity . In contrast to the phenotypes in flowering plants , the gametangiophores of the M . polymorpha mutants did not form an excess number of reproductive structures , such as additional lobes in the receptacles or fasciated stalks . The lobes and stalks are both sexual extensions of thallus and the numbers of these structures depend on the branching of the meristems during gametangiophore development . Thus , the control of proliferative activity by MpCLE1 signaling is uncoupled from meristem branching . In addition to the functional similarity , the expression pattern of MpCLE1 relative to its receptor gene , MpTDR , also resembles to that of CLV3 relative to CLV1 in Arabidopsis , in which the peptide ligand is expressed at/around stem cells while the receptor is expressed in a neighboring region ( Fig 7E ) [44 , 45] . In Arabidopsis , receptor expression in the partially overlapping domain with the ligand is thought to establish different cell fates among neighboring target cells [11 , 12 , 46] . In M . polymorpha , proliferative cells in the dorsal region of apical meristem could also interpret MpCLE1 signals from apical cells as a positional cue . However , the morphological phenotypes of knock-out and overexpression mutants for MpCLE1 were found more significantly in the internal tissue . Further identification of downstream signaling components that suppress proliferative activity would clarify the role of MpCLE1-mediated communication in meristem homeostasis . The slow but continued thallus growth in MpCLE1 overexpression lines indicates that ectopic MpCLE1 does not terminate stem cell activity , and indeed apical cells are maintained in overexpression lines . This contrasts with Arabidopsis CLV3 overexpression , which terminates shoot growth by abolishing stem cells in the meristem [12 , 13] and with Arabidopsis TDIF mutants that lose vascular stem cells at a certain frequency [47] . Loss of MpCLE1 also resulted in abnormal cell division planes in the meristem , similar to both loss and ectopic expression of TDIF in Arabidopsis [21 , 22] . Therefore , the cellular function of MpCLE1 signaling appears to be a combination of CLV3 and TDIF signaling observed in Arabidopsis . Likewise , it was recently demonstrated that CLV3 orthologs in P . patens also regulate both cell proliferation and cell division orientation planes in gametophore shoots and that this function is conserved in Arabidopsis [48] . Coupled with our results , these observations hint that perhaps the ancestral functions of both the TDIF and CLV3 pathways may have encompassed regulation of both cell proliferation and cell division orientation planes . Since TDIF and TDR are genetically conserved in vascular plants , lack of TDIF and TDR in the moss P . patens suggests that TDIF signaling has been lost within the moss lineage . Thus , TDIF signaling is not an indispensable regulatory module for P . patens development despite its significant contribution to M . polymorpha development , perhaps due to the major differences in body plan between mosses and liverworts [49] . In contrast , M . polymorpha retains both TDIF ( MpCLE1 ) and CLV3 ( MpCLE2 ) orthologs suggesting each possesses unique functions in liverworts . Our work reveals a general association of CLE peptide signaling with both gametophytic and sporophytic meristems in land plants although it’s still unclear if CLE is involved in the control of sporophytic meristems in bryophytes . Meristems in both generations evolved in the ancestral land plant and are not present in algae which do not possess CLE peptide signaling . One major difference in the body plans of these organisms is that the land plant body is built from meristems with 3 or more cutting faces , whereas algal bodies are largely constructed from modifications of filamentous growth so that three-dimensional co-ordination of cell behavior is not required . Thus , mechanisms to focus and control meristem growth evolved concomitantly with meristems , with one mechanism being via CLE peptide-based cell-to-cell communication .
Phylogenetic analysis was performed as described previously [25] . Sequences for phylogenetic analysis were summarized in S1 Table . The sequences were first aligned with Clustal X and Bayesian phylogenetic analyses were performed on the alignments using MrBayes 3 . 2 . 1 [50] . Columbia-0 ( Col-0 ) line of Arabidopsis thaliana and Takaragaike-1 ( Tak-1 ) and BC3-38 lines of M . polymorpha were used as wild type in phenotypic analyses and as the genetic background for transgenic lines . Growth conditions for Arabidopsis and observation methods of vasculature were as described previously [14] . M . polymorpha plants were grown at 22°C on half-strength Gamborg B5 medium ( pH 5 . 5 , 1 . 4% agar ) under continuous light . For the induction of gametangiophores , far-red light was supplemented . For the observation of antheridia , antheridial receptacles were dissected with forceps under a stereoscopic microscope ( Stemi 2000-CS , Zeiss , Jena , Germany ) . To measure the area of thalli , gemmae were grown on the half-strength Gamborg B5 medium for 14 days and images of plants were analyzed using ImageJ . Peptides were synthesized by Fmoc chemistry with a peptide synthesizer ( CS136XT , CSBio , CA , USA ) . Analytically pure peptides were obtained by reverse-phase HPLC . Primers , plasmids and transgenic plants are summarized in S2 Table and S3 Table [51–54] . Transformation of M . polymorpha was performed using spores ( cross between Tak-1 and Tak-2 ) or regenerating thalli according to Ishizaki et al . [51] and Kubota et al . [52] . Homologous recombination-mediated gene targeting was performed according to Ishizaki et al . [53] . Transformation of Arabidopsis was performed with the floral dip method [55] . Total RNA was extracted from 11-day-old Arabidopsis seedlings with RNeasy plant mini kit ( Qiagen , Hilden , Germany ) . Three independent RNA samples were used for cDNA synthesis with SuperScript III first-strand synthesis system ( Thermo Fisher Scientific , MA , USA ) . Two technical replicates are made for each RNA sample and the average was used as a single data point . Primers used for qPCR are described in S2 Table . The qPCR assay was performed on a LightCycler 96 system ( Roche ) using KAPA SYBR FAST qPCR kit ( KAPA BIOSYSTEMS , MA , USA ) . Amounts of cDNA input to the qPCR reactions were normalized by the AtTUA4 expression levels . For the comparison of AtCLE41 and MpCLE1 gene expression levels , the expression levels were normalized by performing qPCR using 10 pg of cloning plasmid ( pENTR-AtCLE41 or pENTR-MpCLE1 ) as template . Mean values of 3 samples ± S . D . were indicated . GUS staining was performed according to Ishizaki et al . [51] . Briefly , M . polymorpha gemmalings grown on agar medium were directly submerged in X-Gluc assay solution containing 50mM sodium phosphate buffer ( pH 7 . 2 ) , 1mM potassium-ferrocyanide , 1mM potassium-ferricyanide , 10mM EDTA , 0 . 01% Triton X-100 and 1mM 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid . After vacuum infiltration , samples were incubated for 3–12 hours at 37°C in dark . GUS-stained samples were cleared with ethanol and mounted with clearing solution ( chloral hydrate-glycerol-water , 8:1:2 ) before imaging with light microscope ( Axio Imager . A2 , Zeiss ) . For histological analyses , GUS-stained samples were rinsed with water before fixation . At least 2 independent transgenic lines were examined for each experiment and representative images are shown . Fluorometric quantification of GUS activity was performed according to Ishizaki et al . [56] with minor modifications . Three biological replicates were sampled . For each replicate , 5 gemmalings ( 5–10 mg in total ) grown for 5 days on the half-strength Gamborg B5 medium with or without 10 μM peptide were collected in a microtube , frozen with liquid nitrogen and homogenized with micropestle in 100 uL of GUS extraction buffer , containing 50 mM sodium phosphate ( pH 7 . 2 ) , 10 mM 2-mercaptothanol , 1 mM EDTA and 0 . 01% Triton X-100 . Debris was removed by centrifugation at 13 , 000 rpm for 5 min at 4°C . Protein concentration was measured using 5 uL of the protein solution using the TaKaRa Bradford Protein Assay Kit ( Takara , Kyoto , Japan ) with the low-concentration protocol according to the manufacturer’s instructions . For the GUS enzyme reaction , 40 μl of each protein solution ( ca 30 μg protein ) , 50 μl of the GUS extraction buffer and 10 μl of 10 mM 4-methylumbelliferyl β-D-glucuronide ( MUG ) was mixed in a microtube and incubated at 37°C for 40 min . The reaction was stopped by adding 900 μl of 200 mM sodium carbonate . Fluorescence ( 460 nm emission/360 nm excitation ) of liberated 4-methylumbelliferone ( MU ) was measured on a microplate reader ( Synergy LX , BioTek , VT , USA ) and normalized by the protein concentration . For preparation of plastic sections , plant samples were trimmed with a razor blade and fixed in FAA solution ( 50% ethanol: 10% formalin: 5% acetic acid in water ) . Fixed samples were embedded into Technovit 7100 resin ( Heraeus Kulzer , Wehrheim , Germany ) and 4 μm sections were prepared with a rotary microtome ( RM2235 , Leica , Heidelberg , Germany ) . Sections were stained with 0 . 02% toluidine blue or with 0 . 002% Safranin-O solution for GUS-stained samples and then mounted with Entellan New ( Merck Millipore , MA , USA ) . WISH was performed by modifying a protocol for Arabidopsis seedlings [57] . For the preparation of RNA probes , MpCLE1 and MpEF1α genes were cloned into pCRII vector ( Thermo Fisher ) by PCR from M . polymorpha cDNA using primers indicated in S2 Table . Digoxygenin-labeled ribo-probes were synthesized using SP6/T7 RNA polymerases ( DIG RNA Labeling Kit , Roche , Basel , Switzerland ) after digestion with Xho I/BamH I restriction enzymes , respectively . For the preparation of plant samples , 10-day-old gemmalings were fixed in a 1:1 mixture of heptane and fixative ( 4% paraformaldehyde ( PFA ) , 15% DMSO and 0 . 1% Tween- 20 in water ) for 45 min on a rotary shaker at room temperature ( RT ) . Following fixation , tissues were placed in 100% methanol twice for 5 min and 100% ethanol three times for 5 min to remove chlorophyll and incubated for 30 min in a 1:1 mixture of ethanol and Histo-Clear . After treatment of 100% ethanol , tissues were rehydrated in 75% ethanol ( v/v in water ) , 50% ethanol ( v/v in phosphate buffered saline; PBS ) and 25% ethanol ( v/v in PBS ) for 10 min each . Tissues were refixed in the fixative for 20 min and rinsed twice for 10 min in PBST ( 0 . 1% v/v Tween-20 in PBS ) at RT . For permeabilization of cell wall , tissues were treated with 0 . 1% cellulase ( final concentration 100 μg/ml ) and 0 . 2% macerozyme ( final concentration 200 μg/ml ) for 30 min at RT and with proteinase K ( final concentration 125 μg/ml ) for 30 min at RT . After stopping the permeabilization with glycine ( final concentration 2mg/ml ) , tissues were refixed in the fixative and rinsed in PBST for 10 min at RT . For labeling , tissues were hybridized with the DIG-probes ( 150 μg/ml ) in the hybridization solution ( 50% formamide in 5×saline-sodium citrate buffer ( SSC ) containing 0 . 1% Tween-20 , 0 . 1 mg/ml heparin and 0 . 1mg/ml herring sperm DNA ) for 16 hours at 55°C . The probe mixture was denatured prior to use for 10 min at 80°C . Hybridized tissues were rinsed in 4×SSC three times for 15 min at 55°C , 0 . 1×SSC three times for 20 min at 55°C and maleic acid buffer ( MAB ) for 5 min at RT . After washing excessive probe , tissues were incubated in 0 . 1% boheringer blocking reagent in maleic acid buffer ( BBR-MAB ) for 30 min at RT , the 1:2 , 000 diluted anti-digoxigenin-AP Fab fragments ( Roche ) in BBR-MAB for 2 hours at RT and 0 . 05% ( v/v ) Tween-20 in MAB three times for 15 min at RT . Bound ribo-probe was detected by overnight staining with nitroblue tetrazolium ( Roche ) and bromo-chloro-indolyl phosphate ( Roche ) for overnight at 4°C . Stained tissues were photographed under a stereoscopic microscope ( MZ16F , Leica ) .
|
Land plants undergo an alternation of generations where both haploid and diploid phases develop multicellular bodies . Their growth relies on the activity of meristems at the growing tips of their bodies . Here we show a CLE peptide hormone acts as an intercellular signal controlling proliferative activity in the apical meristem of Marchantia polymorpha . Our finding reveals a general association of CLE peptide signaling with meristem homeostasis , a feature that evolved in the ancestral land plant , in both haploid and diploid phases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"brassica",
"bryology",
"plant",
"physiology",
"notch",
"signaling",
"plant",
"science",
"model",
"organisms",
"experimental",
"organism",
"systems",
"plants",
"nonvascular",
"plants",
"arabidopsis",
"thaliana",
"research",
"and",
"analysis",
"methods",
"animal",
"studies",
"proteins",
"biochemistry",
"signal",
"transduction",
"eukaryota",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"post-translational",
"modification",
"vascular",
"plants",
"biology",
"and",
"life",
"sciences",
"meristems",
"cell",
"signaling",
"signal",
"peptides",
"organisms"
] |
2019
|
Control of proliferation in the haploid meristem by CLE peptide signaling in Marchantia polymorpha
|
Rift Valley fever virus ( RVFV ) causes outbreaks of severe disease in livestock and humans throughout Africa and the Arabian Peninsula . In people , RVFV generally causes a self-limiting febrile illness but in a subset of individuals , it progresses to more serious disease . One manifestation is a delayed-onset encephalitis that can be fatal or leave the afflicted with long-term neurologic sequelae . In order to design targeted interventions , the basic pathogenesis of RVFV encephalitis must be better understood . To characterize the host immune responses and viral kinetics associated with fatal and nonfatal infections , mice were infected with an attenuated RVFV lacking NSs ( ΔNSs ) that causes lethal disease only when administered intranasally ( IN ) . Following IN infection , C57BL/6 mice developed severe neurologic disease and succumbed 7–9 days post-infection . In contrast , inoculation of ΔNSs virus subcutaneously in the footpad ( FP ) resulted in a subclinical infection characterized by a robust immune response with rapid antibody production and strong T cell responses . IN-inoculated mice had delayed antibody responses and failed to clear virus from the periphery . Severe neurological signs and obtundation characterized end stage-disease in IN-inoculated mice , and within the CNS , the development of peak virus RNA loads coincided with strong proinflammatory responses and infiltration of activated T cells . Interestingly , depletion of T cells did not significantly alter survival , suggesting that neurologic disease is not a by-product of an aberrant immune response . Comparison of fatal ( IN-inoculated ) and nonfatal ( FP-inoculated ) ΔNSs RVFV infections in the mouse model highlighted the role of the host immune response in controlling viral replication and therefore determining clinical outcome . There was no evidence to suggest that neurologic disease is immune-mediated in RVFV infection . These results provide important insights for the future design of vaccines and therapeutic options .
Rift Valley fever virus ( RVFV ) is a zoonotic arbovirus that has serious public health , veterinary and economic impacts throughout Africa and the Arabian Peninsula . RVFV outbreaks are often triggered by periods of heavy rain and the subsequent emergence of RVFV-infected mosquitoes , and are characterized by the subsequent widespread abortion storms in livestock . People may become infected directly by the bite of infected mosquitoes , or by exposure to infected animal tissues . In humans , RVFV infection is generally described as a self-limiting febrile illness marked by severe arthralgia , myalgia , photophobia and headache . Although most individuals recover without long-term sequelae , some human infections progress to a serious hepatitis , hemorrhagic syndrome or delayed onset encephalitis . In these most severely afflicted patients , case fatality can be greater than 20% [1] , [2] , and survivors of the encephalitic syndrome frequently suffer from long-term neurological complications . Fatal RVFV encephalitic disease was first documented in the 1974–76 South African outbreak , and has since been a consistent finding in subsequent epidemics [1] , [3] , [4] . RVFV is a member of the family Bunyaviridae , genus Phlebovirus , and has a tripartite negative-sense RNA genome composed of a large ( L ) , a medium ( M ) and a small ( S ) segment . The L segment encodes the RNA-dependent RNA polymerase . The structural glycoproteins Gn and Gc , as well as a nonstructural protein NSm , are encoded on the M segment . NSm is thought to function by inhibiting virus-induced apoptosis [5] , [6] as well as by mediating replication in the mosquito vector [7] . The ambisense S segment encodes the nucleoprotein ( N ) that is required for RNA synthesis , and the nonstructural NSs protein . NSs is the major virulence factor and functions broadly by inhibiting the host immune response [8] via generalized downregulation of host transcription [9] , post-transcriptional degradation of protein kinase R ( PKR ) [10] , [11] and TFIIH p62 [12] , and repression of the interferon-β ( IFNβ ) promoter [13] . The critical role of NSs in RVFV pathogenesis has been clearly illustrated in a number of studies , including the RVFV mouse model . Following infection with wild-type RVFV , C57BL/6J mice succumb to a fatal hepatitis within 2–3 days of infection , regardless of the dose or route of administration . In contrast , subcutaneous inoculation of a highly attenuated virus lacking NSs ( ΔNSs ) results in a widely disseminated infection that is cleared with no indication of clinical disease . We recently demonstrated that viral clearance and protection from clinical disease is dependent on a functional CD4+ T cell response . In that study , 30% of mice depleted of CD4+ T cells succumbed to neurologic disease following ΔNSs infection , in association with reduced antibody titers and evidence of altered CD4+ T cell regulatory function [14] . The results indicated that severe neurologic disease developed following a reduced systemic immune response , but also raised the question of whether neurologic disease resulted from direct viral damage to the CNS or to pathology resulting from disruption of the blood brain barrier and an unregulated immune response . To address this question , a mouse model of RVFV infection with route-dependent disease was utilized as a platform to explore the differences in the host immune response between animals that developed fatal RVFV encephalitis and those that had subclinical infections . Here we report that intranasal ( IN ) infection of the highly attenuated ΔNSs virus resulted in consistently fatal neurologic disease 7–9 days after infection , in contrast to the uniformly subclinical infection that follows subcutaneous ΔNSs infection . Intranasal virus inoculation has been commonly utilized in mouse models of human viral encephalitic diseases , including Herpes simplex virus [15] , influenza A virus [16] , Sindbis virus [17] , Venezuelan equine encephalitis virus [18]–[20] , western equine encephalitis virus [21] and Hendra virus [22] . This method of inoculation can be a proxy for aerosol exposure , which has been shown to be an important route of RVFV exposure in laboratory settings [23] , and is thought to play a role in the infection of individuals handling infected animal tissues . Aerosol exposure of wild-type RVFV has been shown to cause uniformly fatal disease in mice [24] , [25] , similar to subcutaneous infection with the same virus . Use of the ΔNSs virus model in this study permitted characterization and direct comparison of the host immune responses and viral kinetics in mice that developed fatal encephalitis ( IN-inoculation ) with those that had a subclinical infection ( FP-inoculation ) . Fatal infections were associated with weak systemic immune responses and a subsequent failure to clear virus from the periphery , and onset of clinical disease correlated with high viral RNA loads and infiltration of activated T cells in the CNS . Interestingly , depletion of T cells did not improve survival , suggesting that in contrast to other encephalitic viruses [26] , [27] , that RVFV neurologic disease is not immune-mediated . These results provide important insights into the pathogenesis of severe RVFV infection that could inform the development of therapies targeted towards treating or preventing RVFV mediated encephalitis .
Animal procedures in this study complied with institutional guidelines , the US Department of Agriculture Animal Welfare Act , and the National Institutes of Health Guidelines for the humane use of laboratory animals . All procedures were approved by the Centers for Disease Control and Prevention ( CDC ) Institutional Animal Care and Use Committee ( IACUC ) ( Protocols 2023 and 2409 ) . All work with infectious RVFV was completed in a biosafety level ( BSL ) -3E laboratory . Female 8–10-week-old C57BL/6J mice were obtained from Jackson Laboratories and were housed within BSL-3E laboratories in microisolator pans in HEPA filtration racks , following standard barrier techniques . In all animal experiments , mice were evaluated for clinical signs of disease at least once daily , and were euthanized according to a pre-determined clinical illness scoring algorithm or if found moribund . Stocks of recombinant RVFV ( strain ZH501 ) lacking NSs ( ΔNSs ) and a GFP-expressing RVFV lacking both NSs and NSm ( ΔNSm/ΔNSs:GFP ) were produced using reverse genetics and propagated as previously described [28] , [29] . Titers of all viral stocks were determined as tissue culture infective dose 50 ( TCID50 ) on VeroE6 cells and visualized by indirect fluorescent-antibody assay ( IFA ) using anti-RVFV hyperimmune mouse ascitic fluid ( HMAF ) primary antibody . Virus sequence identity was verified by full-length genome sequencing with 6-fold redundancy prior to virus use . For virus inoculated intranasally ( IN ) or subcutaneously in the left rear footpad ( FP ) , mice were briefly anesthetized with isofluorane . For the initial LD50 study , mice were inoculated with 10-fold dilutions of virus , ranging from 1×105 TCID50 to 10 TCID50 per nare . In subsequent experiments , mice infected with ΔNSs IN received 1×104 or 1×105 TCID50 in 10 µL sterile Dulbecco's modified Eagle's medium ( DMEM ) per nare . Mice infected in the left rear FP received 2 × 105 TCID50 in 20 µL of sterile DMEM subcutaneously . Sham-inoculated mice were given sterile DMEM . Significant differences between survival curves were analyzed using a log-rank ( Mantel-Cox ) test ( GraphPad Prism; GraphPad Software , Inc . ) . C57BL/6J mice were depleted of CD4+ and CD8+ T cells using monoclonal anti-CD4 ( GK1 . 5 ) and anti-CD8 ( YTS169 ) antibodies , or mock-depleted using an isotype control antibody ( LTF2 ) ( all obtained from Bio X Cell ) . Antibodies were diluted in sterile phosphate buffered saline ( PBS ) ; 300 µg of GK1 . 5 and 300 µg of YTS169 were administered to each mouse intraperitoneally on days -3 and -1 prior to virus infection . Depletion efficiency was determined by flow cytometry to be greater than 99% . One hour prior to euthanasia , mice were injected intraperitoneally ( IP ) with 800 µL of 1% Evan's blue dye . Mice were deeply anesthetized with isofluorane and perfused with 10 mL of PBS . Brains were removed and examined for evidence of dye uptake . Blue coloration of spleens , livers and kidneys were controls for effective distribution of dye . Serum was collected and used for RVFV anti-N IgG ELISA as described previously [30] , [31] . Briefly , purified RVFV N protein or negative control LASV G1 protein was used at a concentration of 200 ng/well . Plates were blocked in blocking buffer ( 5% skim milk , 5% fetal bovine serum ( FBS ) and 0 . 1% Tween-20 in 1 × PBS ) at 37°C for 1 h . Plates were incubated with serially diluted sera in blocking buffer for 1 h at 37°C . Plates were washed 3 times in 1 × PBS with 0 . 1% Tween-20 ( PBST ) and incubated with goat anti-mouse horseradish peroxidase ( HRP 1∶10 , 000 ) ( Jackson ImmunoResearch ) in blocking buffer for 1 h . Plates were washed 3 times in PBST and ABTS substrate was added according to manufacturer's instructions ( KPL Inc . ) . Reactions were stopped with the addition of 1% SDS and read at 405 nM . Absolute values obtained from negative controls were subtracted prior to analysis . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( n = 3/group/time point; *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . Stock ΔNSm/ΔNSs:GFP virus was diluted to 100 TCID50 in 50 µL DMEM without FBS . Sera were heat-inactivated at 56°C for 30 min . In a 96-well plate , 1∶5 , 1∶10 , 1∶20 , 1∶40 , 1∶80 , 1∶160 , and 1∶320 serum dilutions were made in 50 µL DMEM . An equal volume of diluted RVFV was added to diluted sera and samples were incubated for 1 h at 37°C . A suspension of approximately 3×104 VeroE6 cells was added to each well , and the plates were incubated for 72 h before visualization of GFP positive cells . VNT100 was defined as the highest dilution that permitted 100% neutralization of virus input . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( n = 3/group/time point; *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . At time of euthanasia , spleens were harvested and immediately placed in sterile Roswell Park Memorial Institute ( RPMI ) media . Single cell suspensions were generated by passing spleens through a 70 µM mesh filter using the blunt end of a 5 cc syringe plunger . Cell suspensions were washed in RPMI and red blood cells ( RBCs ) were lysed using RBC Lysis Buffer according to manufacturer's instructions ( Sigma-Aldrich ) . Cells were washed two additional times in PBS and suspended in RPMI for quantitation of total cells using a hemocytometer . Whole brains were removed and immediately placed in sterile RPMI . Brains were macerated with a #10 blade and single cell suspensions were produced by passing tissue through a 70 µM filter using the blunt end of a 5 cc syringe . Cells were suspended in RPMI and layered on a 70/30 Percoll gradient in a 15 mL conical tube . Gradients were centrifuged at 800 × g for 20 minutes . The interface was taken and washed in 1x Hank's balanced salt solution ( HBSS ) . RBCs were lysed according to manufacturer's instructions ( Sigma-Aldrich ) . Cells were suspended in either flow buffer or RPMI for subsequent assays . Approximately 1×105 splenocytes or brain PBMC were stained in flow buffer ( PBS with 3% FBS ) using 1∶200 dilutions of the following antibodies: fluorescein isothiocyanate ( FITC ) anti-CD4 ( 553651 ) , FITC anti-CD8 ( 553031 ) , phycoerythrin ( PE ) anti-CD3 ( 555275 ) , PE anti-CD45 ( 553081 ) , PE anti-CD11b ( 561689 ) , allophycocyanin ( APC ) anti-CD69 ( 560689 ) , or FITC anti-CD19 ( 561740 ) , all from BD Biosciences . Staining was performed at 4°C for 20 min and then cells were washed twice in PBS before being counted . From each well , 50 , 000 events were collected on an Accuri Flow Cytometer ( BD Biosciences ) . CFlow Sampler software ( Accuri ) was used to gate on the lymphocyte population . CD3 by CD4 or CD8 plots were then generated , a second gate was placed to delineate the CD3/4 or CD3/8 T cells , and the expression of CD69 on these cells was determined . Plot quadrants were set based upon the patterns observed in unstained cells . Specificity of staining was confirmed by the concomitant use of appropriate isotype control antibodies and fluorescence minus-one controls ( all from BD Biosciences ) . Compensation was performed using single antibody stains . Three to five animals from each group of mice were tested independently on each day . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . MultiscreenHTS filter plates ( EMD Millipore ) were activated with 35% ethanol , and then washed in sterile PBS . Activated plates were coated with 1 µg/well of anti-mouse IgG or with 100 µL of inactivated RVFV virus ( 107 TCID50/mL ) , and with 200 ng/well of affinity purified glutatione S-transferase ( GST ) -tagged RVFV N protein and allowed to bind overnight at 4°C . Plates were then washed in sterile PBS and 1×105 splenocytes were placed per well with 3 fold dilutions . Cells were incubated for 24 hours at 37°C . Plates were washed with PBS and incubated with anti-mouse IgG HRP ( 1∶5000 in PBS ) for 1 hour at room temperature and washed 5 times with PBS . Spots were developed using DAB substrate ( Sigma-Aldrich ) and plates were read on an ELISPOT reader using KS ELISPOT software ( Zeiss ) . RVFV-specific antibody production is presented as relative to total antibody production . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( n = 5/group/time point; *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . MultiscreenHTS filter plates ( EMD Millipore ) were activated with 35% ethanol , and then washed in sterile PBS . Activated plates were coated with anti-TNFα ( BD Biosciences ) , anti-IL-4 , anti-IL-2 , or anti-IFNγ ( Mabtech Inc . ) capture antibodies and incubated overnight at 4°C . Plates were washed and 1×105 splenocytes/well were plated in 2-fold dilutions and incubated for 24 h at 37°C . Plates were washed with PBS and incubated with biotinylated detection antibodies for 2 h at room temperature . Plates were washed and incubated with streptavidin HRP for 1 h at room temperature . After a final series of washes , spots were detected using DAB substrate ( Sigma ) and read on an ELISPOT reader using KS ELISPOT software ( Ziess ) . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( n = 3/group/time point; *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . Mice were deeply anesthetized with isofluorane , terminally bled and perfused with 10 mL PBS . Specimens of olfactory bulb , cerebral cortex , cerebellum , brainstem , liver , spleen , popliteal lymph node , sciatic nerve and whole blood were collected . RNA was extracted using MagMax Total RNA Isolation kit ( Ambion ) . With the exceptions of the sciatic nerve and popliteal lymph node , in which cases the entire structures were used , approximately 100 mg of tissue were placed directly in lysis buffer and homogenized using a high-throughput tissue grinder ( GenoGrinder2000 ) . Approximately 50 µL of whole blood was added directly to lysis buffer with isopropanol . Homogenates were extracted using the MagMax Express-96 Magnetic Particle Processor ( Ambion ) according to manufacturer's directions including a DNase treatment step . Liver , spleen , sciatic nerve , popliteal lymph node , olfactory bulb , cerebrum , cerebellum , brainstem and whole blood RNA were tested by RVFV qRT-PCR as described previously , using 1 µL of total RNA [29] , [32] , [33] . To normalize between samples , 1 µL of total RNA was used for a ribosomal RNA qRT-PCR assay ( ABI Biosciences ) . Reactions were run on an ABI 7500 quantitative PCR machine ( ABI Biosciences ) . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( n = 5/group/time point; *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . RNA extracted from the cerebrum of IN or FP-inoculated mice was used to determine inflammatory gene expression . Primer probe sets for IL-6 , TNFα , CCL2 and CCL5 ( Applied Biosciences ) were used with Superscript III Platinum qRT-PCR reagents ( Invitrogen ) . Samples were run in 50 µL reactions: 1 . 5 µL of primer probe mix was combined with 25 µL Master Mix , 1 µL Taq enzyme mix , 21 . 5 µL sterile water and 1 µl of total RNA . To normalize between samples , 1 µL of total RNA was used for a GAPDH qRT-PCR assay ( ABI Biosciences ) . Reactions were run on an ABI 7500 quantitative PCR machine ( ABI Biosciences ) . Relative fold changes between mock-infected and IN- or FP-infected animals were determined using the comparative CT method as described by Schmittgen and Livak [34] . Data were analyzed using 2-way ANOVA with Bonferroni post-tests ( n = 5/group/time point; *p<0 . 05; **p<0 . 01; ***p<0 . 001 ) ( GraphPad Prism; GraphPad Software , Inc . ) . Whole brains from mice inoculated IN with ΔNSs virus were euthanized after development of severe neurologic signs 7 and 8 dpi . Brains were fixed by immersion in 10% neutral buffered formalin for 7 days . Tissues were processed , paraffin-embedded , and sectioned following routine methods , and stained with hematoxylin and eosin ( H&E ) for histological examination . Immunohistochemistry ( IHC ) assays were performed using a polymer-based indirect immunoalkaline phosphatase detection system with colorimetric detection of antibody/polymer complex with Fast Red Chromogen ( Thermo Fisher Scientific ) as previously described [35] . Pre-treatment of tissue sections was applied as needed using Proteinase K ( PK ) ( Roche , Mannheim , Germany ) or Antigen Retrieval ( AR ) ( Biocare Medical ) prior to IHC staining . Tissues were evaluated using a polyclonal ( PK , 1∶1000 ) and a monoclonal ( PK , 1∶500 ) Rift Valley Fever virus antibody provided by the CDC ( SPB ) . In addition , tissues were evaluated using CD68 ( PK , 1∶100 , DAKO Cytomation ) , CD3 ( AR , 1∶100 , Dako Cytomation ) , CD8 ( AR , Ready-to-Use , Leica Microsystems , UK ) and CD20 ( PK , 1∶200 , Dako Cytomation ) cellular markers . Appropriate positive and negative control serum and tissues were tested in parallel for each assay .
As described previously [14] , subcutaneous FP inoculation of 2 . 0×105 TCID50 ΔNSs virus resulted in uniform survival with no indication of clinical signs , whereas mice inoculated with ΔNSs virus IN developed dose-dependent clinical disease ( p<0 . 0001; Figure 1A ) . Mice infected with the highest virus doses , 2 . 0×105 and 2 . 0×104 TCID50 , displayed signs of severe neurologic disease and consistently succumbed to infection 7–9 dpi ( Figure 1A ) . At the time of death , IN-inoculated mice had significantly higher virus RNA loads in the brain relative to the liver ( p<0 . 0001; Figure 1B ) . Of the mice given 2 . 0×103 TCID50 , 80% succumbed to infection with a prolonged time to death . Lower doses did not result in clinical disease . The striking differences between intranasal and subcutaneous inoculation provided a unique platform to characterize and compare the immune responses elicited during fatal and nonfatal infections , respectively . To determine the kinetics of virus spread within the nervous system , 5 FP- and 5 IN-inoculated mice were euthanized , perfused with PBS and samples of olfactory bulb , cortex , cerebellum , brainstem and sciatic nerves were taken for RVFV-specific qRT-PCR on 1 , 2 , 3 , 4 , 5 , 6 , 7 and 8 dpi . The olfactory bulbs of all IN-inoculated mice tested positive for virus RNA 2 dpi and viral RNA loads steadily increased over time ( Figure 2A ) . Subsequently , virus RNA was found in the cerebrum ( Figure 2B ) , cerebellum ( Figure 2C ) and brainstem ( Figure 2D ) . In contrast , small numbers of FP-inoculated mice had low levels of virus RNA present throughout the brain at late points in infection but these virus RNA loads were significantly lower than IN-inoculated mice . To better illustrate the temporal pattern of virus spread in the CNS of IN-inoculated mice , the virus RNA data from the olfactory bulb , cerebrum and cerebellum are presented together ( Figure 2E ) . All FP-inoculated mice had significant virus RNA levels present in the sciatic nerve 2 dpi , whereas all IN-inoculated mice had RVFV-positive sciatic nerves 5 dpi ( Figure 2F ) . These data are consistent with both the route of infection ( FP-inoculated mice ) and with the dissemination of virus throughout the nervous system ( IN-inoculated mice ) . In order to determine the effect of the route of infection on peripheral virus kinetics in vivo , we measured viral RNA loads by qRT-PCR in the blood , liver , spleen and popliteal lymph nodes . Viral RNA levels were significantly higher in the blood , liver and spleen of FP- versus IN-inoculated mice 1 dpi ( Figures 3A , 3B and 3C , respectively ) . However , by 3 dpi , blood , liver and spleen viral RNA loads were indistinguishable between the FP- and IN-groups . FP-inoculated mice cleared virus from the liver by 7 dpi but IN-inoculated mice were unable to control virus replication in the liver . FP-inoculated mice had significantly higher virus RNA loads in the ( draining ) popliteal lymph node than IN-inoculated mice 1–3 dpi ( Figure 3D ) , consistent with the route of infection . To explore differences in humoral immunity following IN or FP inoculation , we compared RVFV anti-N IgG and neutralizing antibody levels and the number of RVFV-specific B cells between groups on 2 , 4 , 6 and 8 dpi . Production of IgG and neutralizing antibodies were delayed by 48 hours in the IN-inoculated mice relative to the FP-inoculated mice . FP-inoculated mice had significantly higher IgG titers 8 dpi , the time at which antibody was first apparent in the IN group ( Figure 4A ) . Only FP-inoculated mice had a detectable neutralizing antibody response 6 dpi , and higher levels 8 dpi than IN-inoculated mice ( Figure 4B ) . FP-inoculated mice also had significantly higher numbers of RVFV-specific IgG-secreting B cells than IN-inoculated mice 4 and 8 dpi ( Figure 4C ) . As a broad measure of T cell activation , we next evaluated the relative proportion of CD69+ T cells in the spleen following IN or FP inoculation . The spleens of FP-inoculated mice had a significantly higher percentage of CD3+/CD4+ T cells 8 dpi than IN-inoculated mice ( Figure 5A ) . The percentage of CD3+/CD8+ T cells was also significantly higher in FP-inoculated mice 4 and 8 dpi ( Figure 5B ) . However , CD4+ T cell activation , as measured by expression of CD69 , was higher in IN-inoculated mice than FP-inoculated mice throughout the course of infection and significantly greater 8 dpi ( Figure 5C ) . A similar trend was apparent in CD69-positive CD8+ T cells; a higher percentage of CD8+ T cells of IN-inoculated mice expressed CD69 on all 3 days examined ( Figure 5D ) . To evaluate the quality of the cellular immune response , we measured IFNγ , TNFα and IL-2 expression in splenocytes using ELISPOT assays . Significantly more T cells from FP-inoculated mice expressed IFNγ and TNFα 6 and 8 dpi ( Figures 6A and 6B , respectively ) . FP-inoculated mice also had more T cells producing IL-2 6 and 8 dpi , but IN-inoculated mice had higher numbers of IL-2 expressing T cells 2 dpi ( Figure 6C ) . Leukocyte migration into the CNS was compared between groups using flow cytometry . In an initial experiment , IN-inoculated mice had a significantly higher proportion of CD45+ cells than FP-inoculated mice 7 and 8 dpi ( Figure 7A ) , and the majority of these appeared to be T cells given positive staining for CD3 ( Figure 7B ) . In a subsequent experiment , leukocytes were harvested from brains of IN- and FP-inoculated mice late in infection , 6 and 8 dpi . In IN-inoculated mice , there were significantly more CD4+ ( Figure 7C ) and CD8+ ( Figure 7D ) T cells , than in FP-inoculated mice . In brains from the IN group , a significantly higher percentage of CD4+ T cells were expressing CD69 ( Figure 7E ) . Similarly , on 6 and 8 dpi , a striking number of CD8+ T cells were also CD69+ in the brains of IN-inoculated mice ( Figure 7F ) . Given the early neuroinvasion and increased numbers of leukocytes seen in the brains of IN-inoculated mice , we sought to determine if RVFV infection resulted in an increase in BBB permeability . 5 FP- and 5 IN-inoculated mice were injected with Evans blue dye one hour prior to euthanasia on 1 , 2 , 3 , 4 , 5 , 6 , 7 and 8 dpi . Mice were then anesthetized and completely perfused with PBS , and whole brains were removed . No blue dye coloration was seen in the brains from FP- or IN-inoculated mice at any time point ( data not shown ) indicating that there was no breakdown of the blood brain barrier during infection . Blue coloration of liver , spleen and kidneys was apparent in all mice , indicating adequate dye perfusion . To gain a more detailed understanding of the inflammatory response in the CNS , we evaluated inflammatory gene expression in the brains of IN- and FP-inoculated mice . We found that both groups demonstrated a mild increase in expression of CCL2 , IL-6 and TNFα 6 dpi . However mice inoculated IN , but not FP , displayed significant upregulation of several key inflammatory cytokines 8 dpi , including CCL2 , CCL5 , IL-6 and TNFα , ( Figure 8 A-D ) that coincided with the development of severe clinical disease . In order to determine the distribution of leukocytes infiltrating the CNS , we evaluated the brains of moribund IN-inoculated mice using histology and IHC . RVFV specific IHC suggested viral entry most likely occurred through olfactory bulbs ( Figures 9A and 9B ) , and multifocal staining throughout the brain was consistent with subsequent vascular and transneural spread of virus . Meningoencephalitis was present and associated with the presence of viral antigen ( Figures 9C and 9D ) . Inflammatory foci were associated with morphologic indications of apoptosis , viral antigen , a CD68+ microglial response and the presence of CD3+ cells ( Figure 9 E-H ) To explore the possibility that the onset of neurologic disease is dependent on immune-mediated pathology , we depleted mice of CD4+ and CD8+ T cells prior to ΔNSs IN infection . However , no differences were observed in terms of clinical signs , mortality , or time to death between CD4/CD8-depleted mice and mock-depleted mice ( Figure 10 ) .
Encephalitis is a serious and occasionally fatal complication of human RVFV infection , however , little is known about the factors contributing to this disease outcome . There are currently no approved vaccines for human use and no treatment exists beyond supportive care . Development of efficacious preventative and therapeutic measures requires a more detailed understanding of disease pathogenesis , including the kinetics of virus replication and spread , the modes of neuroinvasion , and the protective and potentially pathologic roles of the host immune response . These studies are ideally approached by comparing data from a fatal mouse model of RVFV encephalitis with data from a nonfatal RVFV infection in mice . Here , we demonstrate that ΔNSs virus intranasal inoculation consistently caused fatal neurologic disease 7–9 dpi , a finding that stands in stark contrast to the absence of clinical disease following subcutaneous administration of the virus . Similarly , route-dependent clinical disease was recently described in a mouse model of Hendra virus; mice inoculated parenterally did not develop clinical signs but fatal encephalitis developed following intranasal exposure [22] . Our results also correlate well with those from recent studies that demonstrated enhanced neurovirulence of wild-type RVFV when administered as an aerosol in nonhuman primates [36] and rodents [24] , [37] . Comparison of immunocompetent mice inoculated FP or IN permitted characterization of the viral kinetics and immune responses associated with nonfatal and fatal outcomes , respectively . Interestingly , viral RNA loads were similar in peripheral tissues by 2–3 dpi , presumably due to hematogenous spread from the initial sites of virus replication . However , the route of inoculation had a dramatic effect on the quality of the systemic host response; FP-inoculated mice mounted a more rapid and effective immune response than IN-inoculated mice . A robust humoral response , as measured by significantly earlier and higher anti-RVFV IgG and neutralizing antibodies , was associated temporally with early viral clearance from peripheral tissues . Later in infection , FP-inoculated mice initiated a strong T cell response , with significantly higher production of IFNγ , TNFα and IL-2 than IN-inoculated mice . Studies of other encephalitis-associated viruses have shown that IFNγ and TNFα can play a critical role in controlling viral replication in the CNS . For example , in a Japanese encephalitis virus ( JEV ) model , IFNγ knock-out mice had higher virus RNA titers in brain than wild-type mice , but peripheral tissues have similar viral RNA loads [38] , a pattern similar to that seen in our RVFV study , where production of IFNγ was associated with better control of virus . Similarly , following peripheral infection with herpes simplex virus , production of TNF protected mice from development of severe encephalitis [39] . These results suggested that FP inoculation , but not lethal IN inoculation , stimulated a highly effective adaptive immune response that supported viral clearance and prevented progression to clinical disease . Paradoxically in IN-infected mice , very early in infection there was evidence of enhanced activation of T cells as compared with FP-inoculated mice . These mice produced significant levels of IL-2 , a cytokine required for the proliferation and differentiation of T cells ( as reviewed in [40] ) and enhanced natural killer cell activity [41] , that correlated temporally with a marked increase in the proportion of T cells expressing the activation marker CD69 . However , in contrast to FP-inoculated mice , the T cells of IN-inoculated mice did not produce IFNγ or TNFα , cytokines associated with the initiation of an appropriate adaptive response and subsequent viral control . The presence of activated T cells that lack effector function has been described previously in models of other neurotropic viruses , including lymphocytic choriomeningitis virus ( LCMV ) [42] and rabies virus [43] . CNS infection with a highly neurotropic strain of rabies virus resulted in increased expression of CD69 on T cells , but reduced production of inflammatory cytokines IL-2 , TNFα and IFNγ . The authors concluded that the apparent T cell ‘unresponsiveness’ resulted from virally-induced immunosuppression specifically targeting the CD4+ helper T cell response [43] . The effect of route of infection on CD4+ helper T cell function during RVFV infection remains to be evaluated in further detail . The profound effect of the route of inoculation on host immune response and clinical outcome may be explained , at least in part , by differences in the nature of the immune cells encountered following the initial inoculation . After FP infection , virus immediately encounters resident dendritic ( Langerhans ) cells and macrophages of the skin . These cells are professional antigen-presenting cells ( APCs ) , and upon recognition of viral antigen , migrate to the draining lymph node to stimulate an early immune response . In contrast , intranasal inoculation resulted in rapid infection in the CNS . There are no detectable dendritic cells in the CNS and microglia are considered poor APCs and limited in the ability to initiate an adaptive response ( reviewed in [44] ) . The delay between initial IN inoculation and systemic infection , coupled with the immunologically privileged nature of the CNS , may help explain the lagging immune response and subsequently enhanced viral replication in IN-inoculated mice . Viral RNA loads were significantly higher in the brains of IN- relative to FP-inoculated mice , despite similar virus RNA loads in peripheral tissues , suggesting that the timing and mechanism of neuroinvasion were route-dependent . When administered IN , the virus rapidly infected the olfactory bulb and spread caudally to the cerebrum and cerebellum . Multifocal distribution of viral antigen was apparent , suggesting hematogenous and transneural spread from the olfactory bulb . Indeed , active RVFV infection of neuroepithelium and the olfactory bulb have been recently reported [24] , strongly supporting neuroinvasion through the olfactory tract in intranasally exposed mice . On the other hand , FP-inoculated mice had virus RNA-positive sciatic nerves 2 dpi , with virus detected several days later in the CNS , and particularly the brainstem , suggesting ΔNSs virus might utilize retrograde transport to access the CNS . Infection of the peripheral nerves of IN-inoculated mice coincided with increasing virus RNA loads in the brainstem , suggesting that axonal virus transport might also occur in the form of anterograde transport . In light of the apparent hematogenous spread of virus and the presence of cellular infiltrate in the CNS , we evaluated the possibility that virally induced alterations in blood brain barrier permeability led to the onset of clinical disease . In models of JEV [45] and Venezuelan equine encephalitis virus ( VEEV ) [46] development of viral encephalitis has been linked to inflammatory-mediated alterations in BBB permeability either as a primary mechanism of neuroinvasion or as a mediator of immunopathology . Late stage RVFV has been associated with an increase in proinflammatory cytokines and chemokines [47] , Furthermore , in this study , upregulation of IL-6 , TNFα , CCL2 and CCL5 gene expression in the brains of IN-inoculated mice was associated with high virus RNA loads and infiltration of large numbers of aberrantly activated CD4+ and CD8+ T cells . However , to our surprise we found no indication of increased BBB permeability in the mice , even at the height of the CNS inflammatory response , indicating that BBB disruption is not required for RVFV pathogenesis . Although there were no overt alterations in BBB permeability , it was possible that T cell infiltration and associated upregulation of inflammatory cytokines alone might have enhanced neurovirulence , as seen in other models of viral encephalitis [26] , [27] , [48] . Induction of these cytokines and chemokines amplify the antiviral response and recruit leukocytes , but if left unchecked , can have destructive effects . In the brains of moribund IN-inoculated mice , a diffuse meningoencephalitis was present and characterized by apoptosis in the presence of viral antigen , T cells and CD68+ microglia . To determine if the T cells played a pathologic role in the development of neurologic disease , we depleted mice of both CD4+ and CD8+ T cells prior to ΔNSs virus intranasal infection . Strikingly , we observed no improvement in survival of T cell-depleted mice relative to mock-depleted mice . These results indicate that the development of RVFV encephalitic disease is not T cell-mediated , and instead suggests that direct viral damage is responsible for the onset of severe disease . In summary , we developed a mouse model of fatal RVFV encephalitic disease utilizing intranasal inoculation with an attenuated RVFV . Fatal human cases of RVFV encephalitis are relatively uncommon , and there are few descriptions of the CNS pathology . However , the available case reports from South Africa and Egypt describe focal areas of necrosis with perivascular cuffing [49] and degeneration of the cerebral neurons with infiltration of microglial cells [50] . Although these findings are fairly nonspecific , they are similar to the lesions seen in the intranasally infected mice , suggesting this is a reasonable model of RVFV encephalitis . Comparison of the course of disease following IN or FP inoculation of ΔNSs virus highlighted the potential roles of the host immune response in determining clinical outcome . IN inoculation resulted in rapid and fulminant virus replication in the CNS via infection of the olfactory bulb , whereas FP-inoculated mice had low virus RNA loads in the CNS and survived without developing disease . Despite different outcomes , peripheral virus kinetics were the same following both routes of inoculation , until viral clearance , associated with a robust antibody response , occurred exclusively in FP-inoculated mice . IN-inoculated mice displayed an inappropriate T cell response; although apparently activated , T cells failed to release critical cytokines for control of virus replication . The ineffective adaptive immune responses of IN-inoculated mice were associated with the development of peak virus RNA loads , strong proinflammatory responses and infiltration of activated T cells into the CNS , followed by severe clinical disease and death . Interestingly , in contrast to other encephalitic viruses , T cells were not required for development of neurologic disease , suggesting that RVFV encephalitis is not immune-mediated . Although we cannot exclude a role for inflammatory cytokines and other immune cells in the development of clinical disease , our studies strongly point to direct viral destruction of the brain as the major cause . These findings have implications for the development of therapeutics targeting RVFV neurologic disease , which may be particularly useful given the relatively delayed onset of RVFV encephalitis in infected humans .
|
Rift Valley fever virus ( RVFV ) is a mosquito-borne virus that causes severe disease in people and livestock throughout Africa and the Arabian Peninsula . Human disease is usually self-limiting , but a small proportion of individuals develop fatal encephalitis . The role of the host immune response in determining disease outcome is largely unknown . In order to compare the quality and character of immune responses in nonfatal and fatal cases , we used an attenuated RVFV to inoculate mice by two routes . Subcutaneous inoculation resulted in a subclinical systemic infection that was rapidly cleared due to a robust adaptive response . In contrast , intranasal inoculation stimulated weaker immune responses that failed to control virus replication and culminated in uniformly fatal encephalitis . With many encephalitic viruses , the onset of disease is mediated by changes in blood brain barrier permeability and often , subsequent injury to the CNS by an uncontrolled immune response . However , our results suggest that development of RVFV disease does not depend on either mechanism , but rather results from direct virus-mediated damage in the CNS . Future therapeutic drug design should take into account all possible routes of virus exposure as well as the role of therapies that boost the adaptive response to better combat disease .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"humoral",
"immunity",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"immunology",
"microbiology",
"animal",
"models",
"model",
"organisms",
"research",
"and",
"analysis",
"methods",
"animal",
"models",
"of",
"infection",
"inflammation",
"viral",
"immune",
"evasion",
"mouse",
"models",
"pathogenesis",
"immune",
"response",
"immune",
"system",
"immunity",
"host-pathogen",
"interactions",
"virology",
"biology",
"and",
"life",
"sciences",
"acquired",
"immune",
"system"
] |
2014
|
Rift Valley Fever Virus Encephalitis Is Associated with an Ineffective Systemic Immune Response and Activated T Cell Infiltration into the CNS in an Immunocompetent Mouse Model
|
The peregrine falcon Falco peregrinus is renowned for attacking its prey from high altitude in a fast controlled dive called a stoop . Many other raptors employ a similar mode of attack , but the functional benefits of stooping remain obscure . Here we investigate whether , when , and why stooping promotes catch success , using a three-dimensional , agent-based modeling approach to simulate attacks of falcons on aerial prey . We simulate avian flapping and gliding flight using an analytical quasi-steady model of the aerodynamic forces and moments , parametrized by empirical measurements of flight morphology . The model-birds’ flight control inputs are commanded by their guidance system , comprising a phenomenological model of its vision , guidance , and control . To intercept its prey , model-falcons use the same guidance law as missiles ( pure proportional navigation ) ; this assumption is corroborated by empirical data on peregrine falcons hunting lures . We parametrically vary the falcon’s starting position relative to its prey , together with the feedback gain of its guidance loop , under differing assumptions regarding its errors and delay in vision and control , and for three different patterns of prey motion . We find that , when the prey maneuvers erratically , high-altitude stoops increase catch success compared to low-altitude attacks , but only if the falcon’s guidance law is appropriately tuned , and only given a high degree of precision in vision and control . Remarkably , the optimal tuning of the guidance law in our simulations coincides closely with what has been observed empirically in peregrines . High-altitude stoops are shown to be beneficial because their high airspeed enables production of higher aerodynamic forces for maneuvering , and facilitates higher roll agility as the wings are tucked , each of which is essential to catching maneuvering prey at realistic response delays .
The stoop is a remarkable attack strategy used by peregrine falcons Falco peregrinus , and a range of other raptors [1–4] . It involves a steep , controlled dive in which the attacker strikes its prey at high-speed with a massive blow in mid-air [1] . The high momentum of the attacker places it at obvious risk of harm , especially when diving into flocks of birds [1] or when pulling out only meters from the ground [3] . Arguably , for the stoop to evolve as an habitual attack strategy , these risks must be outweighed by certain survival advantages , and stooping has therefore been proposed either to save energy [5] , or to enhance catch success [6] . These hypothetical advantages remain unproven , however , because it is challenging to compare the success rates of different attack strategies empirically . Success rates are confounded by a variety of factors , including the experience [1 , 5] and reproductive status [7] of the attacker , the season of the attack [8] , and the species of prey [6] . Even the seriousness of the attacker’s behavior may be an important source of variation: falcons seem to not always focus on achieving a high success rate , and appear sometimes to be practising or playing with their prey [9] . Moreover , the outcome of the stoop is often difficult to observe due to its high speed [6] . There presumably exists a trade-off between different factors influencing catch success in a stoop . On the one hand , it has been proposed that the high speed of the attack provides an element of surprise , leaving little time for the prey to evade [5 , 10] . On the other hand , it is possible that the high speed of the attack decreases the precision of interception [2] , and makes it harder for the attacker to follow the prey if it turns sharply [11] . Such trade-offs are difficult to investigate empirically , and we therefore turn to modeling and simulation . Because physical and physiological constraints influence catch success , we use an embodied cognition approach [12] . We investigate the success of different attack strategies by incorporating in a physics-based simulation model the aerodynamics , flight mechanics , guidance , and control . Such detailed simulations have already proven useful in work on missile guidance: the increasing demand for better performing missiles forces the inclusion of the detailed dynamics of the missile and its target when comparing the effectiveness of different guidance systems [13] . The nonlinear nature of these dynamics restricts the use of analytic methods , such as linear-quadratic optimal control , and the effectiveness of different mechanisms must therefore be examined through parametric variation of the system between repeated simulations of interception . Here , we study the general intercept problem under the particular dynamics of flapping and gliding bird flight . Interception in this biological context differs from that of missiles in that gravity plays a pivotal role in determining the best attack strategy: in missiles , the speed and acceleration are so high that the effects of gravity are marginal , but in birds , the acceleration due to gravity dominates the dynamics . The flight performance of the model-birds in our simulations depends on their flight morphology , and differs considerably between predator and prey . To maneuver , model-birds flap , glide , and vary their wing span . We use the model to study attacks by peregrine falcons on a habitual prey species , the common starling Sturnus vulgaris . We simulate three different patterns of flight by the prey—straight flight , smooth turning , and non-smooth turning ( see Fig 1 ) . This approach to varying the target motion is standard in simulations of missile guidance systems , and broadly summarizes the main options for the target to maneuver [13 , 14] . It also captures the range of different prey behaviors found in nature . For instance , if a bird is caught by surprise while commuting , then typically it will be flying in a straight line . Conversely , when turning , birds usually maneuver smoothly , but they will also fly erratically if a threat is detected , in a kind of non-smooth maneuvering flight known as jinking . Here we investigate whether , when , and why stooping increases catch success in each of these three situations ( see S1–S5 Videos for a visualization of attacks in each scenario ) . Because falcons often attack in wide , open spaces at high altitude , there are no objects or boundaries in our simulation space . We parametrically vary the falcon’s initial position relative to its prey to simulate a continuum of possible attack strategies ( e . g . stoops versus level chases ) . Predator and prey are each free to move with 6 degrees of freedom in translation and rotation , and are subject to gravitational and aerodynamic forces , which they manipulate by controlling their wings ( Fig 2 ) . The model-bird’s flight controller determines the changes in wing shape and motion that best meet the accelerations commanded by its guidance system , under a quasi-steady blade element model of the aerodynamics . In model-falcons , the guidance system commands turning toward the prey in closed-loop , whereas in model-starlings the guidance system is a forcing function that is set to ensure that the prey remain within approximately ±20 m of their starting altitude . Because we assume that birds maximize their flight speed during escape and pursuit , model-birds always generate the maximum possible forward acceleration given their instantaneous velocity and orientation , subject to the constraint that they must simultaneously meet , as closely as possible , the acceleration demanded normal to their flight direction by their guidance system . The falcon’s closed-loop guidance is essential in commanding the changes in velocity that are needed to intercept prey , whether maneuvering or not , and to deal with the effects of steering error . Our model-falcons use a guidance law called pure proportional navigation , which has been shown to fit the empirically measured attack trajectories of peregrine falcons closely [15] . Proportional navigation is also favored as a guidance law in missiles , because it provides a simple way of implementing the geometric rule known as parallel navigation or constant absolute target direction ( CATD ) , according to which the attacker holds the geographic direction of the line-of-sight to target constant through time [14 , 16–18] . This geometry guarantees interception if the attacker is closing range , because at every instant it is set on a collision course with its target ( i . e . would hit its target if both continued flying at constant velocity thereafter ) . Under pure proportional navigation , the attacker turns at an angular rate proportional to the angular rate of the line-of-sight to target . Although this guidance law can be written in three-dimensional vector form , it is more intuitively explained in the two-dimensional case , for which: γ ˙ = N λ ˙ ( 1 ) where γ denotes the bearing of the attacker’s velocity vector and λ denotes the bearing of the line-of-sight to target , both measured in an inertial reference frame; the dot notation denotes the time derivative , and N is called the navigation constant . The numerical value of N determines the rate of convergence to a parallel navigation ( CATD ) course: in missiles , low values of N result in slow convergence , whilst high values can cause overshoot , leading to control instability [14] . Partly for these reasons , intermediate values of N between 3 and 5 are typical in most missile applications . The overall objective of our simulations is to identify the attack strategy that maximizes the catch success of the falcon for a given prey motion , a range of assumptions regarding the delay in the falcon’s response , and the error in its vision and control ( see Materials and methods ) . Here , an attack strategy is defined as some particular combination of the predator’s navigation constant N , and its initial vertical and horizontal distance from the prey . The optimization was conducted via parametric variation of the attack strategy , in combination with Generalized Additive Modeling ( GAM ) , which we used to interpolate between the 106 randomly chosen attack strategies that we simulated in each optimization [19 , 20] . We hypothesise that stooping maximizes catch success , and that it does so as a direct consequence of the flight physics in our simulation model . We test this by asking whether a model-falcon’s catch success is maximized by attacking from a high altitude , which couples into a high flight speed in our simulations . Remarkably , we find that the optimality of stooping depends not only on the motion of the prey , but also on the tuning of the underlying guidance law . Specifically , we show that stooping is only expected to evolve in conjunction with the same low values of the navigation constant N that have been identified empirically in peregrine falcons [15] .
Flight performance is expected to be a key determinant of catch success in a chase . Clearly , any prey species that can fly faster than a falcon will be able to outrun its attacker in straight flight . In practice , peregrine falcons fly much faster than starlings , and our aerodynamic model predicts that they hold a considerable speed advantage in both level flight ( maximum speed: 29 versus 23 ms−1; Fig 3a ) and vertical dives ( terminal speed: 123 versus 52 ms−1; Fig 3b; see section K for a comparison between flight performance in the model and empirical measurements ) . Even so , escape is possible if the prey species can outmaneuver its attacker . For instance , if at a given flight speed the prey can produce a higher aerodynamic force relative to body weight than its attacker ( i . e . produce a higher load factor ) , then it may escape by turning more tightly than its attacker in a smooth maneuver called a turning gambit [21 , 22] . Our aerodynamic model shows that starlings can indeed sustain higher load factors than peregrine falcons flying at the same speed ( Fig 3c ) , and that although falcons can achieve even higher load factors by flying faster ( Fig 3c ) , the net effect is such that a starling will always be able to turn on a tighter radius than a faster-flying falcon ( Fig 3e ) . Similarly , if the prey can achieve a higher roll acceleration than the falcon , then it will be able to redirect its lift faster , and hence outmaneuver its attacker in a non-smooth jinking maneuver . Our aerodynamic model predicts that a starling can indeed produce a higher roll acceleration than a falcon flying at the same speed ( Fig 3d ) . So great is a model-starling’s advantage in this respect that a model-falcon can only be expected to match a model-starling’s maximum roll acceleration by diving at close to terminal velocity ( i . e . at close to its maximum speed ) . Hence , model-starlings may often escape model-falcons in our simulations , even though their maneuvers are not implemented as an evasive response to the falcon . In summary , a starling can always outmaneuver a falcon that is flying at a similar speed , but a falcon can always beat the load factor and roll acceleration of a starling by diving at a sufficiently higher speed . Whether this strategy enhances catch success will presumably depend on the flight pattern of the prey , and the complex ways in which the predator’s flight speed , response delay , and errors in vision and control interact to affect its guidance . We explore the outcome of these complex interactions in the simulations presented below . The catch success of model-falcons was always maximized by entering a steep dive , but the optimal starting altitude varied greatly between the three different flight patterns of the prey ( Table 1; see asterisked points in Fig 4 ) . Catch success in attacks on straight-flying prey ( Fig 4a ) was maximized by stooping from a low altitude ( < 200m ) , leading to a low flight speed at the point of intercept ( 35 − 45ms−1 ) . Optimal stoop altitude was somewhat higher ( c . 350m ) when prey maneuvered smoothly ( Fig 4b ) , leading to a moderate intercept speed ( 50 − 55ms−1 ) . Catch success with non-smoothly maneuvering prey ( Fig 4c ) was maximized by stooping from a very high altitude ( c . 1500m ) , leading to a very high intercept speed ( > 100ms−1 ) approaching the terminal velocity of the model-falcon ( see Fig 3 ) . Interestingly , catch success barely declined when the model-falcon attacked from a higher altitude than the optimum ( Fig 4 ) , but was greatly reduced if the model-falcon attacked from a lower starting position ( Fig 4 ) , so stooping from a high altitude is never a bad strategy provided that the guidance system is appropriately tuned ( see below ) . The attack strategy of a model-falcon encompasses both its initial position relative to the prey , and the setting of its navigation constant N . The global optima that we have so far discussed ( asterisked points in Fig 4 ) assume joint optimization of the predator’s initial attack position and its navigation constant N , and the optima for both parameters depend on the motion of the prey ( see also Table 1 ) . Selection on N is expected to be strongest when prey execute non-smooth maneuvers , for which high catch success is achieved over only a narrow range of N ( compare width of dark blue area denoting high catch success in Fig 5c with the equivalent areas in Fig 5a and 5b ) . Interestingly , for all three types of prey motion , the optimal setting of N tends to be lower the faster the stoop ( see dashed lines in Fig 5 plotting the optimal setting of N conditional upon the speed at intercept ) . Conversely , for a given setting of N , the optimal intercept speed becomes lower the higher the value of N ( see solid lines in Fig 5 plotting the optimal speed at intercept conditional upon the setting of N ) . Thus , for any given type of prey motion , high-speed , high-altitude stoops only maximize catch success over a small range of comparatively low values of N . At higher values of N , catch success is maximized by using a low-speed ( Fig 5 ) , low-altitude ( Fig 4 ) attack , but this is generally less successful than a high-speed , high-altitude attack at a lower value of N . In summary , it turns out to be essential for our model-falcons to set their navigation constant appropriately: if a sub-optimal value of N were used , then stooping might no longer be the best attack strategy , because of poor catch success . For instance , if a model-falcon were to use the optimal value of N for smoothly maneuvering prey ( N = 5 . 6 ) against prey executing non-smooth maneuvers , then a high-altitude stoop would be unlikely to result in prey capture ( third panel of Fig 4c ) . This does not necessarily mean that a falcon must actively adjust N to match the maneuverability of its prey: the best attack strategy of a model-falcon against the best defensive flight pattern of a model-starling ( i . e . non-smooth ) involves entering a high-speed , high-altitude stoop at N ≈ 3 . This minimax strategy not only yields maximal catch success against non-smoothly maneuvering prey , but also yields near-maximal catch success against prey that are flying straight or maneuvering smoothly ( second row of Fig 4 ) . Hence , subject to the assumptions of our model , we expect falcons to adopt a general strategy of stooping from high-altitude at N ≈ 3 , because this strategy is effective against all of the different patterns of prey flight that we have tested here . Some interesting flight trajectories emerge at N < 2 ( see S2 Fig ) . In this case , the model-falcon exerts most of its acceleration towards the end of its attack ( see also [14] ) , often diving below its prey before looping upward to intercept . This upward-curved trajectory is regularly observed in nature [4 , 23] , and has previously been suggested to be a strategy of a falcon to fly into the blind spot of its prey’s vision [9] . Our model provides a more parsimonious explanation for these flight paths , which can emerge naturally from the dynamics of the underlying feedback law . The most important factor that causes the reduction in catch success observed at high values of the navigation constant N is the response delay of the model-falcon . A robustness analysis ( Fig 6a ) shows that high values of N are no longer associated with a low catch success if the reactions of the falcon are effectively instantaneous ( compare catch success as a function of N at τ = 0 . 1ms delay with the equivalent line for the default τ = 50ms delay used in our baseline model; see Materials and methods ) . Conversely , if the falcon’s actual response delay is greater than the default assumed in our baseline model , then the optimal value of N is driven towards an even lower value ( Fig 6a ) . Visual error also affects the optimal value of N in our simulations: if the falcon is subject to greater visual error than the default value assumed in our model , then the navigation constant is again driven towards an even lower value of N ( Fig 6b ) . This reflects the fact that the propagation of this visual error into the commanded acceleration is directly proportional to N ( see Fig 2 ) . In contrast , the optimal value of N is robust to the error assumed in the control system itself ( Fig 6b ) . How do response delays , or errors in vision and control , impact the success of a given attack strategy ? As expected , catch success declines as each of these quantities increases ( Fig 2 ) . Remarkably , however , the optimal starting altitude becomes lower when the visual error or control error is increased ( Fig 6c and 6d ) . Thus , a high-speed , high-altitude stoop only maximizes catch success if the falcon is accurate in both vision and control . A high-speed stoop maximizes catch success for all of the response delays that we tested , noting that any much longer delay would have resulted in very low catch success ( Fig 6e ) . On the other hand , if the falcon’s response is effectively instantaneous ( τ = 0 . 1ms ) , then 100% catch success is attained even in a low-altitude dive from < 200m . This implies that the falcon’s flight performance is sufficient to catch a starling in a low-level stoop , but that delays in the model-falcon’s response hamper its ability to catch prey . The lower catch success that results from having a slower response can be ameliorated by diving from higher altitudes at lower N . When a falcon stoops from high altitude , its attack is characterized by both a very high flight speed , and a very steep descent angle—either of which could promote catch success . To investigate the effect of steepness of the descent , we altered the initial conditions of the simulations so as to model a horizontal attack at very high initial speed ( 112 ms−1 ) . This effectively simulates the final approach of a falcon that stoops from a very high altitude to gain speed before levelling off to intercept . Remarkably , the maximum catch success of these model-falcons is only 3% lower than for those intercepting their prey at the same speed in a steep dive ( 61 vs 64% ) . This implies that the steep descent angle is not directly responsible for the overwhelming success of a stoop , and hence that the key reason for starting from a very high altitude is to gain airspeed by converting potential energy to kinetic energy . The very high airspeed attained in a stoop enables model-falcons to exceed the model-starlings’ maximal load factor and roll acceleration ( Fig 3 ) . To test which of these two dimensions of flight performance causes an increase in catch success in a stoop , we artificially capped the maximal load factor or maximal roll acceleration of our model-falcons . We thus investigated the catch success of a bird flying at the same high speed achieved in a high-altitude stoop ( > 100ms−1 ) , but with the lower maximal acceleration associated with sustained level flight ( 30ms−1 ) . Limiting either component of the model-falcon’s flight performance resulted in a substantial drop in catch success ( 51% when limiting roll acceleration and 42% when limiting load factor ) . This suggests that the high speed that a falcon attains in a stoop is important partly because of the higher load factors and higher roll accelerations that can be achieved in high-speed flight . Interestingly though , model-falcons flying at high speed still performed considerably better than model-falcons in sustained level flight ( 31% vs 26% ) even though the maximal load factor and roll acceleration was made the same , and even when these fast falcons levelled off before interception . This implies that a high flight speed is beneficial in and of itself , independent of the higher acceleration performance that is usually also associated with fast flight . This result might seem surprising , because model-falcons fly faster than model-starlings even in sustained level flight ( Fig 3a ) , and the most obvious consequence of increasing the falcon’s flight speed further is to increase its turning radius , potentially causing it to overshoot when attacking sharply turning prey . However , the flight speed of a falcon varies continuously in our model , on account of its varying acceleration demand , and work on missile guidance and control has shown that the accelerations commanded in response to variations in speed are lower when the angle between the current line-of-sight and the expected point of intercept is smaller . The faster the falcon , the smaller this angle , which reduces the risk of control saturation , and thereby decreases the probability of missing the target .
Although our physics-based model is realistic enough for its intended purpose , there are obviously further constraints in nature that we have not modelled here , including the effects of unsteady aerodynamics , the dynamics of pitch and yaw instability , and the mechanics of catching or knocking the prey with the talons at intercept . There are also other complicating factors that we have not modelled , including the effects of explicit evasive maneuvers by the prey , or the impact of intra- and inter-specific variation in flight morphology and physiology , and hence variation in the flight performance of predator or prey . These factors can be studied through extensions of the model and through parametric variation of the model between simulations , and will be considered elsewhere . Nevertheless , our approach to studying the dynamics of aerial predation is unique among behavioral studies of complex systems in combining guidance and control laws inspired by missile theory [14] with a detailed simulation model of the biology and physics of animal flight . The underlying feedback laws are well-founded in the theory of optimal guidance [14] , and their validity as a phenomenological model of guidance and control in peregrine falcons has already been verified in nature [15] . Furthermore , the simulation approach that we have used proves necessary because of the complexity of the flight dynamics , which precludes an analytical approach [13] . Even setting aside the aerodynamic complexities that we have handled using a blade-element model of flapping flight , the mere fact that the birds must reorient their body to redirect their lift vector generates dynamics that are known to have no analytical solution in the most-closely analogous case of bank-to-turn missiles [25] . Our modeling therefore follows an embodiment approach , which states that behavior emerges through feedback-loops between the brain , the body , and the world . Aspects of cognition , such as the guidance laws used to intercept prey , are shaped by properties of the body and therefore bodily traits need to be considered to fully understand behavior [12 , 26] . In summary , our agent-based simulation approach provides insights into the optimization of attack strategies by an aerial predator that could not have been reached in any other way , and thereby paves the way for a new generation of studies into the optimization of complex multi-agent flight behaviors .
In our simulations , model-birds fly with six degrees of freedom through an open three-dimensional space without objects or boundaries . In each simulation run , a model-falcon aims to intercept a lone model-starling in mid-air , using a pure proportional navigation guidance law ( Eq 1 ) . In model-starlings , the guidance command is a forcing function that ensures that they either fly linearly , or execute smooth or non-smooth maneuvers , always keeping within ±20m of their initial altitude . Model-birds are subject to gravitational and aerodynamic forces , and flap , glide and retract their wings to manipulate the aerodynamic forces . Model-birds maximize their forward acceleration at a given speed and orientation , subject to the constraint that they meet the normal acceleration commanded by their guidance system . A flight controller determines the changes in wing shape and motion that best meet the desired acceleration . When the commanded normal acceleration cannot be met , model-birds simply exert the maximum attainable lift force . At the start of an attack , the model-starling is located at the origin of the global coordinate system , with its body coordinate system oriented randomly . This variation in initial orientation ensures sufficient randomization to avoid artificial results due to coupling of highly specific initial conditions of the falcon and starling . The model starling begins flying at an initial speed of 11 ms−1 , calculated as the airspeed at which the cost of transport is minimized under the model . The model-falcon initially flies at a speed of 16 ms−1 , with its longitudinal body axis pointing directly towards the starling , and its lateral body axis horizontal . We parametrically vary the falcon’s initial position relative to the prey , and vary the navigation constant N ( i . e . the one free parameter of the falcon’s guidance law; see below ) to simulate a continuum of different attack strategies . For each attack , we sample at random from a uniform distribution , sampling the navigation constant N between 1 and 20 , the falcon’s initial altitude above the prey between −200 and 1500 m , and the initial horizontal distance to the prey between 0 and 800 m . The simulation ends when the falcon either intercepts the starling or is unsuccessful in its attempt to intercept , according to the criteria defined below . For a visualization of the simulations , see SI videos . Every simulation ends in either the success or failure of the model-falcon to catch the model-starling . A catch is defined as occurring when the model-falcon comes within 0 . 2m of the model-starling . Failure occurs if either the falcon has not caught the starling within 40 s , or if it experiences a near-miss from which it cannot recover . A near-miss occurs when the model-falcon comes within 5 . 0 m of the model-starling , but subsequently finds itself further than this from the model-starling and with the model-starling in the blind zone of the model-falcon ( a cone of 45° behind the bird ) such that the falcon would effectively need to begin a new engagement in order to re-acquire its target . In order to analyze how the model parameters affect catch success , we apply Generalized Additive Modeling ( GAM; [19 , 20] ) . This is a nonlinear regression method which places no assumptions on the shape of the relationship between predictor and outcome . The estimation of the smoothing functions is conducted by automated cross-validation procedures ( quadratic penalized likelihood ) , which reduce the likelihood of over-fitting and therefore ensure that our ( conditional ) maxima are not spurious . We applied GAMs with a logit link function , with catch success as the outcome variable and with the navigation constant N , initial-altitude and horizontal-distance as the independent variables . We built separate models for each combination of prey motion , response delay , and error . No constraints on the effective degrees of freedom were applied . Model simulations were programmed in C++ , using openGL for graphics rendering . Hildenbrandt’s StarDisplay model [27] was used as the framework for graphical display . Optimization studies of the blade-element model were conducted in MATLAB 2014a , and the mgcv package [28] of R statistics [29] was used for GAM regression . Here we explain the detail of our simulation model , using the block structure depicted in Fig 2 , and discussing each of the following four segments of the block diagram in turn: A . Kinematics , B . Vision , C . Guidance , D . Control and E . Aerodynamics . We justify each variable and mechanism by parameterization to empirical data , and justify mathematical argument in terms of physics or optimality . For symbol meanings , see Table 2 .
|
Peregrine falcons are famed for their high-speed , high-altitude stoops . Hunting prey at perhaps the highest speed of any animal places a stooping falcon under extraordinary physical , physiological , and cognitive demands , yet it remains unknown how this behavioural strategy promotes catch success . Because the behavioral aspects of stooping are intimately related to its biomechanical constraints , we address this question through an embodied cognition approach . We model the falcon’s cognition using guidance laws inspired by theory and experiment , and embody this in a physics-based simulation of predator and prey flight . Stooping maximizes catch success against agile prey by minimizing roll inertia and maximizing the aerodynamic forces available for maneuvering , but requires a tightly tuned guidance law , and exquisitely precise vision and control .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"acceleration",
"ornithology",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"dynamics",
"classical",
"mechanics",
"bird",
"flight",
"vertebrates",
"starlings",
"biological",
"locomotion",
"animals",
"flight",
"(biology)",
"animal",
"anatomy",
"zoology",
"birds",
"wings",
"community",
"ecology",
"physics",
"eukaryota",
"ecology",
"predation",
"physiology",
"trophic",
"interactions",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"amniotes",
"organisms",
"aerodynamics"
] |
2018
|
Physics-based simulations of aerial attacks by peregrine falcons reveal that stooping at high speed maximizes catch success against agile prey
|
Human African Trypanosomiasis ( HAT ) is a neglected disease targeted for elimination as a public health problem by 2020 . Elimination requires a better understanding of the epidemiology and clinical evolution of HAT . In addition to the classical clinical evolution of HAT , asymptomatic carriers and spontaneous cure have been reported in West Africa . A genetic component to human susceptibility to HAT has been suggested to explain these newly observed responses to infection . In order to test for genetic associations with infection response , genetic polymorphism in 17 genes were tested ( APOL1 , IL1B , IL4 , IL4R , IL6 , IL8 , IL12B , IL12RB1 , IL10 , TNFA , INFG , MIF , HLA-G , HLA-A , HP , HPR and CFH ) . A case-control study was performed on 180 blood samples collected from 56 cases and 124 controls from Cameroon . DNA was extracted from blood samples . After quality control , 25 samples ( 24 controls and 1 case ) were eliminated . The genotyping undertaken on 155 individuals including 55 cases and 100 controls were investigated at 96 loci ( 88 SNPs and 8 indels ) located on 17 genes . Associations between these loci and HAT were estimated via a case-control association test . Analyses of 64 SNPs and 4 indels out of 96 identified in the selected genes reveal that the minor allele ( T ) of rs8062041 in haptoglobin ( HP ) appeared to be protective against HAT ( p = 0 . 0002395 , OR 0 . 359 ( CI95 [0 . 204–0 . 6319] ) ) ; indicating higher frequency in cases compared to controls . This minor allele with adjusted p value of 0 . 0163 is associated with a lower risk ( protective effect ) of developing sleeping sickness . The haptoglobin related protein HPR and HP are tightly linked and both are duplicated in some people and may lead to higher activity . This increased production could be responsible of the protection associated with rs8062041 even though this SNP is within HP .
The study was conducted in three active sleeping sickness foci of the forest region of Southern Cameroon ( Fig 1 ) . The Cameroonian population is made up of more than 240 ethnic groups that can be grouped into Bantu ( e . g . : Beti , Bassa , Bakundu , Maka , Douala , Pygmie ) , Semi Bantu ( e . g . : Bamileke , Gbaya , Bamoun , Tika ) and Sudano-Sao ( e . g . : Fulbe , Mafa , Toupouri , Shoa-Arabs , Moundang , Massa , Mousgoum ) . The composition varies considerably between HAT foci and even within the same HAT focus . The three HAT foci where this study was undertaken were Bipindi and Campo in the Southern region and Fontem in the South-west region of Cameroon . The Campo focus ( 2°82'00"N , 9°85'20"E ) is located in the tropical forest and extends from the Atlantic coast along the Ntem river which delimits the Cameroon–Equatorial Guinea border . It is a hypo-endemic focus with no history of epidemic outbreaks [23] and a cumulative number of 98 cases were detected between 1998 and 2013 . The main source of livelihood for the inhabitants of the Campo focus is agriculture , fishing and hunting . It is a cosmopolitan area with several ethnic groups including mainly the Iyassa , Kwasse , Maabi , Mvae and Ngoumba , most of whom are Bantu speaking . Other minor ethnic groups are semi Bantus and Sao-Sudanese and can be found at Campo for administrative and socioeconomic purposes . The Bipindi HAT focus ( 3°82'00"N , 10°82'20"E ) is located at about 75 km from the Atlantic coast in the South of Cameroon . It is an old HAT focus that has been known since 1920 . During the last two decades , the Bipindi focus was among the most active HAT foci of Cameroon with around 83 HAT cases diagnosed from 1999 to 2011 [24] . About 95% of the inhabitants of the Bipindi HAT focus are Bantu speaking and belong to ethnic groups such as Ngoumba , Nti , Fan and Pygmies . The remaining 5% of inhabitants ( semi Bantus and Sao-Sudanese ) are there for administrative and socioeconomic purposes . The main livelihood for people in this focus is hunting , farming and seasonal harvesting of fruits . The Fontem focus ( 5°40’00”N , 9°55’00”E ) is located in the South-West Region of Cameroon where HAT has been known to occur since 1949 [25] . The Fontem focus was previously among the most active HAT foci of Cameroon [26] , but in recent decades , it has become hypoendemic with about 8 patients detected among 16 , 000 persons examined between 1998 and 2007 [27] . In this focus , the Mundani , Bamoua and Bangwa are the major ethnics groups . Other minor ethnic groups such as Banyangue and Bamileke are also found .
The blood samples were collected during medical surveys performed jointly with the national sleeping sickness control program of Cameroon . The sampling was done at Campo in 2014 and for Bipindi and Fontem , in 2015 . During these surveys , all inhabitants were screened with CATT test [28] on whole blood . All inhabitants with positive CATT test were subjected to CATT dilution on plasma and each inhabitant positive on a CATT dilution ≥1/8 was subjected to parasitological examination ( capillary tube centrifugation ( CTC ) [29] and minianion exchange centrifugation technique ( mAECT ) ) [30] . For all inhabitants with CATT dilution ≥1/8 and negative for all parasitological tests , 90μl of plasma were spotted on a Whatman paper disc ( divided in four equal parts with each bearing a spot of 30μl ) that was sent to CIRDES in Burkina Faso for the trypanolysis test [31] . Beside the CTC and mAECT , lymph node aspiration followed by a microscopic examination was performed to search for trypanosomes in all individuals showing enlarged lymph nodes . A new HAT case was defined as an inhabitant in whom trypanosomes were seen by at least one parasitological method . Beside these new HAT cases , old HAT cases were also resampled . Old HAT cases were residents in whom trypanosomes had been previously seen on at least one parasitological test after passive or active case detection . Old HAT cases were only included in this study if the information regarding the clinical status , the CATT test and all parasitological tests were available in hospital records . A control was considered as any individual negative for the CATT test , all parasitological tests including CTC , mAECT and lymph node examination and when possible the trypanolysis test . With these sampling criteria , 5ml of peripheral venous blood samples were collected from cases and controls into EDTA coated tubes . In the field , the tubes were mixed gently and stored at 4°C in an electric cooler before being transported to the laboratory . The protocol of this study was approved by the Ethical Committee of the Ministry of Public Health of Cameroon reference number N°2013/11/364/L/CNERSH/SP of 21 November 2013 . The local administrative and traditional authorities of each HAT focus were also informed and gave their approval . Subsequently , the review board ( LAMAS ) of Laboratory of Microbiology and Anti-microbial substances of the Department of Biochemistry of the Faculty of Science of the University of Dschang gave their approval . All adult subjects provided informed consent , and a parent or guardian of any child participant below 18 years old provided informed consent on their behalf . Each informed consent was written because all individuals enrolled in this study gave their approval by signing an informed consent form and a Certificate of Confidentiality . During analyzes , data of each subject were anonymized . Blood samples were centrifuged at 5000rpm for 3 minutes and the buffy coat was collected . Genomic DNA was extracted from the Buffy-coat with the QIAamp DNA Blood Midi/Maxi kit ( Qiagen ) according to the manufacturer's instructions . The DNA was eluted with 200μl of elution buffer and stored at -20°C until use . Power calculations were undertaken using the genetics analysis package gap in r [32] The choices of candidate genes were based on previous observations . The cytokines IL4 , IL6 , IL10 , IL8 , INFG , TNFA , HP , HPR and MHC gene HLA-G were selected because they have been previously associated with HAT [14–16 , 20 , 33 , 34 , 35 , 36 , 37] . In addition , two genes for factors involved in the lysis of trypanosomes , APOL1 and haptoglobin-related protein ( HPR ) were also included [17 , 18 , 19] . Five further genes that had previously been reported to play an important role in the susceptibility to other infectious diseases were selected: Human Leukocytes Antigen A ( HLA-A ) [38 , 4 , 39] , IL1B [10] , Complement factor H ( CFH ) [6 , 10] , IL12B and IL12RB1 [5 , 11] and Macrophage migration inhibitory factor ( MIF ) [40 , 41 , 42] genes were also included . Most SNPs and indels for testing were selected after a Linkage scan ( r² = 0 . 5 ) and quality control with Plink version 1 . 9 [43] using whole genome sequencing data . These data were obtained from a merged dataset between the African populations data from the 1000 Genomes Project combined with low fold coverage ( 8-10x ) whole genome shotgun data generated from 230 residents living in regions ( DRC , Guinea Conakry , Ivory Coast and Uganda , European Genome Archive A accession number ) where trypanosomiasis is endemic [44] . The 88 SNPs and 8 indels loci were selected by two strategies: 1 ) by linkage scan of SNPs and indels ( r2 < 0 . 5 ) across the gene; 2 ) by selection of SNPs and indels with published associations with HAT . Linked SNPs were identified for IL6 , IL4 , IL8 , IFNG and HLA-G genes . For APOL1 , HPR , HP , HLA-A , IL1B , IL12B , IL12RB1 , IL4R , CFH , IL10 , MIF and TNFA genes , individual published SNPs and indels were identified and selected based on literature searches . Samples which had low DNA concentration or did not satisfy the quality control criteria were excluded prior to genotyping . Genotyping was performed by two commercial service providers: 1 ) “Plateforme Genome Transcriptome” at INRA of Bordeaux in France; 2 ) LGC Genomics Hoddesden , UK with approximately 1μg of genomic DNA per sample . At INRA , genotyping was carried out with a Multiplex design ( two sets of 40 SNPs or indels ) using Assay Design Suite v2 . 0 ( Agena Biosciences ) . For each SNP and indel , the genotyping was done with the iPLEX Gold genotyping kit ( Agena Biosciences ) for the Mass-Array iPLEX genotyping assay according to the manufacturer’s instructions . Products were detected on a Mass-Array mass spectrophotometer and data were obtained in real time with Mass-Array RT software ( Agena Biosciences ) . SNP clustering and validation was carried out with Typer 4 . 0 software ( Agena Biosciences ) . A summary of the candidate genes , and SNPs and indels is shown in the supplementary data S1 Table . Some SNPs and indels that failed genotyping at INRA and some additional SNPs and indels were genotyped at LGC Genomics , Hoddesden , UK where SNPs and indels were genotyped using the PCR based KASP assay [45] . This was a case-control study where no familial controls were collected during sampling . The raw genotypic data were converted to PLINK format and quality control ( QC ) procedures implemented using the PLINK v1 . 9 package [43] . The Spearman Chi-square test was used to compare frequencies of observed and expected genotypes under Hardy–Weinberg equilibrium ( HWE ) and LD using R/Rstudio version 3 . 3 . 2 ( 2016-10-31 ) —‘Sincere Pumpkin Patch’ and Plink [43] . After quality control and filtering , poorly performing SNP loci with missing genotypes ( ≥10 ) and samples with missing loci ( ≥4 ) were removed . In addition , all loci with a MAF below 1% or a HWE P value < 1 × 10−4 were removed . SNP in linkage with adjacent SNP ( r² > 0 . 5 ) were also pruned . These filters are as described by Anderson et al . [46] to minimize the influence of genotype-calling artifacts in a candidate gene study . The association between individual SNPs and indels within genes and HAT were tested using the Fisher exact test with Plink v1 . 9 software . Results were adjusted for multiple testing by Bonferroni correction . To show significant association during multiple tests , a single marker ( SNP ) must show , after Bonferroni correction , an alpha value ( obtained P value before correction/number of SNPs analyzed ) below 0 . 000746 ( 0 . 05/68 ) . The Bonferroni correction assumes that each of the statistical tests is independent; however , this is not always true due to the possibility of linkage disequilibrium among the SNPs . In instances where the assumption is not true , the correction is often too strict , leading potentially to false negatives . A less stringent correction for multiple testing was also employed . The Benjamini-Hochberg false discovery rate ( FDR ) estimates the proportion of significant results ( P < 0 . 05 ) when the Bonferroni correction considers them as false positives [47 , 48] . FST is a measure of differences between populations . The analysis of FST was run to check for significant allele frequency difference between the cases and controls while Principal Component Analysis ( PCA ) was used to check for population stratification that might confound the analysis using Plink [43] . European Genome Archive A accession number: EGAS00001002602 .
This study was one of six studies of populations of HAT endemic areas in Cameroon , Cote d’Ivoire , Guinea , DRC , Malawi and Uganda . The studies were designed to have 80% power to detect odds ratios ( OR ) >2 for loci with disease allele frequencies of 0 . 15–0 . 65 and 100 cases and 100 controls with the 96 loci genotyped . Overall , 216 individuals were included in this study: 56 ( 25 . 93% ) HAT patients and 160 ( 74 . 07% ) controls . The 216 individuals belonged to 22 different ethnic groups . The mean age ( range ) of HAT cases was 44 . 94 ( 15–82 ) years , while that of controls was 37 . 08 ( 9–86 ) . The overall sex ratio ( male/female ) was 1 . 02 ( 109/107 ) , with HAT cases being 0 . 75 ( 24/32 ) and controls 1 . 12 ( 84/75 ) . Given that only 56 cases were available from our study area , the power of this study was reduced and it had 80% power to detect an OR >3 with disease allele frequencies of 0 . 1–0 . 45 with the 96 loci genotyped . One hundred and eighty ( 56 HAT cases and 124 controls ) of the 216 samples were sent for genotyping . After DNA quantification and quality control on each of these 180 samples , 25 were excluded from genotyping . 155 samples were genotyped: 55 ( 34 . 48% ) HAT cases and 100 ( 64 . 52% ) controls . 96 loci containing 88 SNPs and 8 indels were tested from 17 candidate genes . The number of SNPs and indels analyzed varied considerably ( from 1 to 18 ) between genes ( Table 1 ) . The highest number of 18 SNPs and indels was observed for HLA-G and the lowest number of one SNP for IL10 , IL1B and CFH . However , it is important to point out that for APOL1 ( three SNPs ) , CFH , TNFA , HLA-A and IL10 , the SNPs considered here are only those that have been already reported in the literature . Of the 88 SNPs and 8 indels used in this study , 24 SNPs and 4 indels ( with 8 removed for MAF ≤1% , 7 for missing loci ≥10% , 5 with HWE P-values <1 x 10−4 and 8 for linkage at r² ≥ 0 . 5 ) of them were excluded during quality control which excluded one gene ( HLA-A ) completely . Four samples were also excluded during quality control due to missing individual data ≥4% . For subsequent analyses , 69 . 29% of loci including 64 SNPs and 4 indels from 16 genes and 151 ( 97 . 42% ) samples will be considered for association analysis ( Table 1 ) . The principal components ( S1 Fig: supplementary data ) and FST values ( S2 Table: supplementary data ) analysis showed that cases and controls were evenly dispersed ( homogenous and samples did not cluster by phenotype ) ; indicating that the population and subpopulation structure is not the driving force in our observations . Alleles for the sixteen genes and 64 SNPs and 4 indels analyzed were all in Hardy–Weinberg equilibrium ( S2 and S3 Tables: supplementary data ) ; suggesting random genetic exchange within the studied populations . The MAF varied considerably across SNP and indel ( S2 Table: supplementary data ) with the lowest MAF at rs11575934 in IL12RB1 ( MAF = 0 . 0067 ) and the highest value at rs371194629 in HLA-G ( MAF = 0 . 5 ) . The minor allele ( T ) of rs8062041 in HP appeared to be protective against HAT ( p = 0 . 00024 ) . An odds ratio ( OR ) of 0 . 359 ( CI95 [0 . 20–0 . 63] ) indicated low frequencies in cases compared to controls . This SNP is located in a copy number variation ( CNV ) essv41754 that spans both HP and HPR ( Fig 2 ) . In addition , the minor alleles of IL4 and HLA-G also appeared protective ( IL4: C rs2070874 , uncorrected p = 0 . 047: and HLA-G: G rs1233330 , uncorrected p = 0 . 011 ) . The OR of 0 . 62 ( CI95 [0 . 38–1 . 01] ) for rs2070874-IL4 C and 0 . 2754 ( CI95 [0 . 093–0 . 81] ) for rs1233330 HLA-G also indicated low frequencies of the major allele in cases compared to controls . However , for HLA-G , the minor allele ( A ) of SNP rs17875389 had a higher frequency in cases than controls ( p = 0 . 042 ) . The OR of 2 . 29 ( CI95 [0 . 97–5 . 39] ) suggests that the A allele may increase the risk of developing HAT . Of the 64 SNPs and 4 indels considered here , only four ( SNP ) of them belonging to three genes were associated with the development of HAT before Bonferroni correction ( Table 2 ) . After Bonferroni correction only one SNP ( rs8062041 T/C ) in HP was associated with HAT . The odds ratio of 0 . 359 suggests that the minor allele has a protective effect within the Cameroonian population with 95 . 3% ( FDR ) chance of this locus being associated with HAT ( Table 2 ) . For the three remaining SNPs where the association was not significant after Bonferroni correction , our results show that the allele frequencies in cases and controls were not the same ( Table 2 ) and that there is some possibility of an association with disease . FDR_BH is the probability of falsely rejecting the null hypothesis that allele frequencies are the same in cases and controls . rs1233330 and rs17875389 in HLA-G had FDR_BH values of 0 . 36 and 0 . 67 respectively; suggesting that there are 64% and 33% probabilities of an association between SNPs at these loci with HAT . For rs2070874 in IL4 , the FDR_BH value of 0 . 722 suggests 27 . 8% chance of an association with HAT . For the other genes ( APOL1 , IL4 , IL6 , IL10 , IL8 , TNFA and INFG ) involved in immune response that have been previously investigated in HAT , our results revealed no statistical association with the disease within the Cameroonian population ( S1 Table ) . No association was also observed with SNP and indel of APOL1 and all SNPs of HLA-A , IL1B , IL12B , CFH , IL12RB1 , IL4R and MIF previously associated with the susceptibility to other infectious diseases ( S2 Table: supplementary data ) .
In this study we obtain good quality genotype data for a total of 64 SNPs and 4 indels in 16 genes to investigate associations with trypanosomiasis . Of these genes selected on the basis of their association with HAT or other infectious diseases , most were not statistically associated with HAT in Southern Cameroon . The most important result of this study is the observation that the T allele of SNP rs8062041-HP with a p-value of 0 . 00024 ( Bonferonni corrected p = 0 . 015 ) and an OR of 0 . 36 is associated with a lower risk ( protective effect ) of developing sleeping sickness . This SNP lies within intron 1–2 of HP of the CNV essv41754 that spans both HP and HPR transcripts ( Fig 2 ) . Although the biological significance of this CNV is not well understood , it is important to point out that HP and HPR have some biological similarities . Haptoglobin is involved in the scavenging of haem from lysed red blood cells . Trypanosome infections induce extensive lysis of red cells releasing haem which is scavenged by HP . In mice , the expression of the haptoglobin receptor ( Cd163 ) on macrophages declines dramatically after infection with T . congolense [49] and is the earliest indicator of infection . HPR also binds haem but is not cleared from circulation after haemolysis . However , HPR is of particular interest because it plays a prominent role in the innate resistance of humans to most Trypanosoma species [50] . This innate resistance is linked to trypanosome lytic factors 1 and 2 ( TLF1 , TLF2 ) which are bound to a minor subclass of high-density lipoprotein ( HDL ) [51] . Both factors harbor APOL1 , which is the trypanolytic component [52] , and HPR which facilitates the uptake of APOL1 via trypanosome haptoglobin–hemoglobin receptors ( HpHbR ) . Interestingly , rs8062041-HP ( T/C ) is located on chromosome 16 ( 16q22 . 2 ) in the CNV essv41754 that spans both HP and HPR ( Fig 2 ) . Such genomic structural variants involving HP/HPR duplication have been reported with higher frequency in people of African descent [53 , 7] . For instance , HP and HPR have been reported in 29 independent studies listed in the Database of Genome Variants [54] . T . b . gambiense protects itself against killing by APOL1 by reducing the abundance and affinity of the receptor for HPR [55] . If rs8062041 , located in the CNV essv41754 spanning HP and HPR is correlated with CNV genotype , then an increase in HPR expression could drive increased uptake of APOL1 and parasite killing . As in other diseases such as heart disease , cancer , malaria and Crohn’s disease , polymorphism in HP could also have direct biological significance in HAT . Polymorphism in the haptoglobin gene may be associated with reduced cholesterol levels in the blood [56] and since cholesterol is specifically taken up by trypanosomes as a nutrient , any reduction in cholesterol might restrict parasite growth rate . There are numerous variants of HP , some of which may have arisen from gene conversion from HPR exons [56] . Single SNP tag these variants poorly ( max r2 = 0 . 44 ) , however SNP haplotypes can tag these variants efficiently ( max r2 = 0 . 92 ) and are more strongly correlated with cholesterol levels than individual SNP [56] . High density genotyping of the HP/HPR locus will be required to understand the role of this locus in the response to trypanosome infection . Although rs8062041 is within HP , the known involvement of HPR in APOL1 mediated killing means that increased expression of HPR is another mechanism by which this SNP could be associated with the observed difference in likelihood of developing HAT . Our results showing an association between HAT and one SNP located within a CNV spanning HP and HPR duplication are not in line with results of Hardwick et al . [18] who observed no association with the HPR duplication allele and HAT in DRC . The difference between these results could be linked to the position of SNP within HP , the genetic diversity between the studied populations as well as the sampling methods . In our study , a case control approach was used while Hardwick et al . [18] used family-based sampling . Bresalier et al . [57] reported an association between polymorphisms at some HPR loci with an increasing risk of developing colon cancer . Similar associations were outlined by Tabak et al . [58] for HPR/APOL1 loci variations in hepatoma and leukemia . There are also examples of CNV mediating different susceptibilities to infectious diseases [59 , 60] . Of the twelve SNPs of IL6 identified and investigated in our study , none of them revealed an association with HAT . However , with similar investigation on the same gene , Courtin et al . [15] showed a T allele of the IL6 ( 4339 ) SNP rs2069849 which was significantly ( Bonferroni corrected p = 0 . 04 ) associated with a decreased risk of developing HAT in the DRC . This SNP was not genotyped in this study because it could not be multiplexed with the others in the panel . The discrepancy between our results and those of Courtin et al . [15] could be due to insufficient linkage between our marker SNP and rs2069849 and or genetic differences between the DRC and Cameroon populations . The study designs also differed; we used a case control approach while Courtin et al . [15] used a family-based design and our study was smaller . It has been suggested that the G1 and G2 alleles of APOL1 which increase the risk of developing kidney disease are under selection because they confer resistance to HAT [17] . Our observations on APOL1 are consistent with Cooper et al . [22] who found no association with APOL1 G1 and G2 in a comparison of cases and active T . b . gambiense HAT . Concerning HLA-G , our results showed a protective effect of developing HAT for the loci rs17875389 G/A ( p = 0 . 0416 and OR of 2 . 291 ) and an increased risk effect for rs1233330 A/G ( p = 0 . 01105 and OR of 0 . 2754 ) . These results support those of Courtin et al . [14] who reported similar results for different SNPs of the same genes in the DRC . The association with IL4 rs2070874 T/C ( p = 0 . 00712 and OR of 0 . 6151 ) is the first time this has been observed in HAT although associations with IL4 have been observed in South American trypanosomiasis [61 , 62] . The presence of IL4 in extravascular tissues promotes alternative activation of macrophages into M2 cells and inhibits classical activation of macrophages into M1 cells . This increase in repair macrophages ( M2 ) is coupled with secretion of IL10 and TGFB that result in a diminution of pathological inflammation [63] . The results discussed above for IL4 and HLA-G are based on FDR_BH values should be used with caution because no association was found after correction for multiple testing . However , we were only able to collect a relatively small number of cases ( 56 ) for this study , despite conducting large-scale field surveys . Whilst our power calculations indicated that effects of the sizes observed could be detected with our relatively small number of samples , larger cohorts of well phenotyped cases and controls may be required to confirm these observations . Therefore , although the present data is only suggestive of an association , the finding of suggestive associations in multiple populations increases the probability that these are genuine associations with disease [64] . This challenge is precisely what the TrypanoGEN network , a consortium of partners in eight African and three European countries seeks to address . The network has collected from seven regions in six countries ( Cameroon , Cote d’Ivoire , DRC , Malawi , Uganda , and Zambia ) a total of 3301 samples from cases and controls to include in a genome-wide-association study [44] which will be used to test the hypotheses generated here .
The results of this study reveal an absence of association between HAT and several SNPs identified in genes previously associated with HAT within inhabitants of sleeping sickness foci of other African countries . An association between one SNP in HP and the susceptibility to HAT was revealed in inhabitants of sleeping sickness foci of Cameroon . Located within a CNV that spans both HP and HPR and given the known involvement of HPR in response to HAT , the association of rs8062041 with a CNV is the most plausible mechanism by which this SNP could be associated with protection against HAT . Our results reveal also that the association between host genetic determinants or gene polymorphisms and the susceptibility to T . b . gambiense infections may vary according to studied populations .
|
Human African trypanosomiasis ( HAT ) or sleeping sickness is a neglected tropical disease targeted for elimination by 2020 . This elimination requires a better understanding of the epidemiology and clinical evolution of this disease . Beside the classical clinical evolution , asymptomatic carriers , seropositive and spontaneous cure of infected persons have been reported in West Africa . Arguments in favor of human genetic susceptibility to HAT have been raised to explain this variability in clinical presentation . This study investigated the genetic polymorphism of 17 genes between controls and sleeping sickness patients in Southern Cameroon in order to improve our knowledge of human susceptibility to trypanosome infections . We identified single nucleotide polymorphisms and indels in 17 selected genes involved in immune responses and carried out a case-control candidate gene association study and demonstrated differences between variants associated with the disease . From these genes , only haptoglobin ( HP ) at the SNP rs8062041 was found to have polymorphisms which were strongly associated with trypanosomiasis . The minor allele ( T ) at this SNP position appeared to be protective against HAT ( p = 0 . 0002395 , OR 0 . 359 ( CI95 [0 . 204–0 . 6319] ) ) reducing the risk of developing disease approximately threefold . The haptoglobin related protein ( HPR ) is adjacent to HP and is a component of the Trypanolytic factor that kills trypanosomes . The HP and HPR locus is duplicated in some people . The rs8062041 variant may be associated with this duplication and it is possible that increased production of HPR is the cause of the protection associated with rs8062041 . The results reported here will contribute to the knowledge of the role of human genetics in disease progression , and thus lead to the identification of novel biomarkers which could involve development of new diagnostics , treatments and intervention strategies .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion",
"Conclusion"
] |
[
"medicine",
"and",
"health",
"sciences",
"african",
"trypanosomiasis",
"variant",
"genotypes",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"genetic",
"mapping",
"protozoans",
"neglected",
"tropical",
"diseases",
"molecular",
"genetics",
"molecular",
"biology",
"techniques",
"genotyping",
"africa",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"cameroon",
"zoonoses",
"proteins",
"protozoan",
"infections",
"trypanosomiasis",
"molecular",
"biology",
"genetic",
"loci",
"people",
"and",
"places",
"biochemistry",
"haptoglobins",
"trypanosoma",
"eukaryota",
"plasma",
"proteins",
"heredity",
"genetics",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2017
|
A polymorphism in the haptoglobin, haptoglobin related protein locus is associated with risk of human sleeping sickness within Cameroonian populations
|
Kaposi's sarcoma-associated herpesvirus ( KSHV ) is a γ-herpesvirus associated with KS and two lymphoproliferative diseases . Recent studies characterized epigenetic modification of KSHV episomes during latency and determined that latency-associated genes are associated with H3K4me3 while most lytic genes are associated with the silencing mark H3K27me3 . Since the latency-associated nuclear antigen ( LANA ) ( i ) is expressed very early after de novo infection , ( ii ) interacts with transcriptional regulators and chromatin remodelers , and ( iii ) regulates the LANA and RTA promoters , we hypothesized that LANA may contribute to the establishment of latency through epigenetic control . We performed a detailed ChIP-seq analysis in cells of lymphoid and endothelial origin and compared H3K4me3 , H3K27me3 , polII , and LANA occupancy . On viral episomes LANA binding was detected at numerous lytic and latent promoters , which were transactivated by LANA using reporter assays . LANA binding was highly enriched at H3K4me3 peaks and this co-occupancy was also detected on many host gene promoters . Bioinformatic analysis of enriched LANA binding sites in combination with biochemical binding studies revealed three distinct binding patterns . A small subset of LANA binding sites showed sequence homology to the characterized LBS1/2 sequence in the viral terminal repeat . A large number of sites contained a novel LANA binding motif ( TCCAT ) 3 which was confirmed by gel shift analysis . Third , some viral and cellular promoters did not contain LANA binding sites and are likely enriched through protein/protein interaction . LANA was associated with H3K4me3 marks and in PEL cells 86% of all LANA bound promoters were transcriptionally active , leading to the hypothesis that LANA interacts with the machinery that methylates H3K4 . Co-immunoprecipitation demonstrated LANA association with endogenous hSET1 complexes in both lymphoid and endothelial cells suggesting that LANA may contribute to the epigenetic profile of KSHV episomes .
Eukaryotic DNA is packaged into chromatin which plays a central role in the regulation of all DNA processes including replication , transcription , and repair . Chromatin contains nucleosomes with DNA wrapped around the core histones H2A , H2B , H3 , and H4 . Nucleosomes carry epigenetic information in the form of post-translational histone modifications . N-terminal histone modifications including acetylation , methylation , phosphorylation , and sumoylation are important in partitioning chromatin into transcriptionally active or repressive domains ( reviewed in [1] ) . In mammalian cells , genome-wide ChIP-seq assays revealed that histone acetylation at H3K9 and H3K4 trimethylation ( H3K4me3 ) correlate with active transcription , while H3K27 trimethylation ( H3K27me3 ) is detected in promoters of repressed genes [2] . The apparently opposite modifications H3K4me3 and H3K27me3 co-localize at some promoters ( “bivalent marks” ) , poising these genes to be transcribed upon signaling . Histone modifications are also detected in regions outside promoters . All three states of H3K4 methylation are highly enriched at insulator sites , while only H3K4me and H3K4me3 are associated with enhancers [2] , [3] . Histone lysine methylation is mediated in mammalian cells by a large family of lysine methyltransferases ( KMTs ) that exist in protein complexes . A single enzyme can be responsible for the three states of methylation in a progressive manner , or different enzymes may be required for different methylation states . Mammalian cells contain 10 different H3K4 KMTs , which include the hSET complex , mixed lineage leukemia 1 to 5 ( MLL1-5 ) complexes , Set7/9 , Smyd1 , Smyd3 , and Prdm9 , which are largely not redundant [3] , [4] , [5] . hSET1 and MLL complexes share three core components: WDR5 , RbBP5 , and ASH2L , and siRNA-mediated knockdown of these proteins leads to a significant reduction of global H3K4 methylation , strongly suggesting that hSET1 and MLL are responsible for the majority of H3K4 methylation [6] , [7] . It has been demonstrated that the hSET1 and MLL complexes can be recruited to specific promoters through interactions with transcription factors or co-activators including E2F , NF-E2 , MAPK , and USF1 [8] , [9] , [10] , [11] . Moreover , for HSV-1 , an α-herpesvirus , it was demonstrated that hSET1 or MLL complexes are recruited to IE promoters through a VP16/HCF interaction , the latter functioning as a scaffold for the activator complex [12] , [13] . Kaposi's sarcoma-associated herpesvirus ( KSHV , also named HHV8 ) is a γ is a named . In addition to Kaposi's sarcoma ( KS ) which targets endothelial cells , KSHV is associated with two lymphoproliferative disorders: primary effusion lymphoma ( PEL ) and a subset of multicentric Castleman's disease ( MCD ) . Although the majority of cells in KS tumors are latently infected , both latent and lytic phases of KSHV infection contribute to pathogenesis and tumorigenesis [14] . During latency , viral gene expression is restricted to a small subset of genes including the latency-associated nuclear antigen ( LANA ) , vCyclin , vFLIP , Kaposins , and viral miRNAs [15] , [16] . LANA is a multifunctional protein with important roles during viral latency . LANA is the only viral protein required for episome maintenance by supporting KSHV latent DNA replication and tethering the viral episome to cellular chromosomes . Tethering is mediated by LANA via interactions with the viral terminal repeats and the core histones H2A/H2B [17] , [18] , [19] , [20] , [21] . LANA associates with several chromatin modifying complexes including the histone H3 lysine 9 ( H3K9 ) methyltransferase SUV39H and the H3K9 demethylase KDM3A [22] , [23] . In addition , LANA binds to the histone acetyltransferase CBP and histone deacetylase complex mSin3 [24] , [25] . LANA contributes to activation and repression of host and viral genes , presumably by interacting with transcriptional activators ( i . e . Brd2/4 or RING3 , Sp1 , Ap1 ) [26] , [27] , [28] , [29] and repressors ( HP1 , Dnmt3 , and mSin3 ) [24] , [30] , [31] and proteins involved in chromatin remodeling ( FACT and CBP ) ( [25] , [32]; for recent review see [33] ) . We hypothesized that LANA plays a role in the establishment and maintenance of the KSHV epigenome . To address this question and to identify viral and cellular genes potentially regulated by LANA , we performed genome-wide ChIP-seq analyses for LANA , Pol II , and histone modifications . Our data confirm that during latency both active H3K4me3 and repressive H3K27me3 marks are associated with the viral episomes [34] , [35] . Interestingly , H3K4me3 marks are highly correlated with LANA occupancy at sites where the silencing mark H3K27me3 is excluded . Furthermore , it was demonstrated that LANA selectively associates with H3K4 lysine methyltransferase ( KMT ) hSET1 complexes . Our data suggest that LANA may directly contribute to the viral epigenome by binding to specific viral promoters and enhancers and by interacting with H3K4 KMT hSET1 complexes .
Histone marks on KSHV genomes have previously been mapped by PCR-based ChIP assays and ChIP-on-chip assays using tiling array hybridization [34] , [35] , [36] , [37] . We investigated the KSHV epigenome using ChIP-seq , which has been applied to the genome-wide analysis of epigenetic modifications in mammalian cells [1] , [2] . Based on studies demonstrating that the patterns for acetylated H3K9 and H3K14 were almost identical with H3K4me3 on the KSHV epigenome [34] , [35] , we characterized the transcription-associated mark H3K4me3 and the repressive mark H3K27me3 in combination with RNA polymerase II ( Pol II ) . ChIP-seq assays were performed in BCBL-1 cells with antibodies against H3K4me3 , H3K27me3 , and Pol II . Sequencing reads were sequentially aligned against KSHV ( accession number NC_009333 ) and human genome hg19 using Bowtie [38] . About 89 . 8–95 . 9% of tags were aligned to hg19 and 0 . 73–2 . 4% to KSHV . To determine reproducibility of ChIP-seq assays , we compared two biological replicate datasets of H3K4me3 ChIP-seq in BCBL-1 cells using a Bland-Altman analysis [39] , [40] , [41] , [42] . As shown in Figure S1 , the 95% confidence interval shown between the green lines indicates high reproducibility . Data have been submitted to NCBI GEO ( accession number GSE52421 ) . ChIP-seq tags mapped to KSHV were used for peak analysis by CisGenome [43] . Genome-wide profiling of H3K4me3 , H3K27me3 , and Pol II occupancy on the KSHV genome in BCBL-1 cells was visualized using the UCSC Genome Browser ( Fig . 1 ) . Previously published nucleotide ( nt ) numbers from Gene Bank accession # U75698 are converted to accession # NC_009333 as referred to in Table S1 . Control IgG gave low background while H3K4me3 and H3K27me3 yielded specific occupancy patterns ( Fig . 1 ) . Within the unique long region , multiple H3K4me3 peaks are located at the KSHV latency-associated region ( KLAR ) including the LANA ( ORF73 ) promoter , a broad region from the beginning of ORF72 coding sequence to the beginning of the miRNA cluster , and the intragenic region in between K12 and miRNAs ( Fig . 1 ) . The latency-associated region is a complex locus containing at least three promoters driving the expression of LANA , vCyclin , vFLIP , miRNAs , and the Kaposin family of proteins [26] , [44] , [45] . Distribution of H3K4me3 in this region is shown in more detail in Fig . S2 , and is consistent with expression of this region during latency . In addition , several lytic genes including ORF8/ORF9 , K4 . 2 , ORF50 ( RTA ) , K7 , K8 , vIRFs and ORF58 were enriched for H3K4me3 at different levels ( Fig . 1 , marked by asterisks ) . Unlike H3K4me3 which forms distinct peaks , H3K27me3 is distributed more broadly across large regions containing late lytic genes that are void of H3K4me3 ( Fig . 1 ) . Pol II occupancy was probed with an antibody that recognizes both elongating and pausing Pol II [46] , [47] , [48] and displayed a number of distinct peaks within the latency-associated region that coincide with H3K4me3 . Outside of this region , the highest Pol II occupancy was detected in a region spanning ORFs K4 . 2 to K7 ( asterisk ) . At most genomic loci H3K4me3 and H3K27me3 are mutually exclusive . Several blocks of lytic late genes , including loci from the beginning of the genome to 9 . 5 K , regions spanning 30 K to 60 K and 77 K to 83 K , are enriched with H3K27me3 but void of H3K4me3 and Pol II , indicating heterochromatin structure ( Fig . 1 ) . We focused in more detail on H3K4me3 , H3K27me3 , and Pol II occupancy at the +/−2kb region surrounding known transcription start sites ( TSS ) of selected viral genes ( Fig . 2 ) . Promoters for the latent genes ORF73 ( LANA ) ( Fig . 2 ) and vIRF-3 are enriched for H3K4me3 and Pol II but depleted for H3K27me3 , as was the transcription start site for vIL-6 ( Fig . 2 ) . Recent transcriptome profiling and chromatin structure analysis showed that the vIL-6 promoter is active in a subpopulation of PEL cells during latency [49] , [50] . Promoters for many additional viral genes displayed enrichment for H3K4me3 and Pol II . For example , the promoter of the lytic gene K7 is significantly enriched with H3K4me3 and Pol II , but depleted for H3K27me3 . Although these epigenetic marks suggest transcriptional activity , it was recently demonstrated by Toth et al . that transcription of K7 is paused at the elongation step through NELF binding to pol II [51] . As expected , lytic late gene promoters are enriched for H3K27me3 as exemplified by ORF25 and ORF38 , encoding a major capsid protein and a tegument protein , respectively ( Fig . 2 ) . The promoter for the lytic immediate early gene RTA is enriched for Pol II and both H3K4me3 and H3K27me3 ( bivalent marks , Fig . 2 ) . For validation , quantitative ChIP-PCR assays confirmed ChIP-seq results for the promoter regions of LANA , vIRF1 , vIL6 , RTA , and K7 , and the coding region of the late lytic gene ORF19 , which was enriched with H3K27me3 ( Fig . 3 ) . In summary , these H3K4me3 and H3K27me3 ChIP-seq profiles are in agreement with the general patterns from previously published ChIP-on-chip studies [34] , [35] . To date , epigenetic modifications have only been mapped in cells of lymphoid and epithelial cells but not in endothelial cells , which give rise to KS . To address this gap in the literature , we performed ChIP-seq assays in long-term infected TIVE-LTC cells , which contain BCBL-1 derived episomes [52] , [53] . Genome-wide occupancy of H3K4me3 , H3K27me3 , and Pol II is depicted in Figure 4A . The viral copy number in TIVE-LTC is less than 5 copies per cell , which is comparable with KS lesions in vivo [53] . As a result , the total number of KSHV-specific sequence tags was 25- to 165-fold lower in TIVE-LTC compared to BCBL-1 ( Compare y-axis in Fig . 4A and Fig . 1 ) . Because the total number of reads for KSHV in TIVE-LTC cells was low , we increased coverage by applying SureSelect target enrichment technology ( Agilent ) . ChIP-seq libraries were incubated with a custom-designed KSHV-specific biotin-labeled RNA bait library which yielded 2 , 000- to 4 , 000-fold enrichment . While numbers of tags per peak increased , ChIP-seq profiles were similar to those observed without enrichment indicating non-biased selection ( comparing corresponding tracks in Fig . 4A to Fig . 4B ) . H3K4me3 occupancy showed significant differences in TIVE cells compared to BCBL-1 cells within the latency-associated region . A prominent H3K4me3 peak within the ORF73 ( LANA ) promoter ( around nt 127 , 600 ) observed in BCBL-1 , was demonstrably reduced in TIVE-LTC cells . Instead , a strong H3K4me3 peak appeared at nt 126 , 280 within the LANA coding region , which was not present in BCBL-1 cells ( Fig . S2 panel B ) . Interestingly , in addition to H3K4me3 and Pol II , this region was also enriched for H3K27me3 thereby creating a bivalent mark . Outside of the latency-associated region occupancy of H3K4me3 was decreased at several areas including the K4 . 2 promoter region ( around 23 K ) , ORF50 promoter region ( around 72 K ) , 96 K , and 103 K region ( Fig . 4 ) . Conversely , H3K27me3 signals increased significantly within the LANA coding region ( shown as two peaks covering nt 124 . 1 K to nt 125 . 5 K and nt 126 K to nt 127 K ) . Pol II occupancy was significantly decreased throughout , especially at the K4 . 2 promoter region ( at about nt 24 K ) . In this context , it is interesting to note that TIVE-LTC cells are tightly latent [53] and we recently demonstrated that a subpopulation of episomes is heterochromatinized at the latency-associated region [49] . Whether the observed increased H3K27me3 and decreased H3K4me3 deposition are causative for the lack of reactivation in TIVE-LTC needs to be further investigated . In summary , except for a few changes affecting H3K4me3 deposition overall histone modification patterns were similar between PEL and endothelial cells . In addition to its role in latent DNA replication and episomal maintenance , LANA is a key regulator of host and viral gene expression . LANA binds to DNA directly in a site-specific manner , or indirectly through protein-protein interactions with multiple chromatin associated proteins including core Histone H2A and H2B , CREB2 , mSin3 , RING3 , MeCP2 , SSRP1 , and P53 ( reviewed in [54] , [55] ) . Hence , we hypothesized that LANA plays a role in the establishment and maintenance of the KSHV epigenome . To determine LANA occupancy on the viral and host genomes and to identify genes potentially regulated by LANA , we performed ChIP-seq using a monoclonal rat LANA antibody . Sequencing generated 5 million tags for rat IgG control , and between 14 . 7 and 36 . 7 million tags for BCBL-1 , and TIVE-LTC cells with and without target enrichment . Genome-wide binding of LANA is depicted in Figure 5 and major peaks observed in BCBL-1 are listed in Table 1 . Three LANA binding sites ( LBSs ) have previously been characterized by EMSA in vitro; two are located within the TR and contribute to latent DNA replication [20] , [21] , [56] and one is upstream of the LANA promoter , which is auto-regulated [24] , [26] , [44] , [57] . LANA ChIP-seq revealed at least 17 distinct LANA peaks in both BCBL-1 and TIVE-LTCs and observed occupancy patterns are almost identical between both cell types ( Fig . 5 ) . However , two LANA peaks ( marked by asterisks in Fig . 5 ) were clearly reduced in TIVE-LTC; one at nt position 11 , 636 within ORF 8 ( Table 1 peak number 7 ) and one within the coding region of K14 at nt position 128 , 517 ( Table 1 peak number 5 ) . The two highest LANA peaks are within the TRs as expected . In agreement with in vitro data [20] , LBS1 and LBS2 are strongly bound by LANA ( Fig . S3 ) . Consistent with a previous report , LANA also binds a region at the beginning of the TR , which is likely indirect since LANA failed to bind this site in vitro [58] . We observed strong LANA binding within the LANA promoter region; however , this LANA peak ( nt 127 , 391 to nt 127 , 833 ) is located just downstream of the LANAp TSS at nts position 128 , 029 ( Fig . S2B ) . Interestingly , the in vitro-characterized Sp1-containing LANA binding site ( nts 128 , 051–128 , 072 upstream of the LANAp TSS ) [26] , [59] , was not bound by LANA in vivo . Furthermore , this LANA peak completely overlaps with three CTCF binding sites ( nt 127 514–127 693 ) [36] , [60] , [61] , suggesting co-occupancy of LANA and CTCF , which was also observed at many host cellular promoters ( discussed below ) . Three additional LANA peaks were detected within the latency-associated region located within the miRNA cluster , the ORF71 coding region , and the K14 ORF ( Table 1 peak 6 , 10 , and 5 respectively ) . Several LANA peaks outside of the TRs and the latency-associated region ( peak # 8 , 9 , 13 , and 15 ) were located within a region previously reported to be bound by LANA in vitro [19] . The fact that none of these sequences showed sequence homology to LBS1/2 indicates that LANA binds either indirectly through protein-protein interactions or directly to sites with novel sequence-specificity . LANA negatively regulates RTA expression and it was demonstrated that the RBP-Jk sites within the RTA promoter are critical for LANA-dependent regulation [62] , [63] . LANA binding was observed within the ORF50 ( RTA ) region; however , this LANA peak was not upstream of the TSS close to the RBP-Jk sites but instead 600 bp downstream within the ORF50 intron ( Figure S3 ) . Rosetto et al . reported LANA binding to oriLyt and modulation of viral lytic replication using in vitro replication assays [64] . However , LANA ChIP-seq did not reveal any LANA binding to oriLyt-L or oriLyt-R in BCBL-1 cells , which display a base level of spontaneous lytic replication . These data demonstrate that LANA binding to chromatin within the LANA and RTA promoters significantly differs from in vitro EMSA assays . Unexpectedly , numerous LANA peaks are located within promoters of IE , E , and late genes , including ORF16 , ORF33 , ORF39 , ORF48 , ORF58 , ORF64 , and vIRF-1 and -3 ( Table 1 ) . To determine whether LANA potentially contributes to their regulation , fragments 2 Kb upstream of their TSSs were inserted into luciferase reporter vectors and co-transfected with a LANA expression vector into HEK293 cells . As shown in Fig . 6 , LANA transactivates the promoters of ORF16 ( E ) , vIRF1 ( E ) , ORF39 ( L ) , and ORF48 ( IE ) in a dose-dependent manner , suggesting that LANA may contribute to lytic gene expression . Interestingly , in this context Wilson et al . identified a second LANA promoter ( LANALTI ) , which is RTA-responsive and induced during lytic replication [45]; however to date no functional role for LANA during lytic replication has been established . Hence , our transactivation data and observed LANA ChIP-seq profiles suggest that LANA binding during latency potentially affects viral genes of different kinetic classes . After viral mapping , unmapped reads from LANA ChIP-seq were mapped to the human genome hg19 , and 2180 and 2951 unique peaks were identified in BCBL-1 and TIVE-LTC , respectively . In agreement with immunofluorescence data on LANA binding to mitotic chromosomes [17] , [65] , we additionally observed a large number of reads that aligned to highly repetitive GC-rich centromere regions . To focus on potential transcriptional targets , we identified LANA peaks within +/−2 kb relative to known TSS , which revealed a strong enrichment for LANA peaks around +/−500 bp in both cell types ( Fig . 7 ) . We identified 412 and 998 peaks located at promoter/enhancer regions upstream of 1295 ( BCBL-1 ) and 3917 ( LTC-TIVE ) annotated transcripts , representing 167 and 505 identified gene symbols ( Table 2 , Fig . S5 ) . Hence , LANA was detected at many more promoters in TIVE cells compared to BCBL-1 . A list of all gene loci enriched by LANA ChIP-Seq in BCBL-1 cells and TIVE-LTC cells is provided in the supplement ( Tables S2 and S3 ) . While the observed LANA binding profile on the KSHV genome was nearly identical in lymphoid and endothelial cells , LANA binding to host genes is mostly cell type specific . Only 26 genes were commonly enriched between BCBL-1 cells and TIVE-LTC cells . While PARL , NIPAL2 , IQGAP3 are BCBL-1 specific , MRPL53 , NFYC , CCDC90B , and HIST2HBE were TIVE-LTC specific , while WDR74 showed nearly identical binding profiles in both cell types ( Fig . S4 ) . Two genes , BIRC6 ( Survivin ) and Id-1 , previously reported to be regulated by LANA [66] , [67] , also contained LANA peaks within promoters in both cell types . Gene Ontology ( GO ) analysis was performed by using DAVID ( Table S4 and S5 ) . Interestingly , albeit the low overlap between both cell types , the two most enriched GO terms for both gene lists were chromosome organization and regulation of apoptosis . LANA binding was observed within promoters of several histone gene variants , explaining the association with chromosome organization , although coverage was stronger in TIVE-LTC . In BCBL-1 cells , putative LANA targets are related to phosphorus metabolic processes , regulation of cellular enzymatic activity , and regulation of cellular response to stress; while in TIVE-LTC cells , putative LANA targets are involved in regulation of macromolecule metabolic process , nutrient levels , and angiogenesis , the latter a hallmark of KS . Recently , Lu et al . performed LANA ChIP-seq in BCBL-1 cells and reported 256 enriched genes [58] and comparison of both data sets gained 15 genes in common including FBXO4 , PARL , and IQGAP3 . For functional validation , we chose IQGAP3 ( IQ motif containing GTPase-activating protein 3 ) , a regulator of cell proliferation in the Ras/ERK signaling pathway [68] . IQGAP3 was the third highest coverage LANA-binding peak in BCBL-1 cells ( Table 3 ) , and observed peaks upstream of the TSS contain two sites with homology to LANA binding sites . A proximal binding site ( BSpro ) is located at −90 and a distal ( BSdis ) is at −700 from the TSS , and both have 4 nts difference compared to the high affinity LBS1 site within the TR ( Fig . 8B ) . A 3 Kb ( −2916 to +84 ) promoter region of IQGAP3 was cloned upstream of a luciferase reporter and co-transfected with a LANA expression vector into HEK293 cells . As shown in Figure 8A , LANA transactivates the IQGAP3 promoter in a dose-dependent manner . Next , putative LANA binding sites were tested in EMSA assays using the C-terminal DNA binding domain of LANA ( LANA-C ) . Mobility of both BSpro- and BSdis-containing probes was retarded in the presence of V5-tagged LANA-C ( Fig . 8B lanes 5 and 8 ) . Adding V5 mAb resulted in supershifting of the complexes of LANA-C with BSdis and with BSpro ( Fig . 8B , lanes 6 and 9 ) . Although the intensity of the LANA-C complexes with BSpro and BSdis were less than with LBS1 , the complexes were stronger than seen for the low affinity LBS2 site ( Fig . 8B ) . No complex was seen when LANA-C was incubated with a control DNA probe derived from a 38 bp portion of IQGAP3 lacking LBS-like sequences , even with prolonged gel exposure . The IQGAP3 sequence tested by Lu et al . did not contain BSpro or BSdis , and did not compete with an LBS1/2 probe for binding to LANA [58] . To test LANA regulation of IQGAP3 in cells , we determined IQGAP3 transcript levels in BCBL-1 and LANA-inducible BJAB cells . IQGAP3 mRNA levels are about 2-fold higher in BCBL-1 cells and induction of LANA in BJAB-Tet on-LANA cells moderately induced IQGAP3 transcription ( Fig . 8C ) . Together these data demonstrate that LANA can directly bind and positively regulate the IQGAP3 promoter . Although demonstrated on a single promoter , these data further validate potential LANA targets identified by ChIP-seq , and suggest that LANA contributes to the regulation of a subset of these genes . LANA peaks were screened for LBS1/2 consensus sites allowing up to four mismatches . In BCBL-1 , 58 out of 2180 ( 2 . 7% ) and in TIVE-LTC cells 205 out of 2951 ( 6 . 9% ) peaks contained sequence similarity to LBS1/2 ( Fig . S5 ) . Hence , some enhancers/promoters may be bound directly by LANA while the majority of LANA peaks result either from protein/protein interaction or from binding unidentified sequence-specificities . To identify potentially novel LANA binding motifs , all DNA sequences enriched by LANA in BCBL-1 and TIVE-LTC cells were analyzed for consensus sequences using “peak motifs” from Regulatory Sequence Analysis Tools ( RSAT ) [69] , [70] . In BCBL-1 cells , 12 , 814 sites contained a unique 14-nts long motif ( Fig . 9A ) . Coverage of three additional motifs was significantly lower ( <3500 ) but contained a similar core sequence . Significantly , in TIVE-LTC cells 20 , 130 sites contained a motif very similar ( 13/14 ) to the one observed in BCBL-1 ( Fig . 9B ) . These results suggest that LANA either directly binds to the motif or associates with other proteins bound to the motif . We searched the known transcription factor binding sites with this motif in the JASPAR database , but failed to identify any known transcription factor with this motif . A single consensus motif without homology to known transcription factor binding sites derived from ChIP-data from two different cell types may point to a novel LANA binding specificity or alternatively , a non-characterized LANA/DNA binding protein interaction . The 15 base sequence ( TCCAT ) 3 formed from overlapping the motifs in Fig . 9AB was tested for binding by LANA-C using EMSA ( Fig . 9C ) . The DNA probe containing the ( TCCAT ) 3 motif formed a complex with LANA-C that was visible on longer exposures comparable to those used to detect binding of LANA-C to LBS2 . The complex was supershifted in the presence of V5 mAb , confirming specificity . This demonstrates that the 15 nt ( TCCAT ) 3 sequence ( Fig . 9AB ) present in high copy numbers in the human genome is a novel LANA binding motif whose affinity is comparable to LBS2 . Next , we asked whether LANA-bound promoters were enriched for specific host transcription factors . Genome-wide occupancy data for many transcription factors are available through the Encyclopedia of DNA Element ( ENCODE ) Consortium . As the closest available cell line to BCBL-1 , we mined ChIP-seq data from GM12878 cells , a B cell lymphoma cell line , for which 43 genome-wide transcription factor binding profiles are available [71] . Since ENCODE contains only six datasets for endothelial cells , this analysis was not performed in TIVE-LTC . ENCODE GM12878 ChIP-Seq data were mapped to the hg19 promoter regions , and compared to LANA occupancy observed in BCBL-1 cells . Transcription factor and LANA peaks within 2 kbp from the TSS were analyzed; furthermore , we calculated and plotted the distance distribution of these peaks . Table 4 lists the number of individual transcription factors peaks co-present at putative LANA regulated promoters , and tabulates the percentage of genes where co-occupancy is predicted . Interestingly , between 83% and 88% of the 167 LANA-binding promoters identified in BCBL-1 contain ZNF143 , CTCF , Whip , STAT1 , or ebf1 binding peaks ( Table 4 ) . Both LANA and CTCF , which contribute to latent and lytic gene expression , co-occupy the LANA and RTA genes within the viral genomes [36] , [60] , [61] . Distance analysis showed that 45% of all LANA peaks are within 200 bps and 60% are within 400 bp of CTCF sites , which expands a role for co-regulation of LANA and CTCF to host genes ( Fig . 10B ) . Similarly , 65% of LANA peaks were within 400 bps of STAT1 binding sites ( Fig . 10D ) . Hence , LANA binding may modulate promoters regulated by STAT1 , a master transcriptional regulator of immunity , cell cycle , and proliferation [72] . While many of the LANA peaks are within 200 bps of STAT1 binding sites , we did not observe LANA binding overlapping STAT1 sites , as was previously reported [58] . Co-occupancy of LANA with transcriptional regulators ZNF143 , a strong regulator of cell cycle control and proliferation ( Fig . 10A ) , whip , a transcription factor involved in DNA damage ( Fig . 10C ) , and ebf1 , a B cell-specific transcription factor ( Table 4 ) , suggest that LANA binding may affect multiple complex regulatory pathways in latently infected cells . The fact that other transcription factors like the ubiquitously expressed zinc finger protein Ying Yang 1 ( YY1 ) were not enriched suggests specificity , which is further supported by the cell type specificity of the observed LANA-bound promoters . Comparison of H3K4me3 and H3K27me3 marks to LANA peaks on viral episomes revealed regions that were enriched for LANA peaks lacked H3K27me3 , but in most cases contained H3K4me3 peaks . On the viral genome H3K4me3 is enriched at all observed LANA peaks ( Fig . 11A ) ; however , there are five regions where H3K4me3 peaks are present in the absence of LANA: 1 ) a wide peak from nt 16 K to nt 18 K spanning ORF11 , K2 and ORF2; 2 ) two sharp peaks up-stream of K4 . 2; 3 ) a minor peak at nt 103 K within the coding region of ORF63; 4 ) a major peak downstream of the miRNA cluster ( nt 119 , 500 ) ; and 5 ) within the K15 gene . Additionally , these regions are void of Pol II ( Figs . 1 and 4 ) , which is characteristic of enhancers that can be enriched for H3K4 methylation but lack Pol II [1] , [2] . These patterns suggest that LANA predominantly binds to active promoters . We extended this analysis to the human genome hg19 and determined that in both BCBL-1 and TIVE-LTC cells , a strong relationship exists between LANA binding at promoters and H3K4me3 ( R2 = 0 . 9 ) ( Fig . 11 BC ) . This suggests that LANA predominantly plays a role as a positive regulator of gene expression . To further address this , we analyzed expression patterns of the 167 genes that showed promoter-associated LANA peaks , by analyzing previously published profiling data [73] , [74] and found that 86% of these genes are indeed expressed in BCBL-1 cells . In summary , these data demonstrate that LANA preferentially associates with promoters that carry H3K4me3 marks and are transcriptionally active . The strong correlation of LANA and H3K4me3 peaks ( Fig . 11 ) raised the possibility that LANA may play a role in the methylation of histone H3K4 , which may contribute to the establishment and maintenance of latency by preventing H3K27me3-dependent silencing of latency-associated promoters . To address this question , we performed immunoprecipitation assays in BCBL-1 cells to determine whether LANA associates with histone H3K4 methyltransferases . In mammalian cells a number of different H3K4 lysine methyltransferases ( KMTs ) exist that function mostly non-redundantly . The MLL/SET1 family , including MLL1-5 and hSET1 , are the major methyltransferases . Members in this family are multi-subunit complexes that share a core complex composed of three proteins: RbBP5 , ASH2L , and WDR5 [3] , [4] , [5] . Accordingly , BCBL-1 cell lysates were immunoprecipitated with LANA-specific monoclonal antibody , and precipitated protein complexes were assayed for the presence of the endogenous MLL1-5/SET1 family core proteins by Western blotting . As shown in Figure 12A , LANA co-precipitated with RbBP5 and ASH2L , the core components of MLL/hSET1 family KMTs . We detected LANA interaction with hSET1 , but not MLL1 , in BCBL-1 cells , which express low levels of MLL1 . The association with both the hSET1 complex core proteins RbBP5 and ASH2L and hSET1 itself was further confirmed in a second PEL cell line ( BC-3 ) and in latently infected endothelial cells ( TIVE-LTC ) ( Fig . 12BC ) . To ask whether LANA directly interacts with the hSET1 core proteins we performed GST pull-down assays with purified GST-ASH2L , GST-WDR5 , or GST-RbBP5 and full-length in vitro translated LANA , but did not detect direct interaction with these proteins ( data not shown ) . In summary , these data show that LANA interacts with endogenous hSET1 complexes either directly or through protein-protein interaction . Recruitment of hSET1 complexes to specific chromatin loci has been reported to be mediated by a number of transcription factors or co-activators , including E2F , NF-E2 , and USF1 [3] , [4] , [5] . The HSV-1 VP16 protein recruits H3K4me3 KMTs to immediate early promoters after de novo infection by interacting with HCF-1 , which subsequently binds to and recruits hSET1 [12] , [13] .
As with two previous studies , we found that in BCBL-1 latency-associated genes are enriched with H3K4me3 and PolII but depleted of H3K27me3 ( reviewed in [55] ) . A number of lytic genes including ORF9 , K4 . 2 , K7 , K8 , and ORF58 also contain H3K4me3 marks . Strong Pol II peaks and H3K4me3 were detected next to the OriLytL and at the TSS of lytic genes K7 , and within the K5 and K6 ORFs . These data are in congruence with Toth et al . , which demonstrated that Pol II transcription of these genes is stalled by association with cellular negative elongation factor NELF [51] . The H3K4me3 deposition pattern was largely identical in TIVE-LTC , which provided the first genome-wide epigenetic analysis of latently-infected endothelial cells . Interestingly , the enrichment of Pol II at oriLyt region was undetectable in TIVE-LTC cells which cannot be efficiently reactivated in culture [53] ( Fig . 4 ) , suggesting that bivalent marks on promoters other than RTA contribute to efficient reactivation . Previously , Chandriani and Ganem performed transcript profiling by limiting dilution PCR during latency in BCBL-1 , SLK . 219 , and HFF . 219 cells and identified vIL-6 expression during latency [50] . We identified both PolII and H3K4me3 on the vIL6 promoter ( K2 ) , in both BCBL-1 and TIVE-LTC cells . The latter cell line is strictly latent and therefore provides evidence for vIL6 transcription during latency in endothelial cells . These data are further supported by recent studies on chromatin structure , which identified a nucleosome-free region at the vIL6 promoter in latently infected BCBL-1 and LTC-TIVE cells [49] , [76] . LANA ChIP-seq from both cell lines identified 17 highly reproducible LANA peaks on the viral genome ( Table 1 and Figure 5 ) . LANA binding patterns are very similar between cells of lymphoid and endothelial origin ( Figure 5 ) . The highest RPKM coverage was seen on the TR where LANA bound at two LANA binding sites ( LBS1/2 ) [20] . In addition to the previously described LANA and RTA signals , we detected LANA binding upstream of IE , E , and late genes ( Table 1 ) . We showed that at least in the context of reporter assays the promoters of ORF16 , 39 , 48 , and vIRF1 all responded to LANA in trans . These data are congruent with previous data showing that LANA can augment transcription from a wide range of promoters [24] , [57] , [59] . A potential role for LANA during reactivation would explain the presence of a second RTA-responsive promoter upstream of the LANA ORF [45] . However , such a role would have to be early since Kim et al . demonstrated that LANA association with viral episomes decreased at about 4 hours post reactivation [22] . Experiments interrogating whether LANA modulates lytic gene expression in the context of reactivation are currently ongoing . Lu et al . reported LANA binding on viral and human genomes in BCBL-1 cells [58] . While both studies utilized an identical antibody and cell line ( BCBL-1 cells ) , Lu reported only two low coverage peaks at the K7 and vIRF2 regions outside of the TR , which were detected in our study; differences with respect to coverage of the viral genome may largely result from usage of different sequencing methods . LANA binds close to TSS of H3K4me3 decorated promoters , and often co-occupies with transcription factors and the boundary element CTCF ( Fig . 7 , 10 and Table 4 ) . LANA binding to host genes is cell type-specific with more promoters bound in endothelial cells . A small number of LANA enriched sequence tags showed sequence homology to the consensus LANA binding sites ( 58 of 2180 in BCBL-1 and 205 of 2951 in TIVE-LTC ) ( Table 1 ) . RSAT motif analysis [70] , [77] revealed a common consensus sequence ( CCATTCCATTCCA ) that is highly prevalent in the human genome and was highly enriched in both cell types ( Fig . 9 AB ) . EMSA and supershift analysis demonstrated direct LANA binding to this novel motif ( Fig . 9C ) . The affinity of LANA-C for the ( TCCAT ) 3 motif was lower than LBS1 but comparable to LBS2 , the low affinity site within the TR [20] . The existence of thousands of copies of this motif , many of which are within repeats ( data not shown ) , suggests that this binding may be biologically significant . In fact , episomal tethering to the chromosome via binding between the C-terminus of LANA and this novel motif is consistent with reports indicating that the C-terminal DNA binding domain of LANA contributes to chromosome binding [78] , [79] , [80] . In summary , the binding of LANA-C to LBS1/2-like sequences and the novel ( TCCAT ) 3 consensus sequence demonstrates that LANA can directly bind to host cellular DNA via two distinct sequence motifs . This interpretation is also supported by cluster analysis using seqMiner algorithms , which revealed a number of distinct binding patterns across all LANA-enriched promoters ( Fig . 11B ) . Mercier et al . [81] recently reported a LANA ChIP-seq analysis performed in PEL cells and in lymphatic endothelial cells ( LEC ) , which were previously shown to display a unique gene expression profile that is markedly different from latency [82] . Several findings agree between both studies , specifically the fact that many more promoters showed LANA binding in endothelial cells versus lymphoid cells and that these were largely cell type specific . Moreover , the findings by both groups that LANA binds close to TSS that are decorated with H3K4me3 and actively transcribed are in agreement . In addition , Mercier et al . performed RNA-Seq experiments in uninfected and KSHV-infected LEC cells and showed that only a small number of host genes bound by LANA were differentially expressed [81] . This is consistent with our observation that numerous viral genes and 14% of cellular promoters that are bound by LANA are not expressed during latency in BCBL-1 cells ( Fig . 1 , 4 ) . Hence , LANA binding alone does not induce transcription in the context of chromatin . LANA may act at the epigenetic level by influencing histone modifications . Alternatively , LANA may merely have a higher propensity to bind to H3K4me3 decorated promoters since they are often transcribed and contain open chromatin . However , as demonstrated here for the IQGAP3 gene ( Fig . 8 ) , and by numerous previous studies , host cellular genes can be directly regulated by LANA [66] , [67] , [74] . With respect to analyzing how LANA binds to DNA , our studies differ in the proportion of LANA peaks that contain LBS1/2-like sequences . Mercier identified a sequence nearly identical to LBS1 in 157/267 ( 58 . 8% ) of the LANA ChIP peaks , but we found 58/2180 ( 2 . 7% ) of peaks from BCBL-1 cells contained sequences resembling LBS1 , and 205/2951 ( 6 . 9% ) from TIVE-LTC cells . These differences may result from using different bioinformatics tools in the two studies . Additionally , we identified and biochemically characterized a novel LANA-binding sequence motif ( TCCAT ) 3 , which occurs with high frequency in the human genome . Motivated by LANA's preferential association with H3K4me3 mark-containing promoters , we asked whether LANA interacts with KMTs and demonstrated that LANA efficiently immunoprecipitates with hSET1 , the main H3K4 methylase in mammalian cells [3] , [4] ( Fig . 12 ) . We did not detect direct binding to the hSET1 core components RbBP5 , WDR5 , and ASH2L [5] , [6] , [7] and do not know whether LANA interacts with any of the remaining hSET1 components or through a bridging factor . A detailed biochemical and genetic study to determine how LANA interacts with and potentially modulates H3K4me3 deposition is currently ongoing . In support of this observation , hSET1 is not the only epigenetic modifier complex shown to interact with LANA . Kim et al . recently demonstrated that LANA association with the histone demethylase KDM3A regulates viral gene expression during both latent and lytic replication [22] . While the LANA-hSET1 interaction is novel for γ-herpesviruses , the HSV-1 VP16 protein is known to recruit hSET1 and MLL complexes to immediate early promoters through an interaction with HCF-1 [12] , [13] . We propose a working model for the establishment of the viral epigenome which integrates recent findings affecting i ) the epigenetic variation of KSHV episomes [49] , [76] , and ii ) novel mechanistic insights into how PRC2 deposits H3K27me3 marks [83] . After de novo infection an early burst of promiscuous transcription which includes both LANA and the RTA gene occurs which leads to co-transcriptional H3K4me3 deposition at many promoters [1] , [75] , [84] . We envision that LANA is recruited to many promoters that are initially active through an hSET1-dependent mechanism . As a result PRC2-dependent silencing is stopped at regions where LANA is bound and H3K4me3 has been deposited . Recent chromatin structure mapping analysis on the LANA , RTA , and vIL6 promoters demonstrated that a subpopulation of episomes in PEL cells carry nucleosome free regions ( NFRs ) [49] , [76] , which recently have been shown to prevent H3K27me3 marks from spreading [83] . Moreover , these NFRs are flanked by CTCF boundary elements as in the LANA promoter [36] , [37] , [60] , [61] , [85]; additionally LANA binding sites were highly correlated with CTCF binding on both the viral and host chromatin ( Table 4 ) . We envisage competition between PRC2 silencing [34] , [35] and LANA recruitment of KDM3A [22] and possibly hSET1 to form euchromatin on a number of latent promoters as well as promoters essential for reactivation . As a result a small number of episomes will carry epigenetic marks that are “permissive for latency” . Conversely , another subpopulation of episomes will be completely silenced by host-dependent heterochromatin formation and as a result will neither contribute to latent nor lytic gene expression , as recently suggested in BCBL-1 and TIVE-LTC cells by single copy chromatin mapping [49] . Further understanding the precise molecular mechanisms by which LANA contributes to maintenance of euchromatin may yield approaches to tip the balance towards complete epigenetic silencing as a novel intervention strategy .
293 cells , the human embryonic kidney cells , and KSHV long term-infected telomerase-immortalized human umbilical vein endothelial cells ( TIVE-LTC ) [53] were cultured in Dulbecco's modified eagle medium ( DMEM ) supplemented with 10% fetal calf serum ( FCS ) and antibiotics at 37°C under 5% CO2 atmosphere . BCBL-1 and BJAB cells were cultured in RPMI 1640 medium supplemented with 10% FCS and antibiotics at 37°C under 5% CO2 atmosphere . Primer pairs were designed to amplify the promoter region ( +/−2 kb relative to the transcription start site ) for IQGAP3 ( 5′-GATCGGTACCACAACCCAGTCTCTAAACCAG-3′ and 5′-GACTACGCGTGTTCCTTAGGCTGCCCC ) , ORF16 ( 5′- CCGCTCGAGCGGTTGTCAACCAACCAGTCAATCACC and 5′- CCCAAGCTTGGGGCAAAACGTCCTCGTCCATT ) , ORF39 ( 5′- GGGGTACCCCGAATGATGTTTGTCTTCGCC and 5′- CCGCTCGAGCGGCGCATGTTTCTCGGTCTTTT ) , ORF48 ( 5′- GGGGTACCCCTCTACCATGGAAGCCGGCAA and 5′- CCGCTCGAGCGGGGATACACACCTCCATGTTC ) , and vIRF1 ( 5′- GGGGTACCCCATGGAAGCCGGCAACAGTCCT and 5′- CCGCTCGAGCGGACACCTCCATGTTCAGTCAC ) . Corresponding promoter regions were PCR amplified , and cloned into plasmid pGL3-basic at appropriate sites . Antibodies used are control rabbit IgG ( Santa Cruz Biotechnology , sc-2027 ) , rat IgG ( Santa Cruz Biotechnology , sc-2026 ) , rabbit against H3K4me3 ( Abcam , ab8580 ) , H3K27me3 ( Upstate , 07-449 ) , rabbit anti Pol II ( Santa Cruz Biotechnology , sc-899 ) , rat LANA monoclonal antibody ( ABI Inc . , LN53 ) , rabbit anti hSET1 ( Bethyl , A300-289A ) , rabbit anti RbBP5 ( Bethyl , A300-109A ) , rabbit anti ASH2L ( Bethyl , A300-489 ) , and rabbit anti MLL1 ( Upstate , ABE240 ) . ChIP experiments were performed as described before with minor modifications [32] . BCBL-1 or TIVE-LTC cells were crosslinked with 1% formaldehyde at room temperature for 10 min . Crosslinking was terminated by adding glycine to a final concentration of 0 . 125M . Cells were washed twice in ice-cold PBS with protease inhibitors and harvested by centrifugation . Every 2×107 cells were lysed in 1 ml ice-cold Farnham lysis buffer ( 5 mM PIPES [pH 8 . 0] , 85 mM KCl , 0 . 5% NP-40 ) with protease inhibitors . The nuclei were spun down and resuspended in 1 ml RIPA buffer ( 1% NP-40 , 0 . 5% sodium deoxycholate , and 0 . 1% SDS in 1×PBS ) with protease inhibitors . Chromatin was sheared to about 250 bp fragments with 5 sets of 30-second pulses using a Sonic Dismembrator ( Fisher Scientific ) set to 50% of maximum power . Chromatin from 6×107 cells was incubated with 10 µg primary antibody ( normal rabbit IgG , rabbit anti H3K4me3 , rabbit anti H3K27me3 , normal rat IgG , or rat anti LANA ) and 100 µl magnetic beads ( sheep anti rabbit-conjugated , or protein A-conjugated ) at 4°C overnight on a rotator . Beads were washed 5 times with LiCl wash buffer and once with TE . The immune complexes were eluted with 200 µl elution buffer twice at 65°C for 1 hour . The combined eluates were de-crosslinked at 65°C overnight . DNA was extracted once with phenol/chloroform and precipitated with ethanol . 20 µg glycogen was added as DNA carrier . DNA pellets were washed once with 70% ethanol and resuspended in 40 µl H2O . 100 ng ChIP-enriched DNA or control IgG ChIP DNA was blunt-ended with T4 DNA polymerase and Klenow DNA polymerase ( NEB ) and phosphorylated with T4 PNK ( NEB ) . Addition of an “A” base to the 3′ ends of the blunt phosphorylated DNA fragments was performed in the presence of Klenow exo- ( NEB ) . Subsequently , adapters ( 5′-pGATCGGAAGAGCGGTTCAGCAGGAATGCCGAG and 5′- ACACTCTTTCCCTACACGACGCTCTTCCGATCT ) were ligated to both ends of the DNA fragments . DNA within the range of 150 bp to 300 bp was gel-purified and PCR-amplified for 18 cycles with primer 5′ -AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT and 5′- CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT . The resulting libraries were gel-purified , quantified with QuantIT dsDNA Assay Kit ( Invitrogen ) and sequenced with Genome Analyzer IIx ( Illumina ) in Department of Molecular Genetics and Microbiology ( MGM ) at University of Florida ( UF ) . The biotin-labeled RNA baits specific for KSHV genome was customized with eArray XD ( Agilent ) with help from Agilent . The RNA baits are 120 nt long with 4× tiling frequency . TIVE-LTC ChIP-seq libraries were constructed as above . After adapter ligation , DNA fragments between 150 bp and 300 bp were gel-selected and amplified with 10 cycles in first PCR . Samples were purified with Agencourt AMPure XP beads ( Beckman Coulter ) . The KSHV-specific DNA sequences were enriched with SureSelect Enrichment System ( Agilent ) according to the manufacturer's instruction . The purified 1st PCR products were denatured and hybridized with KSHV RNA baits at 65°C for 48 hours in PCR machine with heated lid . The RNA-DNA hybrids were recovered with Dynal MyOne Streptavidin T1 magnetic beads ( Invitrogen ) . The captured DNA was eluted and purified . The DNA was re-amplified for 22 PCR cycles with primers ( 5′-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT and 5′-CAAGCAGAAGACGGCATACGAGCTCTTCCGATCT ) , and purified using Agencourt AMPure XP beads . The libraries were quantified with QuantIT dsDNA Assay Kit ( Invitrogen ) and sequenced as above . HTS sequencing generated 20 . 8 million , 17 . 1 million , and 20 . 5 million tags for H3K4me3 , H3K27me3 , and Pol II ChIP-seq , respectively . After enrichment 42 . 76% , 65 . 4% , and 87 . 58% of H3K4me3 , H3K27me3 , and Pol II ChIP-seq tags were mapped to the viral genome compared to 0 . 014% , 0 . 028% , and 0 . 007% without enrichment; hence enrichment efficiency was about 2 , 000- to 4 , 000-fold .
|
KSHV is a DNA tumor virus which is associated with Kaposi's sarcoma and some lymphoproliferative diseases . During latent infection , the viral genome persists as circular extrachromosomal DNA in the nucleus and expresses a very limited number of viral proteins , including LANA , a multi-functional protein . KSHV viral episomes , like host genomic DNA , are subject to chromatin formation and histone modifications which contribute to tightly controlled gene expression during latency . We determined where LANA binds on the KSHV and human genomes , and mapped activating and repressing histone marks and RNA polymerase II binding . We found that LANA bound near transcription start sites , and binding correlated with the transcription active mark H3K4me3 , but not silencing mark H3K27me3 . Binding sites for transcription factors including znf143 , CTCF , and Stat1 are enriched at regions where LANA is bound . We identified some LANA binding sites near human gene promoters that resembled KSHV sequences known to bind LANA . We also found a novel motif that occurs frequently in the human genome and that binds LANA directly despite being different from known LANA-binding sequences . Furthermore , we demonstrate that LANA associates with the H3K4 methyltransferase hSET1 which creates activating histone marks .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"functional",
"genomics",
"cell",
"biology",
"chromosome",
"biology",
"viral",
"persistence",
"and",
"latency",
"virology",
"viruses",
"and",
"cancer",
"genetics",
"biology",
"and",
"life",
"sciences",
"microbiology",
"genomics"
] |
2014
|
LANA Binds to Multiple Active Viral and Cellular Promoters and Associates with the H3K4Methyltransferase hSET1 Complex
|
Chemokines have been shown to be effective bactericidal molecules against a variety of bacteria and fungi in vitro . These direct antimicrobial effects are independent of their chemotactic activities involving immunological receptors . However , the direct biological role that these proteins may play in host defense , particularly against intestinal pathogens , is poorly understood . Here , we show that CXCL9 , an ELR- chemokine , exhibits direct antimicrobial activity against Citrobacter rodentium , an attaching/effacing pathogen that infects the gut mucosa . Inhibition of this antimicrobial activity in vivo using anti-CXCL9 antibodies increases host susceptibility to C . rodentium infection with pronounced bacterial penetration into crypts , increased bacterial load , and worsened tissue pathology . Using Rag1-/- mice and CXCR3-/- mice , we demonstrate that the role for CXCL9 in protecting the gut mucosa is independent of an adaptive response or its immunological receptor , CXCR3 . Finally , we provide evidence that phagocytes function in tandem with NK cells for robust CXCL9 responses to C . rodentium . These findings identify a novel role for the immune cell-derived CXCL9 chemokine in directing a protective antimicrobial response in the intestinal mucosa .
The intestinal tract is a site of continuous interaction between host and microbe . Tight regulation of immune surveillance and activation maintains the integrity of this interface during non-infectious periods while preserving the ability to launch immediate action upon pathogen exposure . Chemokines are a vital component of this protective response , linking innate and adaptive immunity by activating and recruiting immune cells to infection sites [1] . Until recently , the function of these molecules was focused on their chemotactic activity induced upon interaction with their receptors on various immune cells [1] . However , mounting evidence has shown a direct antimicrobial function for a number of chemokines that relates to their cationic surface properties , similar to antimicrobial host defense peptides [2 , 3] . Host defense peptides ( or antimicrobial peptides ) , are produced by a wide variety of cell types and form an important component of innate host defenses [4 , 5] . Although the exact bactericidal mechanism for cationic antimicrobial peptides remains debated [6] , membrane-disrupting activity appears to be a common feature , facilitated by cationic charge distribution and amphipathicity allowing for attachment to and insertion into bacterial membranes . In mammals , antimicrobial peptides help protect the gut mucosae from infection and maintain intestinal homeostasis [7] . In vitro , the chemokine CXCL9 has potent bactericidal activity against Escherichia coli , Listeria monocytogenes , and Bacillus anthracis [3 , 8] , likely related to a cationic C-terminal domain resembling well characterized antimicrobial peptides [3 , 9] . Additionally , antimicrobial activity of CXCL9 appears to play a key role in protection of mucosal surfaces against pathogen infection [8–11] . However , outside of its role in T cell homing and activation , a potential role for CXCL9 in conferring antimicrobial protection of the gut mucosae against intestinal infections has not been investigated . Citrobacter rodentium is an intestinal pathogen of mice used to model infections with the attaching and effacing ( A/E ) human pathogens , enteropathogenic E . coli and enterohemorrhagic E . coli [12] . C . rodentium colonizes the cecum prior to traversing to the primary site of infection in the distal colon . Like other A/E pathogens , Citrobacter forms intimate attachments with the epithelial surface in the cecum and colon by using a type III secretion system and effectors that it injects into the host’s intestinal epithelial cells [13 , 14] . The host response to infection includes an early and robust chemokine response [15–17] . Infection clearance in resistant mice requires Th1 adaptive responses [13 , 18 , 19] mediated by the IFNγ-stimulated chemokines CXCL9 , CXCL10 , and CXCL11 acting in a CXCR3 dependent manner [20] . Interestingly , of all the ELR- chemokines , CXCL9 is the most highly expressed during C . rodentium infection [15] , suggesting that its role in protecting the host from this intestinal pathogen might be dominant . Previous work by our laboratory found that depletion of IFNγ-producing NK1 . 1+ CD3- cells reduced CXCL9 levels in the C . rodentium-infected gut , leading to increased host susceptibility to infection [19] . Whether this was solely related to T cell recruitment or additional antimicrobial activity was not known . Here , we uncover a direct antibacterial function for CXCL9 that protects intestinal crypts from bacterial infiltration .
ELR- chemokines have an emerging role as potent antimicrobial agents , owing to their cationic C-terminal domain rich in positively charged amino acids . Since the level of CXCL9 increases dramatically following Citrobacter infection of resistant mice [15] , we first determined whether CXCL9 exerted antimicrobial activity against C . rodentium in vitro . Treatment of C . rodentium with CXCL9 resulted in a dose-dependent bacterial killing , as measured by viable colony counts of residual survivors ( Fig . 1A ) . Exposure to CXCL9 at ∼4 μg/ml ( 270 nM ) resulted in 100% killing , and 85% killing at ∼0 . 4 μg/ml ( 27 nM ) . This concentration is biologically relevant as CXCL9 levels in rectal perfusions from the inflamed human intestine can reach up to 2 μg/ml ( 138 nM ) [21] . Time-kill curves showed that killing was rapid , reaching near maximal effect at ∼5 min post-exposure to 27 nM CXCL9 ( Fig . 1B ) . Importantly , in addition to antimicrobial action against the mouse pathogen C . rodentium , we also observed similar susceptibility of human pathogens , EHEC and EPEC to CXCL9-directed bacterial killing ( Fig . 1C ) . Many host-adapted pathogens have evolved resistance to antimicrobial peptides through LPS modifications controlled by the two-component system PhoP-PhoQ [22–24] . To determine whether residual survival of C . rodentium following treatment with lower doses of CXCL9 was PhoP-PhoQ dependent , we treated wild type bacteria and a phoPQ deletion mutant with a concentration of CXCL9 that produced ∼10–20% survival of wild type bacteria , and measured viable bacteria after 1–2 h . The ΔphoPQ strain was more susceptible to CXCL9 , with only 1–2% residual survivors , similar to a peptide-sensitive strain of E . coli . A similar result was seen using the human α-defensin , HD5 , an abundant defense molecule produced by Paneth cells in the gut ( Fig . 1D ) . Interestingly , we found that a 1000-fold lower molar concentration of CXCL9 could elicit similar killing when compared with HD5 . To confirm that this killing was directly linked to CXCL9 , we performed killing assays in the presence of purified anti-CXCL9 antibody or control IgG . Whereas CXCL9 killed ∼100% of C . rodentium , this activity was completely blocked by anti-CXCL9 antibody but not an IgG control ( Fig . 1E ) . Antimicrobial peptides like polymyxin B kill bacteria by inducing membrane permeability [25] , which can be measured by a fluorescence increase following membrane incorporation of the neutral hydrophobic molecule 1-anilino-8-naphthalene-sulfonate ( ANS ) [26] . Indeed , injection of polymyxin B into a culture of C . rodentium led to an immediate increase in ANS fluorescence ( Fig . 1F ) . To better understand how CXCL9 might exert its antimicrobial activity , we measured ANS fluorescence following CXCL9 injection into C . rodentium culture . An increase in fluorescence , which occurs when ANS partitions into exposed membrane , was observed on a similar time scale and magnitude upon injection of CXCL9 as that seen with polymyxin B ( Fig . 1F ) . Together , these data established a direct antimicrobial activity for CXCL9 on C . rodentium in a manner consistent with membrane disruption , similar to classic antimicrobial peptides . Previous work investigating the role of CXCL9 in the gut has focused solely upon its chemotactic properties . However our in vitro results suggested that CXCL9 might have biological significance as a direct antimicrobial molecule . To investigate this , we first measured the levels of CXCL9 in the gut following C . rodentium infection by ELISA . CXCL9 was readily detectable in a variety of tissue samples , including fecal pellets , rectal perfusions , scrapings of the colonic mucosae , and in total homogenized colon samples ( S1A Fig . ) . Next , we depleted CXCL9 in Rag1-/- mice infected with C . rodentium and monitored host survival and bacterial load . Rag1-/- mice were used in order to study the protective effect of CXCL9 independent of its ability to recruit T cells . C . rodentium-infected mice depleted of CXCL9 died , on average , 2 days earlier than mice receiving IgG control . This was significant , and specific to CXCL9 depletion , as depletion of another antimicrobial ELR- chemokine , CXCL10 , had no effect on host mortality over 10 days ( Fig . 2A ) . This difference between CXCL9 and CXCL10 in affecting disease severity may be due to the differential expression of the two chemokines , where CXCL9 is expressed at much higher levels in the infected colon , suggesting a more dominant role [15] . In line with these data , C . rodentium burden was 10–100 times higher in the fecal output from CXCL9-depleted mice compared to controls over the 10-day infection period ( Fig . 2B ) . In order to confirm that anti-CXCL9 antibodies were reaching the lumen of the gut , we measured IgG levels in fecal samples following intraperitoneal delivery of anti-CXCL9 antibody or control IgG into uninfected Rag1-/- mice and showed an IgG accumulation in the feces ( S1B Fig . ) . As a secondary biological readout for CXCL9 depletion , we measured the number of infiltrating CD3+ cells in the distal colon 10 days after C . rodentium infection of immunocompetent C57BL/6 mice and in mice depleted of CXCL9 . As expected , in C . rodentium-infected immunocompetent C57BL/6 mice depleted of CXCL9 , there was a ∼75% reduction in the number of CD3+ cells in the distal colon ( S1C–D Fig . ) . To investigate the impact of CXCL9 depletion on disease severity , we examined the gross pathology of the gut during necropsy . C . rodentium-infected mice depleted of CXCL9 had ( i ) increased evidence of colitis in the cecum , which was shrunken and partially emptied; ( ii ) shortening and thickening of the colon; ( iii ) increased incidence of watery stool; and ( iv ) hematomas along the length of the cecum and colon ( Fig . 2C ) . As a measure of diarrhea in the stool , we measured fecal water content at day 10 , which was significantly greater in CXCL9-depleted mice compared to controls ( Fig . 2D ) . These results were independent of significant differences in the levels of TNFα , IFNγ , IL12p40 , or IL-10 in colonic explants on day 10 post-infection , which were all similar in control mice and mice depleted of CXCL9 ( S2 Fig . ) . Together , these data established an important role for CXCL9 in host defense against C . rodentium infection and in limiting C . rodentium burden in the gut . The gross pathological differences in C . rodentium-infected mice upon CXCL9 depletion suggested a worsened immunopathological response to infection , which correlated with increased bacterial load . Since the pathologic impact on the host can be affected by less than a log-change in peak C . rodentium load [27] , we scored colonic histopathology on day 10 in C . rodentium-infected control mice and animals depleted of CXCL9 . CXCL9 depletion was associated with increased numbers of sloughed epithelial cells in the lumen and immune cell infiltration , with destruction of the epithelial architecture ( Fig . 3A ) . These changes resulted in significantly greater transmural pathology in the colon ( Fig . 3B ) . Similar findings were observed for cecal pathology upon C . rodentium infection of CXCL9-depleted mice , with a more pronounced pathology in the mucosa and submucosa regions of depleted animals ( S3 Fig . ) . In a typical C . rodentium infection , the bacteria attach to the intestinal epithelium , but do not commonly penetrate deep into intestinal crypts [28 , 29] . To localize C . rodentium in infected mice in the presence or absence of CXCL9 , we performed immunohistochemical localization of C . rodentium using an antibody specific to C . rodentium LPS . Indeed , we observed the majority of C . rodentium in close association with the colonic epithelial cell surface in uncontrived Rag1-/- mice , with only marginal evidence of crypt penetration . In contrast , in mice depleted of CXCL9 , C . rodentium was commonly found to penetrate deeply into crypts in the colon ( Fig . 3C and 3D ) and cecum ( S3 Fig . ) . CXCL9 is mainly dependent on IFNγ for its expression [30] , and IFNγ-/- mice have impaired resistance and greater pathology following C . rodentium infection similar to that seen in our CXCL9 depletion studies [31] . Given our results following infection of CXCL9-depleted mice , we hypothesized that IFNγ-/- mice would be similarly susceptible to crypt penetration by C . rodentium due to the attendant decrease in CXCL9 expression . We tested this by localizing C . rodentium in colonic tissues of IFNγ-/- mice and C57BL/6 wild type controls by immunohistochemical staining . In these experiments , we found that C . rodentium was localized mainly to the epithelial surface in C57BL/6 mice , whereas bacteria were commonly found penetrating into colonic crypts of IFNγ-/- mice ( Fig . 3E and 3F ) . Together , these data indicated that the antimicrobial action of CXCL9 helps maintain epithelial barrier defenses against C . rodentium by preventing crypt penetration by invading bacteria . The loss of this defense upon CXCL9 depletion allows for bacterial penetration deep into intestinal crypts with an attendant increase in pathology . To firmly establish the biological significance of direct antimicrobial activity of CXCL9 during C . rodentium infection , we infected CXCR3-/- mice that lack the CXCL9 chemokine receptor and thus do not mount CXCR3-dependent responses following ligand interactions . Based on our prior data , we hypothesized that the host susceptibility to C . rodentium infection following CXCL9 depletion would persist in CXCR3-/- mice . Indeed , CXCR3-/- mice depleted of CXCL9 carried a significantly increased burden of tissue-associated C . rodentium in the colon ( Fig . 4A ) . Furthermore , CXCL9-depleted CXCR3-/- mice had increased pathology scores in both the distal colon ( Fig . 4B and Fig . 4C ) and in the cecum ( S4 Fig . ) . In agreement with our previous data for a direct role for CXCL9-mediated protection of intestinal crypts , uncontrived CXCR3-/- mice were able to restrict C . rodentium to the epithelial cell surface with virtually no penetration by bacteria into intestinal crypts . In contrast , depletion of CXCL9 in C . rodentium-infected CXCR3-/- mice produced a striking invasion of bacteria deep into intestinal crypts in both the colon ( Fig . 4D and 4E ) and the cecum ( S4 Fig . ) . These data confirmed that the antibacterial activity and host protection afforded by CXCL9 was independent of CXCR3-ligand-mediated effects . Previous work examining transcript levels of CXCL9 in the C . rodentium infected colon found the predominant source to be CD11c+ cells , and was therefore attributed dendritic cells ( DCs ) [15] . However , recent work into understanding the role that DCs and macrophages play in intestinal homeostasis , as well as inflammation , has revealed that the intestinal tract is more heavily populated with macrophages [32] , and that some previous work attributing function to CD11c+ DCs has instead been misidentified macrophages [33–35] . Indeed , macrophages have been shown to be a significant source of CXCL9 in other inflamed tissues [36 , 37] . We measured CXCL9 release from GM-CSF induced bone marrow-derived DC ( BMDC ) and M-CSF induced macrophages ( BMDM ) in response to heat-killed Citrobacter and IFNγ . Unstimulated BMDC and BMDM did not produce detectable levels of secreted CXCL9 . However , exposure to either IFNγ alone , or C . rodentium alone stimulated intermediate levels of CXCL9 , which was significantly boosted in response to a combination of both stimuli ( Fig . 5A ) . These data were consistent with that of others , showing that maximal CXCL9 expression is induced by IFNγ , in combination with additional microbial stimuli [38 , 39] . We examined the potential sources of IFNγ responsible for driving CXCL9 expression within DCs , and macrophages . It is well established that natural killer ( NK ) cells are an important early source of IFNγ , and previous work by our laboratory has shown that depletion of NK1 . 1+ cells , a common NK cell marker , in Citrobacter-infected mice results in decreased CXCL9 expression in the colon [19] . Unstimulated BMDM , BMDC , or NK cells did not express CXCL9 . However , co-culture of BMDM and NK cells , or DCs and NK cells , lead to CXCL9 secretion that was dependent on the presence of C . rodentium ( Fig . 5B ) . NK cells were important for maximal CXCL9 release as CXCL9 levels were reduced by ∼50% in the absence of NK cells . Interestingly , we observed that macrophages were far more capable at not only inducing their own expression of CXCL9 upon exposure to C . rodentium , independent of NK cells , but also released more CXCL9 upon co-stimulation with either IFNγ and Citrobacter , or NK cells and Citrobacter in paired experiments directly comparing macrophages and inflammatory DC populations ( Fig . 5B ) . Previous studies have shown that macrophages are capable of producing IFNγ in a TLR signaling-dependent fashion [40 , 41] . Using NK cells isolated from wild type mice , or IFNγ-/- animals and BMDM derived from these mice , we measured the contribution IFNγ produced by macrophages and NK cells on CXCL9 release following C . rodentium stimulation . We found that IFNγ expression by both NK cells and macrophages was critical for maximum CXCL9 release in the presence of heat-killed C . rodentium ( Fig . 5C and 5D ) . Equivalent CXCL9 levels were released from C . rodentium-stimulated BMDM from wild type mice when NK cells were absent , or in the presence of IFNγ-deficient NK cells . These data indicated that the release of CXCL9 was dependent upon IFNγ expression by NK cells , and not other co-stimulatory , and/or NK cell-directed alternative cytokine expression mechanisms . Macrophage-derived IFNγ was also important for this response as release of CXCL9 was significantly reduced by ∼50–60% from macrophages unable to produce their own IFNγ ( Fig . 5D ) . This was further supported in experiments that showed CXCL9 release from C . rodentium-stimulated IFNγ-/- macrophages was significantly less than CXCL9 levels from wild type macrophages ( Fig . 5C ) . Together , these data indicated that IFNγ produced by both NK cells and macrophages is necessary to achieve maximal CXCL9 expression in response to C . rodentium . Finally , we measured CXCL9 levels in the gut following C . rodentium infection and found that , similar to our in vitro results , chemokine levels in the cecum and colon were significantly blunted in IFNγ-/- mice ( Fig . 5E ) . Together , these data indicate that IFNγ produced by NK cells and macrophages is necessary to achieve maximal CXCL9 expression in response to C . rodentium .
The intestinal tract is an environment that requires a balanced set of immunological responses , capable of tolerating the host’s commensal microbiota , while remaining primed to respond to invasion by pathogenic bacteria . Early immune responses to pathogens are critical for controlling both bacterial burden , and disease pathology , which the host achieves through combined cellular and innate antimicrobial responses . Chemokines are an important facet to this host protection , by linking innate antimicrobial activity with cell homing to the site of infection . We found that CXCL9 , a chemokine known previously as an important modulator of CXCR3-dependent cellular homing following C . rodentium infections , has an important additional function in innate antimicrobial defense of the gut . This antimicrobial activity is independent of the CXCR3 receptor , or other aspects of adaptive immunity , and helps to control bacterial burden while protecting intestinal crypts from pathogen invasion . IFNγ-/- mice are more susceptible to C . rodentium infection [31] . The basis for this was thought to be the loss an IFNγ-dependent antimicrobial factor expressed in the colon . Our results are consistent with CXCL9 , a chemokine induced by IFNγ in combination with other microbial stimuli , as a likely mediator of host protection in IFNγ-competent hosts . Previous worked showed that p38α expression in T cells regulates host defense against C . rodentium infection [42] . This study revealed that T cells lacking p38α had a significant reduction in IFNγ production following C . rodentium infection , a decreased infiltration of inflammatory cells into the colon , and yet increased tissue damage , a result that could be linked to the increased invasion of C . rodentium in mice with p38α T cell deficiency . Interestingly , treatment of these mice with IFNγ restored host defenses against C . rodentium , leading to lessened tissue damage , and more importantly , normalization of the tissue-associated bacterial burden . While the authors attributed these findings to increased T cell homing to site of infection , an alternative or additional interpretation , given our current results , is the attendant increase in innate defense mediated by IFNγ-stimulated CXCL9 release . The increased tissue damage resulting from loss of IFNγ , despite the blunted immune cell infiltration [19 , 42] , is likely due to the pathogen itself gaining access to the privileged host niche within intestinal crypts . In a typical C . rodentium infection of resistant hosts , the host restricts the pathogen to the lumen or epithelial surface . However , depletion of CXCL9 , or loss of IFNγ production allows for C . rodentium to penetrate deep within the crypts . A correlation between invasion of C . rodentium into crypts and increased host pathology has been observed in other studies [43 , 44] and so protecting this niche against intestinal pathogens is a key function for the innate immune system . Of note , we found that IFNγ-/- mice were capable of producing a modest , but non-protective level of CXCL9 in the colon following C . rodentium infection . Although CXCL9 production is typically considered to be dependent on IFNγ , an alternative induction pathway has been described macrophages involving IFNα/ signaling . For example , low-level expression of CXCL9 was described in IFNγ-/- mice following infection with vaccinia virus [45] . In addition , STAT1 activation by IFNα in primed macrophages also boosts CXCL9 expression [46] . The quantitative contribution of this IFNγ-independent production of CXCL9 on host protection , however , has not yet been defined . In this work , we observed that macrophages exhibited greater capacity for CXCL9 expression in the presence of microbial stimuli , and/or IFNγ , compared to DCs . Previous observations attributed the greatest levels of CXCL9 transcript to CD11c+ DCs in the Citrobacter-infected colon [15] . However this study relied upon CD11b and CD11c markers to differentiate phagocytes , typical surface markers routinely used to identify DCs and macrophage populations . However , recently it has become clear that many markers previous attributed to a homogenous phagocyte populations , in particular CD11c and CD11b , are expressed on multiple cell types [33] . Therefore , further examination , and in particular direct cell staining for CXCL9 expression , is necessary to identify the population ( s ) essential for robust expression of this chemokine . In addition to DCs , various cell types have been found to express CXCL9 , including epithelial cells , neutrophils , and macrophages [36–38 , 47] . Recently , staining for CXCL9-producing cells within the inflamed tonsils also revealed macrophages to be the predominate source of the chemokine [37] . Preliminary data from our laboratory has shown co-localization of CXCL9 expression with the F4/80 macrophage marker within the colon of C . rodentium infected mice ( data not shown ) . Given these data , macrophages appear to be a significant source of CXCL9 within the inflamed colon of C . rodentium infected mice , however the utility of current molecular tools to investigate the cellular sources of CXCL9 in the gut appear , in our hands , to be limited . This could potentially be overcome by directly labeling native CXCL9 in transgenic mice [48] , however additional work is required . Some host-adapted bacteria have evolved mechanisms of resistance towards antimicrobial host defense peptides through enzymatic cleavage-based mechanisms [49 , 50] . Evidence for bacterial resistance to the antibacterial activity of CXCL9 has also been observed , further implicating it as an innate host defense that can be a selectable target of resistance . For instance , the streptococcal inhibitor of complement ( SIC ) protein , secreted by Streptococcus pyogenes , can inhibit the antimicrobial activity of the CXCL9 C-terminal domain [9] . SufA from Finegoldia magna can also block the antimicrobial activity of CXCL9 by cleavage , while leaving its chemotactic activity intact [10] . C . rodentium does not appear to have such intrinsic resistance mechanisms; however , it is possible that other host-derived mechanisms may play a role in certain infections . For example , interaction of Streptococcus dysgalactiae with human serum albumin blocks some CXCL9-directed killing activity [51] . Whether such a mechanism is relevant in intestinal infections is not known . In summary , our data indicate that CXCL9 plays an important role in antimicrobial defense in the infected and inflamed gut . This activity , independent of the chemokine receptor CXCR3 or an adaptive immune response , protects the gut from crypt invasion by C . rodentium and the tissue damage that ensues . These data add to the growing body of evidence to support this chemokine as an innate antimicrobial defense molecule at mucosal surfaces .
All animal experiments were conducted according to guidelines set by the Canadian Council on Animal Care using protocols approved by the Animal Review Ethics Board at McMaster University . Six to eight-week old C57BL/6 , Rag1-/- , IFNγ-/- , and CXCR3-/- mice were purchased from Jackson Laboratories . Survival experiments were performed with 4-week old Rag1-/- mice . All animals were housed in a specific pathogen-free unit under Level 2 conditions at the Central Animal Facility at McMaster University . For all infections , mice received 2 x 108 CFU/mL via orogastric gavage from an overnight culture of Citrobacter rodentium ( DBS100 ) . For infections , bacteria were pelleted , washed , and resuspended in 10 mM HEPES ( pH 8 . 0 ) , 0 . 9% NaCl . Bacterial burden was monitored at designated time points in feces throughout experiments . At day 10 post infection , mice were sacrificed , and C . rodentium burden was determined in the cecum and colon as previously described [19] . For antibody neutralization experiments , mice were given either 200 μg/mL rabbit anti-mouse CXCL9 , or 200 μg/mL control rabbit IgG on day -1 , 0 , 1 , and then every 3 days via intraperitoneal injection . All neutralization antibodies were column-purified from rabbit antisera , which was a kind gift from Dr . Cory Hogaboam ( University of Michigan/Cedars-Sinai Medical Center ) . Bacterial killing assays were performed with wild type C . rodentium , enterohemorrhagic E . coli , enteropathogenic E . coli , S . Typhimurium , a ΔphoPQ C . rodentium mutant , and E . coli K12 . The phoPQ deletion was generated by Lambda Red mutagenesis according to published methods [52] using primers BRT151 ( tta gcc gtc ctt ctg ccc cgg ctg ctg tcg gcc aaa aat gac ctc cat gtg tag gct gga gct gct tcg ) and BRT152 ( atg cgc gtt ctg gtt gtt gag gat aat gcg tta cta cgt cac cac ctg cat atg aat atc ctc ctt a ) . Stationary phase ( 16–18h ) cultures were sub-cultured 1:50 in LB , and grown at 37°C with shaking until OD600 = 0 . 5 . Bacteria were pelleted , washed and resuspended in 10 mM HEPES buffer ( pH 7 . 4 ) to a concentration of 105 CFU/mL . Killing was initiated by mixing bacteria with 50 μg/mL human α-defensin ( HD5; Prospec ) , CXCL9 ( 0 . 39 μg/mL , or otherwise indicated concentration; Peprotech ) , or sterile water . Bacteria were incubated at room temperature for 2 h , unless otherwise indicated . Cultures were diluted 1:10 with PBS to quench killing , and viable bacterial counts were assessed on solid agar . All data was normalized to the water control and expressed as survival relative to time zero . For CXCL9 neutralization in the bacterial killing assays , 200 μg/mL antibody ( or similarly diluted PBS ) was pre-incubated with 5 μg/mL CXCL9 for 20 min prior to the assay . To assess integrity of bacterial cellular membranes , the fluorescent probe , 8-anilino-1-naphthylenesulfonic acid ( ANS; Sigma-Aldrich ) was used according to previous published protocols [53] . In brief , stationary phase cultures of C . rodentium were sub-cultured 1:50 in LB , and grown at 37°C with shaking until OD600 = 0 . 5 . Bacteria were pelleted , washed , and resuspended in sterile 10 mM HEPES Buffer ( pH 7 . 4 ) , 5 μM carbonyl cyanide 3-chlorophenylhydrazone ( CCCP; Sigma Aldrich ) , and 5 mM glucose . Bacteria were incubated for 30 min at room temperature . For each sample , 93 μL of bacteria was added to each well of black , clear bottom , 96-well plate ( Costar , Corning , Inc . ) with 2 μL 3 mM ANS , and fluorescence was monitored on a Synergy HT microplate reader ( BioTek ) ( excitation , 375nm; emission , 510nm ) . After 5 min , 10 μg/mL CXCL9 , 10 μg/mL Polymyxin B ( Sigma-Aldrich ) , or water control was injected and fluorescence was monitored for an additional 30 min . At 10 days post infection , segments of cecal tip or distal colon were collected and either fixed in buffered 10% formalin , or flash frozen in optimal cutting template compound ( OCT; Sakura , Fisher ) . Segments were fixed for 72 h , paraffin-embedded , sectioned into 6 μm slices , and stained with hematoxylin and eosin ( H&E ) , anti-CD3 antibody ( 1:1000; Labvision ) , or anti-Citrobacter antibody ( 1:4000; Statens Serum Institute ) . H&E sections were used for assessing pathology according to published scoring protocols [54] . All fixed sections were visualized using a Leica microscope . A minimum of 6 views were analyzed for each sample . Evaluation of C . rodentium crypt invasion was determined through enumeration of bacterially penetrated crypts from 4–6 views per sample . Anti-CD3 treated sections were enumerated using ImageJ software . At day 8 post infection , wild type C57BL/6 mice were euthanized , and CXCL9 levels were detected in various tissues . Rectal perfusions were performed with 0 . 5mL PBS with protease inhibitors ( PBS-I , 10 mL PBS per 1 tablet complete Mini , EDTA free ( Roche ) ) . Colon was excised , flushed with 5 mL ice cold PBS , and then opened longitudinally . Mucus scrapings were performed by running a razor blade along the length of the open colon according to established protocols [55] . Collected scrapings were diluted 1:10 in PBS-I , and vortexed for 3 min . Remaining colonic tissue was placed in 0 . 5 mL PBS-I with a metal ball and homogenized in a Mixer Mill ( Retsch ) . Fecal pellets were similarly homogenized in 0 . 5 mL PBS-I . Solid particulates from all homogenized samples were pelleted through centrifugation prior to addition to ELISA plate wells . CXCL9 was quantified using a Duoset murine CXCL9 ELISA ( R&D Systems ) according to the manufacturer’s protocol . At day 10 post infection , cecum and colon were removed , flushed of contents , and washed in ice-cold PBS , pH 7 . 4 . Tissues were cut into 5 mm pieces , and placed in 1 mL RPMI , 50 μg/mL gentamicin . Tissues were incubated for 24 h at 37°C , 5% CO2 . Levels of CXCL9 , IFNγ , TNFα , IL-12p40 , and IL-10 were assessed using Duoset Quantikine murine ELISA kits ( R&D Systems ) according to the manufacturer’s protocols . For IgG analysis of fecal pellet homogenate supernatants , fresh fecal pellets were collected on day 2 post final injection from naïve , uninfected animals receiving two injections of 200 μg/mL anti-CXCL9 antibody , control IgG , or PBS . Fecal pellets were homogenized as described above , and then centrifuged at 14 , 000 g for 20 min . Supernatants were collected , and frozen at -80°C until analysis . IgG ELISA was performed on samples by coating high-binding 96-well plates overnight at room temperature with goat anti-rabbit anti-IgG antibody ( 1:10 , 000; MP Bio ) . Plates were washed 3 times with PBS ( pH 7 . 4 ) , 0 . 05% Tween-20 , and then blocked with PBS , 1% BSA for 1 h at room temperature . Plates were washed , and samples were added and incubated for 2 h at room temperature . For each sample , a 1:2 dilution ( in PBS , 1% BSA ) was made , and 100 μL was added to each well . After incubation , the plate was again washed three times . Bound antibody was detected by addition of anti-rabbit IgG antibody conjugated to HRP ( 1:10 , 000; GE Healthcare ) , and incubated at room temperature for 2 h . Plates were washed , and developed with 100 μL solution A and B ( R&D Systems ) for 20 min . The reaction was stopped by addition of 50 μL 1N H2SO4 . Plates were read at 450 nm using a plate reader . NK cells were purified from whole splenocytes from wild type C57BL/6 or IFNγ-/- mice using a NK cell ( CD49b+ ) negative selection enrichment kit from StemCell technologies according to the manufacture’s protocol . Purified NK cells ( 1x105 ) were cultured in the presence or absence of 3x105 bone marrow-derived macrophages ( BMDM ) or dendritic cells ( BMDC ) derived from uninfected wild type C57BL/6 mice or IFNγ-/- mice with or without 1 ng/mL IFNγ , heat killed ( hk ) C . rodentium ( 3x106 bacteria ) and 8 ng/ml recombinant IL-2 for 24 h at 37°C in 5% CO2 . Data was analyzed using GraphPad Prism ( ver . 5 . 0d ) . Significance was assessed using the Student’s t test or ANAOVA as indicated in the figure legends . P-values less than 0 . 05 were considered significant .
|
Host defense peptides are an essential part of the innate immune response to pathogens , particularly at mucosal surfaces . Some chemokines , previously known for their ability to recruit immune cells to a site of inflammation , have been identified to have direct antimicrobial activity in vitro against a variety of pathogens . Despite this , it was unknown whether chemokines play a role in protecting the gut mucosa against enteric pathogens , independent of their immunological receptors . Using a mouse model of enteric pathogen infection with both wild type mice and genetic knockouts , we showed that the chemokine CXCL9 has direct antimicrobial activity against pathogen infection . This antimicrobial activity prevented the invasion of bacteria into intestinal crypts , thus protecting the host from immunopathology . Neutralization of this CXCL9-dependent antimicrobial activity increased host susceptibility to infection , leading to bacterial penetration into intestinal crypts and increased tissue pathology . These data support the importance of a receptor-independent role for chemokines in host defense at mucosal surfaces and may offer alternative treatment strategies for infections , particularly in regards to organisms that are resistant to conventional antibiotics .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
CXCL9 Contributes to Antimicrobial Protection of the Gut during Citrobacter rodentium Infection Independent of Chemokine-Receptor Signaling
|
Tolerance to drugs that affect neural activity is mediated , in part , by adaptive mechanisms that attempt to restore normal neural excitability . Changes in the expression of ion channel genes are thought to play an important role in these neural adaptations . The slo gene encodes the pore-forming subunit of BK-type Ca2+-activated K+ channels , which regulate many aspects of neural activity . Given that induction of slo gene expression plays an important role in the acquisition of tolerance to sedating drugs , we investigated the molecular mechanism of gene induction . Using chromatin immunoprecipitation followed by real-time PCR , we show that a single brief sedation with the anesthetic benzyl alcohol generates a spatiotemporal pattern of histone H4 acetylation across the slo promoter region . Inducing histone acetylation with a histone deacetylase inhibitor yields a similar pattern of changes in histone acetylation , up-regulates slo expression , and phenocopies tolerance in a slo-dependent manner . The cAMP response element binding protein ( CREB ) is an important transcription factor mediating experience-based neuroadaptations . The slo promoter region contains putative binding sites for the CREB transcription factor . Chromatin immunoprecipitation assays show that benzyl alcohol sedation enhances CREB binding within the slo promoter region . Furthermore , activation of a CREB dominant-negative transgene blocks benzyl alcohol–induced changes in histone acetylation within the slo promoter region , slo induction , and behavioral tolerance caused by benzyl alcohol sedation . These findings provide unique evidence that links molecular epigenetic histone modifications and transcriptional induction of an ion channel gene with a single behavioral event .
Drug tolerance can be defined as reduced responsiveness to an effect of a drug caused by prior exposure to the drug [1] . With regard to recreational drugs , tolerance can cause an increase in self-administration and thereby speed the user down the path to addiction and/or to overdose . Most organic solvents are potent central nervous system depressants that produce sedation if inhaled or consumed in sufficient quantities . These properties have led to the use of such solvents both as anesthetics and as drugs of abuse . We have used the model system Drosophila melanogaster to study the neuronal basis of tolerance to organic solvent sedation . When exposed to a small sedating dose of an organic solvent , such as benzyl alcohol ( BA ) or ethanol , Drosophila acquire rapid tolerance to subsequent solvent sedation . Rapid tolerance is defined as reduced drug responsiveness caused by a single prior exposure to the drug . In flies , this manifests itself as a reduction in the duration of sedation . Changes in the expression of the slo Ca2+-activated K+ channel gene have been linked to the production of rapid tolerance . It has been shown that sedation by a variety of methods induces slo gene expression in the nervous system , that slo mutations block the acquisition of behavioral tolerance , and finally , that transgenic induction of slo phenocopies the tolerant phenotype . Thus , the transcriptional regulation of the slo gene appears to be of general importance for the production of tolerance to sedation by various organic solvents [2–4] . We wish to understand how sedation with organic solvent anesthetics modulates the slo transcriptional control region . We continue to use BA as a model organic solvent anesthetic in this endeavor because it is well tolerated by flies and can be easily administered . One of the first steps in transcriptional activation is commonly thought to be the alteration of chromatin structure . Specific amino acids in the N-termini of core histones can be modified by phosphorylation , acetylation , methylation , or ubiquitylation [5] . In particular , histone acetylation is believed to relax chromatin to make the DNA more accessible for recognition and binding by the transcriptional machinery [6] . Histone acetyltransferases ( HATs ) and histone deacetylases ( HDACs ) are enzymes that modulate histone acetylation states . Many transcription activators , such as cAMP-response element binding protein ( CREB ) binding protein ( CBP ) /p300 , have HAT activity , and some transcription repressors , such as Sin3 and RPD3 , have HDAC activity ( for review , see [7] ) . Recent studies indicate that histone acetylation contributes to the regulation of neural excitability and synaptic plasticity . It has been shown to have roles in learning and memory , in the production of circadian rhythms , and in the response to seizure , and has been identified as an important component underlying cocaine-induced neural plasticity [8–16] . Moreover , the administration of HDAC inhibitors has been shown to enhance long-term memory , arrest neurodegeneration , and alter cocaine responses [11 , 12 , 17 , 18] . Transcriptional regulation of the Drosophila slo channel gene is very complex . It has a 7-kb regulatory region that includes at least five transcriptional promoters that mediate developmental and tissue-specific gene expression . The two upstream promoters are neural specific . Neural expression is requisite for the acquisition of tolerance to both the anesthetic BA and to ethanol [2 , 4 , 19 , 20] . Within the slo transcriptional control region , there exist multiple putative CREB DNA binding sites . The history of CREB makes it an attractive candidate for the sedation-mediated activation of the slo gene and the production of tolerance . CREB is a key transcription factor involved in regulating neural plasticity in response to various environmental stimuli [21] . Activated CREB binds to the cis-regulatory DNA element cAMP-response element ( CRE ) to induce target gene expression . In mammals , CREB has been implicated in the production of neuronal changes associated with drug tolerance and addiction . Exposure to addictive drugs , including opiates , cocaine , nicotine , and ethanol , induces CREB phosphorylation and CRE-mediated gene expression in the mouse nucleus accumbens ( NAc ) , a major brain reward region [22–27] . The elevated CREB function appears to cause tolerance to and dependence on drugs of abuse . For instance , increased CREB function in the NAc reduces the sensitivity of mice to the rewarding effects of cocaine , whereas decreased CREB function induces cocaine sensitivity [28] . Although the CREB target genes that contribute to these specific effects are still not known , it has been shown that genes that encode some peptide neurotransmitters ( NPY and CRF ) , neurotrophic factors ( BDNF and IGF1 ) , and even transcription factors ( c-Fos ) are downstream targets of CREB in specific brain regions [29–34] . However , these are thought to represent only a small subset of the genes regulated by CREB , and many CREB target genes have not been identified [30] . One way that CREB can stimulate transcription is by recruiting CBP , which is a HAT [35] . By inducing histone acetylation , CBP makes DNA more accessible to the transcriptional machinery [5] . It has been shown that the histone acetylation at c-Fos gene promoter , which is mediated by the recruitment of CBP , contributes to cocaine sensitization [36] . In Drosophila , the cAMP-PKA signaling pathway , a pathway activating CREB , has been implicated in the development of drug tolerance . Drosophila carrying a mutation in amnesiac , a gene encoding a neuropeptide that activates the cAMP pathway , show increased sensitivity to alcohol [37 , 38] . Conversely , fly mutants with increased PKA activity show decreased sensitivity to ethanol and cocaine [39] . Here , we show that a sedating dose of BA changes the chromatin acetylation state of the slo promoter region and that these changes are linked to increased slo expression , which produces behavioral tolerance to the sedative effect of BA . Furthermore , we show that the CREB transcription factor has an essential role in the production of acetylation changes , slo induction , and behavioral tolerance to organic solvent sedation .
A strong link exists between gene induction and the acetylation of amino-terminal lysines in histone H4 in promoter regions [6 , 40 , 41] . We used the chromatin immunoprecipitation ( ChromIP ) assay to detect changes in histone H4 acetylation across the slo transcriptional control region , following BA sedation . The antibody used in ChromIP recognizes histone H4 acetylated at lysines K5 , K8 , K12 , and K16 in Drosophila and mammals [10 , 42] . The amount of DNA associated with acetylated histone H4 was quantified by real-time PCR using primers specific for conserved portions of the slo transcriptional control region ( 4b , 6b , cre1 , 55b , and cre2 ) , the two neural promoters ( C0 and C1 ) , and the muscle promoter ( C2 ) ( Figure 1A ) . As a control , the histone acetylation levels at the promoter of the Gpdh gene were measured . Neither the Gpdh mRNA abundance nor the acetylation state of the Gpdh promoter was altered by BA sedation . No significant change in histone acetylation was observed 30 min after BA sedation ( Figure 1B ) . However , we observed a finely focused increase in histone H4 acetylation in the vicinity of 55b 4 h after BA sedation ( Figure 1C ) . This acetylation change is likely to represent an early step in gene induction at which increased slo expression is not yet apparent ( Figure 1G ) . The detection of increased message abundance may require time for the product to accumulate . Six hours after BA sedation , we observed a broad acetylation peak that included both neural promoter C0 and C1 ( Figure 1D ) . This coincided with increased neural expression of slo ( Figure 1H ) . By 24 h , slo expression appeared to be in decline ( Figure 1I ) , and the enhanced histone H4 acetylation was again finely focused . However , this time , acetylation was centered over 6b , which is about 300 bp upstream of neural promoter C1 ( Figure 1E ) . These data indicate that histone acetylation across the slo transcriptional control region is dynamically modulated after sedation with the anesthetic BA . By 48 h , the acetylation level and slo mRNA level have returned to the pretreatment level ( Figure 1F and 1J ) . To determine whether the observed changes in histone acetylation are related to slo induction and to the acquisition of BA sedation–induced BA tolerance , we used sodium butyrate to induce histone acetylation . Sodium butyrate is an extremely well-tolerated , nonspecific HDAC inhibitor that has been used to artificially induce histone acetylation both in mammals and insects [11 , 18 , 43–46] . Newly eclosed wild-type Canton S flies were split into two groups that differed only in the presence or absence of sodium butyrate in their food . Flies in the experimental group were fed food that contained 0 . 05 M sodium butyrate , and control group flies were fed food that lacked sodium butyrate . After 3 d , fly heads were collected , and ChromIP assays were performed to determine the acetylation state of histone H4 . Overall histone H4 acetylation levels , expressed as the ratio of DNA associated with the acetylated chromatin to input DNA , were increased in the experimental group . As previously reported [18 , 42 , 43] , sodium butyrate administration increased the global histone H4 acetylation level by approximately 30% ( Figure 2A ) . The increase in acetylation is not uniform across the slo control region . Statistically significant increases in acetylation were observed only at region 6b ( Figure 2B ) . Sodium butyrate , a nonspecific HDAC inhibitor , alters acetylation at the slo promoter region because it raises global histone acetylation levels . However , BA does not behave as a nonspecific enhancer of histone acetylation , because sedation with BA does not enhance global acetylation levels ( Figure 2A ) . Our working hypothesis was that BA sedation increases histone acetylation , which in turn stimulates slo expression , and that this increase in slo expression leads to behavioral drug tolerance . In support of this hypothesis , we observed that sodium butyrate consumption both induced slo expression and phenocopied BA tolerance . As shown in Figure 3A , sedation with BA caused a 50% increase in neural slo mRNA abundance 6 h after sedation , whereas the consumption of sodium butyrate–laced food produced a 2-fold induction in neural slo messenger abundance ( Figure 3A ) . A typical example of BA-induced behavioral tolerance is shown in Figure 3B . As previously described [2] , flies recover more rapidly from their second sedation than from their first sedation . Figure 3C shows that the flies fed food laced with sodium butyrate exhibited a tolerance-like phenotype . Importantly , we observed that BA sedation of sodium butyrate–fed flies did not further enhance the expression level of slo ( Figure 3A ) or the relative degree of BA resistance ( Figure 3D ) . This observation is consistent with the proposal that BA and sodium butyrate affect these characteristics through a common , saturable pathway . Sodium butyrate increases the level of acetylated histones globally [42 , 43] . Therefore , the tolerance-like phenotype caused by sodium butyrate consumption might be unrelated to the slo gene . To exclude this possibility , we determined whether slo mutations interfere with the capacity to acquire tolerance . The slo4 allele is a loss-of-function slo mutation . It has been characterized as a null mutation genetically , molecularly , behaviorally , and electrophysiologically [47–49] . The slo4 homozygous flies show subtle behavioral differences from wild-type flies , one of which is that they do not acquire tolerance following a single BA or ethanol sedation [2 , 3] . To test whether sodium butyrate–induced BA resistance was related to the slo gene , we asked whether the slo4 mutation could block the capacity to acquire sodium butyrate–induced BA resistance . The consumption of sodium butyrate did not induce the tolerance-like phenotype in slo4 homozygotes , but the wild-type Canton S flies , raised and tested under the same conditions , displayed the tolerance-like phenotype ( Figure 4A and 4B ) . This indicates that the induction of BA resistance by sodium butyrate is dependent on a functional slo gene . Previous studies have indicated that only neural expression of slo is involved in the production of BA tolerance [2] . Therefore , we asked whether sodium butyrate–induced BA resistance could be blocked merely by eliminating neural expression of slo . The ash218 mutant chromosome carries a deletion that removes the two neural promoters of slo , but not the promoters responsible for expression in muscle , tracheal cell , or epithelia tissues . The deletion is a recessive lethal because it removes the neighboring gene [2 , 49 , 50] . Therefore , we used ash218/slo4 transheterozygotes as a way to specifically eliminate slo expression in the nervous system . As shown in Figure 4C , the elimination of slo expression in the central nervous system prevents the induction of BA resistance by sodium butyrate . Figure 4D shows that sodium butyrate–induced BA resistance is a recessive phenotype of the slo4 mutation . In Drosophila , two CREB family genes have been identified , dCREB-A and dCREB2 ( also called dCREB-B ) . dCREB-A is weakly expressed in the adult brain [51] and has not yet been linked to behavioral phenotypes . The dCREB2 gene is most similar to mammalian CREB and CREM genes and is considered to be the Drosophila homolog of these genes [52] . dCREB2 is expressed in the adult brain , and has been implicated in the formation of learning , memory , and circadian behavior [52–54] . As with other transcription factors , CREB has functionally independent activation ( KID ) and DNA binding ( bZIP ) domains . In Drosophila , it has been shown that the KID domain of most dCREB2 proteins exists in a phosphorylated active state . As a result , in Drosophila , CREB activity has been postulated to be controlled at the level of DNA binding to its DNA elements [55] . Two putative CRE sites were identified close to 55b in the slo control region by sequence analysis . Furthermore , the highly conserved 55b element contains an AP-1 binding site motif . Both AP-1 and CREB belong to the b-ZIP family of transcription factors which share a similar DNA binding motif , the leucine zipper . The AP-1 response element ( TGACTCA ) is quite similar to the CREB response element , with only one base difference ( CRE sequence: TGACGTCA ) . Both CREB and AP-1 can induce local histone acetylation by recruiting CBP HAT [35] . We performed the ChromIP assay using an anti-CREB antibody that recognizes the Drosophila dCREB2 DNA binding domain ( Santa Cruz Biotechnology , http://www . scbt . com ) . We determined the relative occupancy of dCREB2 protein within the slo promoter region both before and after drug sedation . The data indicate a clear drug-induced increase in CREB binding at and around the 55b region ( Figure 5 ) . As previously described , BA sedation altered histone H4 acetylation across the slo promoter region . Shortly after sedation , acetylation was induced at 55b; a conserved site that is flanked by two CRE sites . It has been shown that binding of CREB at CRE sites can induce local histone acetylation by recruiting CBP and thereby activate gene expression [13 , 56] . To test whether the early acetylation peak over 55b within the slo control region is caused by the recruitment of CREB and CBP , we asked whether a dominant-negative CREB transcription factor could prevent the early histone acetylation peak at 55b . To do this , we transgenically expressed the repressor isoform of the dCREB2 . As in mammals , the dCREB2 transcript is alternatively spliced . The dCREB2a splice variant acts as a transcriptional activator , whereas the dCREB2b isoform has repressor activity [53] . A dominant-negative transgenic line that expresses dCREB2b under the control of a hsp70 heat-shock promoter ( hs-dCREB2b ) provided a way to inducibly inhibit CREB-mediated gene activation [54 , 57] . Previous studies have shown that a brief heat pulse ( 37 °C for 30 min ) induces the transgene ( hs-dCREB2b ) and that elevated protein levels persist up to 24 h [54] . In this study , the dCREB2b transgene was induced , and histone acetylation levels were measured after BA sedation . These acetylation levels were compared to those from flies of the same genotype in which the transgene had not been activated . We observed that the induction of dominant-negative CREB by a brief heat pulse 1 h prior to BA sedation eliminated the histone acetylation peak around 55b ( Figure 6 ) . The same heat shock protocol has no effect on the histone acetylation changes at 55b in wild-type Canton S flies . This result suggested that the formation of an early acetylation peak requires the binding of CREB to CREs at slo transcriptional control region . The previous data indicate that CREB activity is involved in the drug sedation–induced hyperacetylation within the slo promoter region . Acetylation changes have been linked to slo induction . Therefore , we asked whether dominant-negative CREB could interfere with the induction of slo after drug sedation . Both hs-dCREB2b transgenic females and wild-type CS females were subjected to 37 °C incubation for 30 min , 1 h before BA exposure , and the relative slo messenger levels were measured 6 h after drug exposure by real-time reverse transcriptase PCR ( RT-PCR ) . We observed that induction of the dominant-negative CREB transgene 1 h prior to BA sedation suppressed slo induction ( Figure 7 ) . The same heat shock protocol did not suppress slo induction in wild-type CS flies following BA sedation . This result suggests that disruption of early histone acetylation pattern by inhibiting CREB activity blocks BA sedation–induced slo up-regulation . The data above indicate that overexpression of dominant-negative dCREB2 can block drug-induced acetylation in the slo promoter region and the up-regulation of slo expression . Because slo induction appears causally linked to the production of tolerance in this behavioral paradigm , the activation of the hs-dCREB2b transgene should also block the acquisition of tolerance . To test this idea , we examined induced and uninduced hs-dCREB2b transgenic flies for their capacity to acquire tolerance . Uninduced hs-dCREB2b flies acquired robust tolerance ( Figure 8A ) . However , induction of dCREB2b transgene by heat shock 1 h before the first BA sedation completely blocked the acquisition of tolerance ( Figure 8B ) . This occurred without a change in the initial sensitivity of flies to drug sedation ( Figure 8C ) . The heat-activation protocol did not alter the relative magnitude of tolerance acquired by a wild-type Canton S stock ( Figure 8D and 8E ) . These data indicate that CREB transcription factor activity is important for the acquisition of rapid tolerance in flies .
It has become clear that drug-induced changes in gene expression play an important role in the pharmacodynamic response to drugs of abuse . Epigenetic modifications to promoter regions are rapidly emerging as an important mechanism for producing these changes [11 , 58] . Epigenetics refers to DNA and chromatin modifications that influence chromatin structure and change the state of gene expression without altering the nucleotide sequence [59] . Hyperacetylation of histone H4 has been linked to neural gene activation in activity-dependent signaling pathways and in synaptic plasticity [10 , 13] . In Drosophila , sedation with BA induces neural slo expression in a dose-dependent manner , and increased slo gene expression has been implicated in the production of rapid tolerance to both ethanol and BA [2–4] . An increase in histone acetylation is generally accepted to make DNA sequences more available to transcription factors . Hyperacetylation at one site might beget subsequent hyperacetylation at a second site . That is , the activity of one transcription factor could modify an area to facilitate the binding of additional transcription factors . The action of these factors may continue to further modify the region . We show that sedation with the anesthetic BA produces a specific spatiotemporal pattern of histone H4 acetylation across the slo promoter region . These epigenetic changes are correlated with changes in slo transcription and with the development of drug tolerance , and are thought to be the molecular footprints of a regulatory cascade that is initiated by BA sedation . The acetylation spike at 55b occurs immediately prior to the induction of slo expression and may make available specific sequences that are necessary to activate the upstream neural promoters . Two hours later , we recorded the greatest increase in slo expression and an increase in acetylation that involves most of the slo neural promoter region . The process of transcription itself can enhance acetylation , and therefore , this broad boost in acetylation may be a direct by-product of transcription [60] . Finally , at 24 h post BA sedation , there is an isolated acetylation spike at 6b , and slo expression remains elevated ( albeit slightly less elevated than at 6 h post sedation ) . This may mean that increased accessibility of 6b is required to maintain slo induction . We propose that this dynamic pattern of hyperacetylation represents an unfolding regulatory program that leads to a transient and self-limiting boost in slo channel expression . We used the HDAC inhibitor sodium butyrate to artificially induce histone acetylation in the genome . We asked whether a simple change in acetylation would induce slo gene expression and whether it could also produce a tolerance-like phenotype . We observed that sodium butyrate consumption induced slo expression and phenocopied the BA sedation–induced tolerance phenotype . In various preparations , sodium butyrate has been shown to enhance histone acetylation and to induce expression of a small fraction of the genes in the eukaryotic genome ( estimates range from 1% to 7% ) [45 , 61 , 62] . Therefore , our result would be irrelevant if we could not directly link sodium butyrate–induced BA resistance to slo , and demonstrate other similarities to the effects of BA sedation . Using slo null mutations , we were able to show that the capacity of sodium butyrate to phenocopy tolerance is dependent on slo expression in the nervous system , as is rapid tolerance induced by BA or ethanol sedation [2 , 3] . In addition , BA sedation of sodium butyrate–fed flies did not further enhance resistance , suggesting that both act through a common , saturable pathway . Finally , we were surprised to find some similarity between sodium butyrate and BA sedation upon the pattern of acetylation on the slo transcription control region . Sodium butyrate also caused hyperacetylation at the conserved 6b sequence , but did not detectably change acetylation at C1 , cre1 , or 55b sites . This mimics the acetylation pattern observed 24 h after BA sedation . Since sodium butyrate inhibits HDAC activity , the enhancement of 6b acetylation by sodium butyrate suggests that it has inhibited an HDAC that is chronically positioned near or at the 6b site . The simplest interpretation is that enhanced availability of 6b permits slo induction . Both 55b and 6b are non–promoter-containing DNA sequences that were originally identified because of their conservation between two Drosophila species [19 , 63] . Conservation alone is a strong indicator that these sequences have important roles in the regulation of the slo gene . Figure 9 shows that they are highly conserved across at least eight Drosophila species . Two putative CREB binding sites are located close to 55b . The cre1 site is TGACGAA and cre2 is TGACGTAA . The cre1 site matches the canonical CRE motif , whereas cre2 differs in a single nucleotide . Both include the first five bases of the consensus CRE sites ( TGACG ) , which are sufficient for CREB-mediated transcription [64] . Furthermore , in 55b , there is a putative AP-1 site ( TGATTCA ) which differs in two nucleotides from the canonical CRE site . AP-1 and CREB carry similar dimerization and DNA binding b-ZIP domains . Studies suggest AP-1 and CREB can form cross-family heterodimers and share same consensus DNA binding elements [65] . The AP-1 and CREB transcription factors have been implicated in the neural response to abused drugs . AP-1 transcription factor complexes are dimers formed from the Jun and Fos family of transcription factors . Acute administration of certain abused drugs causes complex changes in the pattern of the Fos family members [27 , 66–68] . These changes affect drug responsiveness and are thought to underlie long-lasting sensitization to cocaine in mammals [67] . The modulation of CREB activity has been linked to drug tolerance and dependence and many other neural responses that can all be considered to be forms of neural plasticity [67 , 69–71] . Both AP-1 and CREB can stimulate gene expression by recruiting cofactor CBP , which helps to position the basal transcriptional machinery [35 , 72] . CBP is a HAT that through histone acetylation causes local decondensation of chromatin [35] . We postulated that AP-1 or CREB might be involved in acetylation at 55b . ChromIP experiments indicated that AP-1 did not bind the slo promoter region ( unpublished data ) . However , ChromIP did show that CREB bound in this region and that drug sedation enhanced CREB occupancy . We further implicated CREB , in drug-mediated slo expression and behavioral tolerance , by showing that a dominant-negative CREB transcription factor could block hyperacetylation surrounding 55b , the induction of slo , and behavioral tolerance caused by BA sedation . These data suggest a chain of events for the regulation of slo in response to BA sedation . That is; drug sedation activates the CREB signaling pathway . CREB binds the CRE sites flanking 55b and perhaps within 55b itself , and then recruits the CBP HAT to acetylate histones in the neighborhood of 55b . This makes sequences at 55b available for binding by factors that lead to further modifications and result in the inhibition of the HDAC positioned at 6b . Inhibition of this HDAC augments 6b acetylation , making it available for binding by a factor ( perhaps HSF , Figure 9 ) that directly stimulates expression from the two neighboring neural promoters . Blockage of any of these events would interfere with the drug-induced slo expression and animals' ability to develop tolerance . Although other more complex models are possible , this simple model will eventually be testable as new tools become available , and is useful for organizing ideas about how slo senses and responds to drug sedation . We have previously postulated that the induction of slo is a homeostatic response that acts to reverse decreased neural excitability associated with sedation by the drug BA . Anesthesia induces slo gene expression , and by itself , the induction of this channel gene phenocopies tolerance [2] . Our data are consistent with the idea that increased BK channel expression reduced the duration of sedation caused by BA and produced a tolerance-like phenotype . This is an unusual role to postulate for a K+ channel . Certainly , in some preparations , increased BK channel activity reduces neural excitability [73–75] . However , in other preparations , BK channel activity has been positively correlated with neural excitability [76–80] . It has been proposed that an increase in BK channel activity limits the instantaneous response of the cell , but augments the capacity for repetitive neural activity by reducing the neural refractory period [76 , 77] . The refractory period is the time that must elapse before the neuron can fire again . Neural pharmacodynamic tolerance to any drug is likely to involve many components [81–84]; however , the slo gene is uniquely positioned to be a homeostatic regulator of neural excitability . The encoded channel has the highest conductance of any neural ion channel , thus small changes in its density can have a large influence on membrane excitability . In conclusion , we propose that in flies , BA sedation causes CREB-mediated epigenetic changes in the slo control region that result in an increase in slo expression , which significantly enhances the excitability of the nervous system to help produce the tolerance phenotype .
Drosophila stocks were Canton S ( wild type ) , slo4 , ash218/slo4 , and the hs-dCREB2b . All fly stocks were raised on standard cornmeal agar medium ( 12/12-h light/dark cycle ) . Newly eclosed flies collected over a 1- to 2-d interval were studied 4–5 d after eclosion . The hs-dCREB2b transgene expresses a dCREB2b cDNA from a hsp70 promoter [54] . Age-matched female Canton S flies were divided into six groups of 15 . Three groups were sedated with 0 . 4% BA , and three were mock sedated as previously described [2] . Sodium butyrate ( 0 . 05 M , 99% purity; Fisher Scientific , https://www . fishersci . com ) in food was fed to 1- to 2-d-old flies for 3 d . Controls were fed unadulterated food . Tolerance was measured as described [2] . In the first exposure , three experimental groups were exposed to 0 . 4% benzyl alcohol until sedation , and three control groups were mock sedated in parallel . Twenty-four hours later , all groups were sedated with BA , then transferred to an anesthetic-free environment , and recovered flies were counted every 30 s . Flies were scored as recovered when they resumed climbing . The log-rank test was used to determine significance between curves [85] . However , error bars represent the standard error of the mean ( SEM ) for each point . About 1 , 500 wild-type flies were either BA sedated or mock sedated for 5 min and then were allowed to recover in a BA-free environment [2] . Thirty minutes , and 4 , 6 , 24 , and 48 h after sedation , flies were collected , frozen in liquid nitrogen , and vortex decapitated , and then heads were collected by sieving . Heads were cross-linked with 2% formaldehyde for 5 min , and chromatin was solubilized and sonicated on ice 6 × 30 s , followed by 1 min cooling on ice to produce fragments of approximately 600 bp with a sonic Dismembrator 250 ( Fisher Scientific ) as described by Orlando et al . [86] . Sheared soluble chromatin was stored at −80 °C . The ChromIP assay was performed as described ( ChIP kit # 17–295; Upstate Biotechnology , http://www . upstate . com ) with minor modifications . One milliliter of soluble chromatin ( 1 mg/ml ) was adjusted to RIPA buffer and then precleared with 50-μl salmon sperm DNA/protein A agarose slurry for 1 h at 4 °C to reduce nonspecific binding . Ten percent of the preimmunoprecipitation lysate ( 100 μl ) was held back and used to determine the input of DNA . The input-level control was processed with the eluted immunoprecipitations ( IPs ) , beginning with the cross-linking reversal step . The polyclonal antibody ( catalog # 06–866; Upstate Biotechnology ) against acetylated H4 at K5 , K8 , K12 , and K16 , and anti-CREB antibody ( sc-186; Santa Cruz Biotechnology ) were used at 1:200 dilution . Five microliters of antibody were added to each sample in 1-ml RIPA buffer and incubated overnight at 4 °C with gentle mixing . Immunocomplexes were recovered by adding 80 μl of the salmon sperm DNA/protein A agarose beads , incubating for 3 h at 4 °C with rotation . The beads were sequentially washed three times in RIPA ( 140 mM NaCl , 1 mM EDTA , 10 mM Tris-HCl [pH 8] , 1% Triton X-100 , 0 . 1% SDS , 0 . 1% sodium deoxycolate ) , twice in high-salt buffer , once in LiCl buffer , and twice in TE buffer , 10 min each . The cross-linking between histones and DNA was reversed by incubating at 65 °C overnight , and DNA fragments were purified with phenol-chloroform extraction followed by acid ethanol precipitation . ChromIP assays were performed more than three times with independent tissue samples . Real-time PCR was performed using the ABI SYBR Green PCR protocol ( Applied Biosystems , http://www . appliedbiosystems . com ) . Within the slo transcriptional control region ( Figure 1A ) , primers were designed to amplify approximately 200-bp fragments at the two neural promoters ( C0 and C1 ) , at one muscle promoter ( C2 ) , and at six evolutionarily conserved areas ( 4b , 6b , cre1 , 55b , S2 , and cre2 ) . As internal controls , we used Gpdh ( Glycerol-3-phosphate dehydrogenase ) and Cyp1 ( cyclophilin 1 ) . Primers sets are: C0 ( 5′-ATCGAACGAAGCGTCCAG-3′ , 5′-CGACGCGCTCAAACG-3′ ) , 4b ( 5′-GACCCGATGATAAAGTCGATGT-3′ , 5′-GCCAGTGACTGACTGACACACA-3′ ) , 6b ( 5′-CCAGCAGCAATTGTGAGAAA-3′ , 5′-CGAAGCAGACTTGAAAGCAA-3′ ) , C1 ( 5′-ACAAACCAAAACGCACAATG-3′ , 5′-AATGGATGAAGGACTGGGAGT-3′ ) , cre1 ( 5′-GATGGGAAAGCGAAAAGACAT-3′ , 5′-CATGTCCGTCAAAGCGAAAC-3′ ) , 55b ( 5′-TACCCAATTGAATTCGCCTTGTCTT-3′ , 5′-CCCACTCTCCGGCCATCTCT-3′ ) , S2 ( 5′-CATTGCTATCCCTTCCCATC-3′ , 5′-ATGCAATGAAGCGAAGAACC-3′ ) , C2 ( 5′-GCACTCGACTGCACTTGAAC-3′ , 5′-AATGAAAAAGTTCTCTCTGTGCAT-3′ ) , cre2 ( 5′-TGGATTGCGACCGAGTGTCT-3′ , 5′-ATCAATACGATAACTGGCGGAAACA-3′ ) , Gpdh ( 5′-GCATACCTTGATCTTGGCCGT-3′ , 5′-GCCCTGAAAAGTGCAAGAAG-3′ ) , and Cyp1 ( 5′-TCTGCGTATGTGTGGCTCAT-3′ , 5′-TACAGAACTCGCGCATTCAC-3′ ) . All amplicons have differences in standard-curve amplification slopes of less than 0 . 1 . Melting curves were used to detect nonspecific amplification . Amplifications were run in triplicate , and the changes on histone H4 acetylation were calculated by the ΔΔCT method . Fold enrichment over control equals 2 ̂ ( CTInput – CTIP ) experiment/2 ̂ ( CTinput – CTIP ) control . The entire protocol was repeated in triplicate a minimum of three times , and the mean and SEM were calculated . Significance was determined by one-way ANOVA . For ChromIP to measure CREB binding , all real-time PCR measurement values were normalized to input DNA in both the BA-sedated and mock-sedated control . The amount of DNA recovered in the IP was expressed as the ratio of input DNA with the equation: IP/input = 2 ̂ ( Ctinput − CtIP ) . The entire protocol has been repeated in triplicate , and the mean and SEM calculated . Significance was determined with the two-way ANOVA . Chromatin from cross-linked fly heads was sonicated and immunoprecipitated with anti-H4 antibody as described above . DNA coimmunoprecipitated with acetylated histone H4 and input DNA were purified by reverse cross-linking followed by phenol-chloroform extraction . DNA was then quantified in a NanoDrop spectophotometer ( NanoDrop Technologies , http://www . nanodrop . com ) . Global histone H4 acetylation levels were expressed as the ratio of the amount of DNA associated with the acetylated chromatin to input DNA . The entire protocol has been repeated four times , and the mean and SEM calculated . Significance was determined with the Student t-test . RNA was isolated from heads using a single-step RNA isolation protocol as described previously [2] and quantified ( NanoDrop Technologies ) . Reverse transcription and real-time RT-PCR were performed in triplicate with slo exon C1-specific primers , which only amplify transcripts from the neural promoters , and cyclophilin 1 primers as described [4] . Fold change was calculated using the standard-curve method ( Applied Biosystems manual ) . Significance was calculated using the Student t-test . Heat-induction protocol for the HS-dCREB2b transgene dCREB repressor ( dCREB2b ) transgene was induced at 37 °C for 30 min in a humidified incubator as described [54] . Immediately after heat shock , flies were sedated with CO2 to reduce the net effect of heat on the neural activity . Half an hour later , flies were stabilized and ready for tolerance assay . Mock induction was performed with Canton S flies ( wild type ) .
The National Center for Biotechnology Information ( NCBI; http://www . ncbi . nlm . nih . gov ) accession numbers for genes mentioned in this paper are ( 32595 ) , dCREB-A ( 39682 ) , dCREB2 ( 32817 ) , Gpdh ( 33824 ) , and slo ( 42940 ) .
|
A startlingly large number of adolescents abuse organic solvent inhalants , common components of glues , paints , and cleaning solutions . Our focus is on the molecular basis of tolerance—reduced response to a drug over time—which promotes increased drug consumption and accelerates the process of addiction . We use the fruit fly Drosophila melanogaster as a model system to determine how the nervous system becomes tolerant to the sedative effects of organic solvents . Sedating Drosophila with an organic solvent ( benzyl alcohol ) increases the expression of the slo K+ channel gene , which accelerates recovery from sedation . To elucidate the molecular mechanics underlying these phenomena , we documented dynamic changes in a chemical modification ( called histone acetylation ) that occurs within the slo regulatory region after sedation . These changes were mediated by a transcription factor and are linked to both slo induction and behavioral tolerance . Increased expression of slo channels is predicted to alter the signaling properties of neurons . This alteration , we propose , directly speeds the recovery from sedation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"mental",
"health",
"pharmacology",
"drosophila",
"neuroscience",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2007
|
Drug-Induced Epigenetic Changes Produce Drug Tolerance
|
Early mammalian development is both highly regulative and self-organizing . It involves the interplay of cell position , predetermined gene regulatory networks , and environmental interactions to generate the physical arrangement of the blastocyst with precise timing . However , this process occurs in the absence of maternal information and in the presence of transcriptional stochasticity . How does the preimplantation embryo ensure robust , reproducible development in this context ? It utilizes a versatile toolbox that includes complex intracellular networks coupled to cell—cell communication , segregation by differential adhesion , and apoptosis . Here , we ask whether a minimal set of developmental rules based on this toolbox is sufficient for successful blastocyst development , and to what extent these rules can explain mutant and experimental phenotypes . We implemented experimentally reported mechanisms for polarity , cell—cell signaling , adhesion , and apoptosis as a set of developmental rules in an agent-based in silico model of physically interacting cells . We find that this model quantitatively reproduces specific mutant phenotypes and provides an explanation for the emergence of heterogeneity without requiring any initial transcriptional variation . It also suggests that a fixed time point for the cells’ competence of fibroblast growth factor ( FGF ) /extracellular signal—regulated kinase ( ERK ) sets an embryonic clock that enables certain scaling phenomena , a concept that we evaluate quantitatively by manipulating embryos in vitro . Based on these observations , we conclude that the minimal set of rules enables the embryo to experiment with stochastic gene expression and could provide the robustness necessary for the evolutionary diversification of the preimplantation gene regulatory network .
At compaction , E2 . 5–E3 . 0 , the outer cells of the embryo become polarized , express the transcription factor Cdx2 , and differentiate into TE . At E3 . 0 , the apical membrane of outer cells expresses several proteins that have a known role in cell polarity [9–13] . The acquisition of polarity starts with compaction at the 8-cell stage , in which apical domain is developed at the contact-free surface . The apical domain is inherited asymmetrically at the next cell division and was shown to play an important role in defining inner and outer cells through cells sorting due to differential contractility [14] . How the TE fate becomes limited to only outer cells is not fully understood , but it is suggested to be a combined effect of contractility , Hippo , and Notch pathways . The Hippo signaling pathway is normally activated at high cell densities and , in this context , is specifically activated in the inner cells [15–18] in which it induces phosphorylation and degradation of the transcriptional coactivator Yes-associated protein 1 ( Yap ) . In outer cells with higher contractility , the levels of phosphorylated Yap are higher [14] . In the absence of Hippo activation , the TEA domain family member 4 ( Tead4 ) binds Yap and cooperates with Notch signaling to induce the transcription factor Cdx2 and specify the TE [19] . In addition , tight junctions are formed between the TE cells in a plane perpendicular to the polarity axis [20 , 21] , and this may further reinforce TE polarity . At E3 . 0 , the inside cells down-regulate Cdx2 but express octamer-binding transcription factor 4 ( Oct4 ) and become ICM . The cells of the ICM initially express both Gata6 and Nanog , but early variations in expression are thought to be propagated by the production of fibroblast growth factor ( FGF ) 4 downstream of Nanog and higher levels of the FGF receptor ( FGFR2 ) in cells expressing higher levels of Gata6 . EPI precursors expressing Nanog secrete FGF4 , promoting a PrE fate in neighboring cells [22–26] . Consistently , ex vivo manipulation of the FGF pathway from E2 . 5 to E4 . 0 can change the fate of ICM cells [27–30] . It has been shown in vitro that Nanog and Gata6 repress each other intracellularly [31–36] . Moreover , FGF/extracellular signal—regulated kinase ( ERK ) signaling enhances Gata6 expression while repressing Nanog [36–39] . Finally , modulating the FGF4 level is sufficient to convert all ICM cells to either PrE ( high FGF4 ) or EPI ( low FGF4 ) [40 , 41] . During the period that cells are making a lineage decision between EPI and PrE , cell movement occurs within the ICM [42] . Chazaud et al . [3] showed that initially EPI and PrE progenitors arise in a heterogeneous mosaic pattern and later physically segregate into the appropriate cell layers , which are finally separated by a basal lamina . It was proposed that the cell movements contribute to cell sorting and may be due to differential adherence of progenitor cell types , which has been observed in vitro [43 , 44] . There have also been several reports on differences in the expression level of the adhesion molecule integrin β1 receptor during PrE differentiation in vitro between the 2 ICM lineages [45–47] . Several other mechanisms contribute to the formation of the “layered” pattern [48] , including down-regulation of transcriptional programs in inappropriately positioned cells or apoptosis [4] . As the blastocyst expands , the ratios of the PrE and EPI are self-regulating , as paracrine interactions control proliferation and apoptosis . In particular , the cytokine Leukemia inhibitory factor ( LIF ) appears to regulate the relative size of the PrE and EPI . LIF is secreted by TE cells , and the corresponding receptor complex is found in the ICM [49] . LIF has been shown to act both on EPI and PrE fate . It blocks maturation in the EPI , and it supports proliferation and cell survival in the PrE [50 , 51] . In addition , atypical protein kinase C ( aPKC ) and platelet-derived growth factor ( PDGF ) signaling promote survival of PrE precursors that reach the surface of the ICM [52 , 53] . Furthermore , a considerable number of ICM cells undergo apoptosis around the time of PrE formation [4 , 54 , 55] . Plusa et al . [4] showed that there is a steady increase in the rate of apoptosis from E4 . 0–E4 . 5 . They reported that PrE precursors are more likely to undergo apoptosis when they are deep within the ICM than when they are positioned along the cavity lining . Here , we hypothesized that a combination of these 4 themes could together explain the robust nature of blastocyst formation . We have conceptualized and unified these themes as rules in a rule-based model to investigate their relative contribution to the robustness of early embryo development . Both the initial specification of the TE and the ICM and differentiation and segregation of PrE and EPI have been modeled in silico at different levels . Chickarmane et al . [56] focused on intracellular transcription networks generating the 3 stable states ( EPI , PrE , and TE ) . Bessonnard et al . [57] modeled 25 static ICM cells on a grid and addressed how cell—cell communication via the FGF/ERK pathway establishes the right proportion of EPI and PrE cells . Krupinski et al . [58 , 59] modeled the mechanical interaction of cells , focusing on the role of polarity in Cdx2 partitioning , as well as differential adhesion and directed movements for the segregation of PrE , EPI , and TE into 3 distinct layers . In these models , the growth of the blastocyst is driven by the growing cavity , and all cells have similar apolar interactions , albeit of different strength . None of these models address the role of polarity in TE cell—cell interaction , apoptosis , or aspects of the emergence of blastocyst scaling [60–62] . The existing in silico models provide important insight into the individual mechanisms driving cell specification during preimplantation development but do not provide a unified framework of early embryo development as a self-organizing system [63] . Recently , such a framework , using rule-based modeling , has been proposed for the specification of synaptic partner cells [64] . Here , we use a similar approach to propose a minimum set of rules to quantitatively model early blastocyst development . Our aim is not to recapitulate the precise timing of mouse development , but to show that with a simple set of rules we could capture blastocyst patterning . As evolution can produce changes in the timing and wiring of the gene regulatory network , the patterning of the mammalian embryo should be able to tolerate stochasticity; our aim is to show that the 4 simple rules can enable this robustness . We focus on 4 rules that include polarity , cell—cell communication via FGF4 signaling , differential adhesion , and apoptosis . Using a series of in silico 2D simulations , we quantify the relative contribution of these 4 elements to early embryonic development . To facilitate comparison to published genetic studies , we have validated this approach in 3D simulations . By introducing polar interactions between TE cells , we show that the development ( including cavity formation ) is self-organized and does not require an a priori assumption of the growing cavity . Moreover , based on these 4 rules , we found that we could effectively simulate experimental embryo manipulation: our model successfully reproduces a range of experimentally observed mutant phenotypes and predicts that the time point of FGF activation could be a clock that dictates the relative size of EPI and PrE in scaling experiments . Consistent with the notion that the timing and duration of FGF/ERK activation is the essential variable in proportioning these 2 lineages , we found that delaying ERK activation by 24 hours resulted in a quantitative reduction in PrE specification .
In a growing blastocyst , cells are tightly packed and adhere to each other . Similarly to earlier in silico models [58 , 65] , we simulate this by introducing an interaction potential in which cells repel each other at a distance smaller than their typical size and attract at longer distances ( see Fig 2 and Materials and methods for details of the potential ) . The interactions between all cell types are the same , except in 2 cases . First , to simulate differential adhesion , the attraction is set to be weaker with and among PrE cells . Second , in contrast to previous models , the physical forces between TE cells are assumed to depend on cell polarity such that 2 TE cells adhere to each other when their polarity is pointing in the same direction , and cells are positioned next to each other in the plane perpendicular to the polarity axis . Biologically , this would correspond to a well-known phenomenon of tight junctions forming in a plane perpendicular to the polarity axis [66] . The modeled blastocyst grows as cells divide . Cell division is simulated by selecting a cell at random and introducing a daughter cell between the mother and its nearest neighbor . In case of TE , 1 daughter cell inherits polarity including the orientation of the polarity from the mother cell . To conceptually capture the 4 major themes outlined in the Introduction , we have formulated the following 4 rules: At E3 . 0 , we define the outer cells by counting the number of nearest neighbors ( shown in Fig 2a ) . Cells with fewer than 5 nearest neighbors are assigned TE fate and polarity , pointing radially outwards from the center of the cell mass . We do not aim at recapitulating how the polarity is established and how inner and outer cells are defined in a real embryo . Instead , we focus on the role of polarity in outer cells after it has been established . To take into account that TE cells are about twice as big as the rest of the cells , each TE cell is simulated by 2 unit circles . Polarity of the TE cells is assumed to lead to polar interactions that can be thought of as tight junctions forming in the plane perpendicular to the TE polarity . We simulate this by a polarity-dependent attraction factor , S ( see Eq 1 ) , that is maximal for 2 neighboring cells if their polarity is oriented in the same direction and perpendicular to the position vector ( illustrated in Fig 2b ) . At every time point , the strength of attraction , S , between 2 neighbor TE cells is given by: S=−1 . 4 ( e^1 × r^12 ) ⋅ ( e^2 × r^21 ) ( 1 ) Where ê is the polarity unit vector of a TE cell , and r^ is the unit distance vector between 2 neighbor TE cells . The prefactor of 1 . 4 assures that TE cells are tightly packed but nonoverlapping . Notice that this polar interaction favors the formation of a single-layer sheet with cells positioned perpendicular to the polarity axis and disfavors compact structures ( Fig 2b ) . The dynamics of the polarity vectors is governed by simple damped dynamic equations as in Eq 3 ( see Materials and methods ) . Proliferation of TE was reported to be about twice as fast compared to ICM [67] , so we set the rate of TE cell division to be 2-fold that of ICM cells . In all lineages , daughter cells inherit the mother’s fate and polarity . From E3 . 0 to E4 . 0 , the FGF signaling pathway becomes important for lineage segregation in the ICM . Bessonnard et al . [57] suggests that the FGF/ERK pathway coupled with the intracellular mutual inhibition between Nanog and Gata6 act together to ensure the fidelity of initial EPI and PrE specification . The simplified logic behind this process can be reduced to the intracellular inhibition and extracellular activation between Nanog and Gata6 ( shown in the network in Fig 1 at E3 . 5 ) : The mutual inhibition between Nanog and Gata6 is , in effect , an intracellular positive feedback loop . When reduced to 1 variable ( e . g . , Nanog ) , the network reveals a combination of intracellular amplifying positive feedback with extracellular inhibition of Nanog in neighboring cells . This representation suggests a Turing mechanism that results in both local amplification and global inhibition ( see S2 Fig ) . Simulations based on this Turing mechanism predict that ICMs would maintain the ratio of PrE ( or EPI ) /ICM , irrespective of embryo size . At E3 . 0 , all ICM cells are in an “undetermined state” coexpressing low levels of both Gata6 and Nanog ( Fig 1 ) , and the cell specification process is started as all these cells begin to express FGF4 [40 , 69] . This initial step follows the same simplified logic outlined above , as long as we assume that once specified , Gata6 cells have a lower concentration of FGF4 in their neighborhood than undetermined ICM cells . In the model , we implement this logic by monitoring the number of nearest-neighbor cells with high FGF4 ( EPI or undetermined ICM ) versus low FGF4 ( PrE ) ( as in Fig 2a ) . At cell division , the likelihood of a mother and a daughter cell to convert to PrE is proportional to the fraction of high FGF4 ( EPI or undetermined ICM ) cells in the neighborhood ( P ( PrE ) =# of high FGF4 neighbors# of ICM neighbors ) , and , conversely , the likelihood to convert to EPI is P ( EPI ) = 1 − P ( PrE ) . As a result of this simple rule , in our simulations , cells undergo the specification into salt-and-pepper pattern of Nanog/Gata6 cells from the initial state of unspecified ICM . In addition , all cells can potentially convert their identity between PrE and EPI later ( E3 . 5 to E4 . 0 ) , as the blastocyst grows . Although the identity of cells can be modulated in ex vivo blastocyst culture in response to FGF treatment or inhibition [27] , once cell identity is established , conversion is quite rare in unmanipulated culture conditions [70] . In our simulations , the rate of conversion is low and asymmetric , which is in line with observations by Xenopoulos et al . [70] . Differential adhesion is activated once the cells specify their identity ( at E3 . 5 ) . It is implemented by a single change in the attraction factor for PrE precursors from S = 0 . 6 to S = 0 . 4 . Biologically , this corresponds to lower adhesive properties of PrE cells . Thus , the attraction factor between 2 EPI , 2 TE cells , or an EPI cell and an ICM or TE cell remains at S = 0 . 6 , while the attraction factor , S , between 2 PrE or PrE and any other cell type is reduced to S = 0 . 4 . These potentials are shown in Fig 2c . TE cells only interact with their nearest neighbors ( i ) . Limiting the range of the TE—ICM potential ( ii ) to about 2 cell diameters allows the model to capture the symmetry breaking event ( at E3 . 5 ) , with ICM and cavity forming at the opposite sides . ICM cells are assumed to interact with all the other ICM cells ( e . g . , by protrusions ) , which is implemented by a global potential without explicit cutoffs ( iii ) . As a result , PrE precursors migrate away from the EPI core and form the PrE layer at the surface of the ICM , facing the cavity . With this rule , the model predicts that if the TE were removed at the blastocyst stage , in the isolated ICM , the EPI would end up surrounded by PrE ( S1 Fig and S3 Movie ) . This is consistent with the experimental observation that in the cluster of mixed EPI/PrE cells , PrE cells migrate to the outer layer surrounding the EPI core [43] . At E4 . 5 , PrE precursors in the EPI core , i . e . , cells surrounded by more than 3 non-PrE precursors , undergo apoptosis to ensure that failures in lineage segregation are not incorporated into EPI development . To compare with experimental results from 3D blastocysts , we used simple scaling relationships , converting between 2D and 3D ( see Materials and methods ) . We have also validated our approach in 3D simulations ( S2 Movie ) . All the major steps were the same as in 2D , with 1 modification: In 2D , TE cells would always have 2 nearest TE neighbors . We identify these 2 TE cells as nearest neighbors if they are within a certain distance . However , in 3D , this approach fails as one may obtain cell centers within a cell diameter that are not nearest but next-nearest neighbors . To account for this and to find the list of “true” nearest neighbors , we have developed a method that separates nearest from next-nearest neighbors . We evaluate if a potential nearest neighbor is closest to the given cell—and thus included as its true nearest neighbor—or if it is closer to another cell in the neighborhood and thus not counted as nearest neighbor ( see Materials and methods for details ) . While this neighborhood assignment is necessary for the stability of the TE in the 3D model , it is not sufficient , as without polarity the TE cells would collapse into a clump , and cavity cannot be formed . In order to quantify the importance of each of the rules , we specified the “successful” configuration of the blastocyst at E4 . 5 to be the one in which ( i ) TE cells are segregated from ICM cells and form a shell; ( ii ) the cavity is formed , and the ICM is positioned at one side of the cavity; ( iii ) ICM cells segregate in 2 distinct layers with the PrE positioned between the cavity and the EPI cells; ( iv ) and no isolated EPI cells are in the PrE layer nor isolated PrE in the EPI . By comparing the outcome of our simulations to the criteria above , we quantified the fraction ( out of 200 simulations for each condition ) of “successful” in silico blastocysts ( Fig 3b ) . Representative screenshots from a successful simulation are shown in Fig 3a and in S1 Movie . We found that with all 4 rules in place , the success rate is high ( 79% , Fig 3b , see also S2 Movie ) , suggesting that these rules together are sufficient for development of blastocysts up to E4 . 5 . We also challenged these 4 rules in 3D simulations and found that they were sufficient to generate 3D blastocysts ( S2 Movie ) . For the sake of simplicity , we will compare the impact of specific perturbations in these rules using 2D simulations . In 2D simulations , the fraction of ICM/total cells was 39 ± 12% , and the EPI/ICM fraction was 44 ± 18% , both of which are in a good agreement with experimental data [48 , 51 , 53] ( Fig 3c and 3d ) . Of the 21% failure in our simulations , about 20% occurred when—as a result of stochastic update—the fraction of ICM cells ( ICM/total cell number ) was low . As a result , there were not enough cells available to close the PrE layer , resulting in a PrE error . In about 1% of the cases , the TE broke , either due to failure of maintaining contacts between the surrounding TE cells or right after polarities have been added at E3 . 0 . This occurs if the embryo is in a “tight configuration” in which adding another ICM cell disrupts the shell of TE cells . This error , we believe , is attributed to our choice of potential and noise parameters and might happen even more rarely if the parameters are fine-tuned . In successful cases , embryos transited through a salt-and-pepper pattern , eventually separating PrE from EPI . To what extent this pattern is salt-and-pepper , i . e . , how big are the regions with the same cell types , depends on several factors . The longer the range of FGF4 signal , the larger are the patches of the same cells; on the other hand , the size of the patches is also increased if the differential adhesion molecules are expressed at the same time as cell specify ( as is assumed in our model ) . While visually we do observe patches of different sizes in published data , validation of this aspect of our model will require single-cell quantification of 3D imaging . To quantify the role of polar interaction , we “switched off” the polarity by setting the attraction factor for all cells to be the same as for undetermined ICM cells ( S = 0 . 6 ) . Without polarity ( ΔPolarity case ) all cells clustered together; consequently , there was no cavitation and no characteristic shell-like layer of TE cells forming ( Fig 3b ) . These results agree well with the observations in mouse mutants and knockdowns targeting polarity pathways: homozygous mutation in downstream regulator of Yap/Taz signaling , Tead4 [60 , 62]; chemical inhibition of RHO-ROCK signaling ( required for apical-basal polarity ) , knockdown of Pard6b ( a component of PAR-aPKC ) by RNA interference ( RNAi ) , disturbing the apical complex aPKC/PAR6 by small interfering RNA ( siRNA ) , downregulating aPKC/PAR3 by injecting double-stranded RNA ( dsRNA ) , or Prickle2 mutants [72–76]—all result in severe polarity defects ( including the absence of tight junctions ) , and all fail to form blastocoel . Elimination of the second rule can be carried out by modulating the FGF concentration either up or down . As expected , low FGF concentration ( −FGF in Fig 1 and S3 Movie ) in our model resulted in no PrE formation and an ICM consisting of only EPI at E4 . 5 . These cells were found in a clump consisting of several layers in one side of the blastocyst . This spatial configuration of EPI cells is in agreement with the observed FGF4- and FGFr2- mutants [40 , 77–80] . Also as expected , the maintenance of a constant FGF/ERK on state ( +FGF in Fig 1 and S4 Movie ) resulted in ICMs composed solely of PrE , consistent with the experimental results from introducing an excess amount of FGF [27 , 40 , 41] ( Fig 3d ) . As a result of stronger adhesion between EPI cells compared to adhesion between PrE cells , the ICM cells clump more in the “EPI only” ( low FGF ) case compared to the “PrE only” ( high FGF ) case . The clumping of the “PrE only” cells is in disagreement with the experimental observation by Yamanaka et al . [27] in which PrE cells are positioned on one side of the blastocyst in 1 layer lining the TE . This disagreement is likely because in our model , the difference between PrE and EPI cells is limited to differences in adhesive properties and does not include the reported apical-basal polarity of the PrE cells [18] . Adding polar interactions to the PrE layer in our model will disfavor “clumping” and make PrE cells line along the TE layer . While polarity of the PrE may add to the robustness of the blastocyst patterning , we chose not to include it into the current model as , within the criteria for success we specified , it does not seem to be necessary for the successful development of the “wild-type blastocyst . ” Noticeably , the failure rate is close to 100% when the differential adhesion between the Gata6 and Nanog positive cells is neutralized ( by setting S = 0 . 5 for all the ICM cells ) ( Fig 3b , ΔDifferential adhesion case , see also S5 Movie ) . At E4 . 5 , PrE progenitors remained distributed in a salt-and-pepper pattern; consequently , a considerably higher fraction of the PrE progenitors underwent apoptosis . The ICM/total cell fraction in this case was 39 ± 11% , and the EPI/ICM fraction increased to 57 ± 17% . Deletion of a number of adhesion molecules is known to produce failures in PrE and EPI segregation [47 , 81–84] . Inhibition of the polarity determinant aPKC [53] at the mid-blastocyst stage results in a failure of PrE/EPI segregation; the increase in inappropriately localized Gata6 cells results in an increased rate of apoptosis within this population , leading to a PrE:EPI ratio of 1:1 , which is within the uncertainty of our results ( Fig 3 ) . Deletion of the fourth rule ( ΔApoptosis ) resulted in 13% of embryos with a PrE precursor positioned deep within the ICM ( S7 Movie ) . As not only misplaced PrE are likely to undergo apoptosis [42] , we tested and found no significant differences in our results when we included up to 20% apoptosis in EPI cells ( see S5 Fig ) . Despite differential adhesion and letting the system reach the equilibrium configuration , those cells were trapped in a local energy minima and could not move towards the cavity . The number of mispositioned PrE cells and , consequently , the rate of apoptosis became higher if the system did not reach equilibrium . As it is not known if the ICM cells are in equilibrium or not , our results suggest that 15% error is the lower bound estimate of how frequently differential adhesion fails to segregate PrE from EPI . To further validate our model , we asked if it can reproduce results of classical scaling experiments [6 , 67 , 71 , 85 , 86] in which single cells from the 2-cell stage embryo were shown to develop into blastocysts , albeit of half the size and at a lower success rate . Dividing the embryo in half at any time point up to the 8-cell stage resulted in “successful” embryos at E4 . 5 in about 59% of cases ( Fig 4a and S8 Movie ) . Furthermore , halved embryos were 50% smaller ( 66 ± 3 cells ) than the unperturbed ones ( 132 ± 3 cells ) . We also observed a 20% increase in failure rate in blastocyst formation . In our simulations , that was predominantly due to PrE error as a result of a smaller ICM and the resulting fluctuations in the ratio of PrE to EPI . In cases in which the PrE/EPI ratio is smaller than in the unperturbed embryo , there are too few PrE cells to form a layer lining EPI core , and EPI cells tend to intercalate into the PrE layer resulting in a PrE error ( Fig 4b ) . Our model also predicts that the rate of apoptosis in successful half embryos will increase , as abolishing the apoptosis rule increases the number of failures from 15% ( Fig 3b ) to about 23% ( Fig 4b ) . This is related to the increase in configurations with PrE error discussed above . The model is also consistent with the recently reported scaling results from aggregating two 8-cell stage embryos [71] , see Fig 5 and S9 Movie . Thus , without any parameter adjustment , our in silico results were in complete agreement with the scaling experiments and allow us to ask which of our rules is responsible for the scaling properties of the mammalian blastocyst . As 3 out of the 4 rules ( Rules 2 , 3 , and 4 ) are conditional on FGF/ERK signaling—apoptosis of PrE and differential adhesion are only possible once ICM differentiated in PrE and EPI—we asked whether stage-specific signaling competence could account for scaling . In these in silico scaling experiments , we have kept the timing of ERK activation unchanged ( at E3 . 0 ) , which would mean that the timing of ERK activation is set at fertilization , either based on expression of the receptor or a rate-limiting factor in the pathway . When we moved ERK activation forward in time , which in effect means delaying the salt-and-pepper pattern , the blastocyst did not fully resolve the salt-and-pepper pattern by E4 . 5 . The model also predicted that delaying ERK signaling should decrease the fraction of PrE cells ( See Fig 3D “Delay FGF4” and S10 Movie ) . To validate this experimentally , we have performed embryo aggregation experiments with and without a potent inhibitor of Mek ( PD0325901 ) [36 , 87] , henceforth referred to as Meki , the kinase that responds to FGFR activation and phosphorylates ERK ( Fig 6 ) . Because of the high level of inherent stochasticity in the model and experiments , we chose to prioritize statistically significant results . The identification of a double positive ( DP ) fraction ( through k-mean clustering , see Materials and methods ) showed large fluctuations between repeat experiments; the fraction of PrE on the other hand was very robust , so we decided to focus on quantifying this cell type as an indicator ICM patterning . To set the time of FGF/ERK competence , we cultured embryos in Meki for 24 hours and then released them from the signaling block . In line with the model predictions , we did observe a decrease in fraction with PrE cells proportional to the duration of the ERK inhibition . Embryos were cultured for 56 hours following manipulation , to an in vitro equivalent of E4 . 5 . During this time window , exposure to FGF/ERK signaling was manipulated in 24-hour intervals . Complete inhibition of Mek for the entire 48-hour period resulted in embryos that were entirely EPI ( Figs 6b , 6c and 7b ) , and this is consistent with previous observations [27 , 71] . However , when embryos were treated for 24 hours ( E2 . 5–E3 . 5 ) with Meki and then released from the block , PrE cells were partially recovered , but their fraction was significantly smaller than in the untreated case . Similar transient inactivation experiments have produced a variety of results [27 , 71] that generally support this observation but without statistical analyses of single-cell quantitation . We also found that the duration of FGF4/ERK activation , or the point in time in which the pathway becomes competent for signaling , delimits the capacity of the ICM lineages to scale ( Figs 6 and 7 ) . Quantification of the relative level of PrE induction ( Fig 6B and S4 Fig ) indicates that normal aggregates maintain constant ratios of EPI/PrE and that delaying ERK activation with Meki resulted in a reduction in PrE specification . When normal and aggregated embryos are pooled , the quantitative reduction in PrE specification was statistically significant by nonparametric rank-sum test . In addition , we observed , that based on total cell numbers , the embryos scaled but not in a perfectly linear fashion . We found that the size of the aggregated embryos increased significantly ( Fig 7a ) , and they contained correctly proportioned ICMs , although we did observe a significantly higher ratio of TE to ICM in triplets ( Fig 7c ) . We believe this could be a result of small differences in the relative number of founder TE cells in the triple aggregates , which proliferate at twice the rate of ICM cells [67] and could amplify these differences 2 days later . We also noticed that our embryos contain slightly lower cell numbers than the recently reported aggregation experiments [71] , but we imagine these difference could be due to strain differences—in this study , we used inbred C57BL/6 , whereas Saiz et al . [71] used the outbred CD1 strain . Taken together , our attempts to delay ERK activation with Meki combined with embryo scaling experiments suggest that the accumulation of FGF4 and/or an important , limiting downstream signaling component accumulates from fertilization to the point at which this pathway can be activated . Notably , the fixed timing of cavitation , reported by Korotkevich et al . [88] , suggests that the timing of TE/ICM differentiation may also start at fertilization . In our model , the timing of TE/ICM differentiation is flexible and only requires a defined time point when polarity is defined .
Without polar interactions we could not form the cavity . These results are in a good agreement with the experiments reporting on the consequences of strong polarity defects [72–76] . In earlier in silico models , which did not include polarity , the cavity was introduced by hand and was assumed to grow and create a positive pressure on TE thus driving their cell division [59 , 58] . In contrast , in our model , the growing cavity is a consequence of dividing TE cells , which form a shell-like layer due to polar interactions . While our results do not rule out the osmotic expansion of the blastocyst , they argue that the expansion by TE proliferation should be considered on equal footing . It is hard to delineate which of the 2 is a driving mechanism as the 2 are tightly coupled . First , even if the blastocyst expands by TE proliferation , the cavity’s osmotic pressure should be maintained at homeostasis . Second , drugs inhibiting TE ion channels do not solely act to decrease blastocyst expansion but also perturb TE metabolism [91] and proliferation . To maintain homeostasis during blastocyst expansion , it is likely that the 2 mechanisms act in tandem , feeding back on each other . When seen from the perspective of one of the markers , e . g . , Nanog in EPI cells , the reduced scheme of FGF/ERK signaling ( Fig 2 and S2 Fig ) can be generalized to a Turing-like patterning mechanism . This mechanism is known for patterning of animal fur , e . g . , emergence of black spots in leopards , and is often summarized as “local amplification and global inhibition . ” In the case of ICM cells , the “local amplification” results from Nanog intracellular positive feedbacks , whereas the “global inhibition” is realized by Nanog cells secreting FGF4 and inhibiting Nanog in neighboring cells ( S2 Fig ) . Thus , as long as FGF4 is produced in predetermined ICM cells , the pattern of intermixed cell types will automatically emerge . However , when we sampled early stages of blastocyst formation , we found that the patterning of the early ICM ( at E3 . 5 ) was similar , but not identical , to the recently reported experiments on quantification found in Saiz et al . , 2016 . In particular , we never detected EPI cells induced in the absence of PrE cells ( S6A Fig ) . Although recent single-cell RNA sequencing data [39 , 92] suggests that undetermined ICM cells do express FGF4 , they appear to do so to a lesser extent than EPI cells , and this was not accounted for in our original model . We therefore decided to test how this may influence our modeling results; we modified the model to make undetermined ICM cells contribute half as much FGF4 as determined EPI cells . We found that the fractions of cells at E4 . 5 did not change ( S6 Fig ) , and that we generated normal blastocysts , indistinguishable from our previous simulations . However , this slight modification now recovers the subpopulation of embryos with EPI but no PrE at E3 . 5 reported by [71] ( S6A Fig ) . Thus , our model not only generated correctly patterned blastocysts , but now also reproduces the earliest phases of lineage segregation with higher fidelity . These simulations make a new , experimentally verifiable prediction that unsegregated ICM cells express lower , but functional , levels of FGF4 than differentiated EPI . It also further demonstrates the importance of self-regulating dynamics in patterning the blastocyst , demonstrating that the final result is not sensitive to initial conditions . The capacity to vary initial conditions without impacting on the final result also provides insight into parameters that can be manipulated in evolution . The FGF/ERK pathway coupled to Nanog/Gata positive feedback was proposed to control PrE/EPI cell proportions in 2 other models , one exploring isolated ICM patterning [57] and the other PrE/EPI specification in embryonic stem cells ( ESCs ) [93] . Neither model captures the dynamic geometry of the growing embryo . Our model incorporates a conceptualized FGF/Nanog/Gata feedback circuit into embryonic development , showing that a form of this mechanism can function in the highly dynamic environment in which cells divide and move due to differential adhesion . As a consequence of this “local amplification , global inhibition , ” PrE and EPI cells in the model are capable of changing their identity at all times during the FGF/ERK competence window , i . e . , the pathway remains active . While in ES cell culture and in FGF4 manipulation experiments by Bessonnard et al . [57] and Yamanaka et al . [27] there is a window in which cells are observed to change their identity , this does not seem to occur frequently under physiological levels of FGF4 in unperturbed blastocysts [70] . It is , however , not known if these changes do not occur because cells are not capable of switching or because they reach an appropriate configuration in which the switching is not required . Our simulation suggest that the latter explanation is correct , and this explains why the cells of the blastocysts remain competent to undergo regulative transformations in response to signaling manipulation while maintaining an apparently deterministic trajectory in normal development . This , and to what extent the number of FGF4-secreting neighbors determines the fate of the cell , can be tested through targeted laser ablation of cells such as to shift the balance between FGF4-secreting and nonsecreting cells in the neighborhood . Simulation results suggest that differential adhesion alone can often ( 62% of embryos ) be sufficient for correct spatial arrangement of PrE and EPI cells . It is believed that the position-dependent apoptosis of Gata6 cells may play an important role in resolving the occasional positional errors: Plusa et al . [4] reports that the isolated Gata6 cells deep inside the ICM apoptose 6-fold more often than the correctly positioned PrE cells facing the blastocoel . One possible mechanism for the positional difference in apoptosis is if the concentrations of the cytokines LIF [50 , 51] and PDGF [52 , 53]—known to promote PrE survival—are lower inside the ICM than at the junction of the EPI , PrE , and blastocoel . While this still remains to be tested experimentally , the current knowledge on LIF is in line with this hypothesis . LIF secreted from TE is likely to accumulate to higher concentrations at the PrE/blastocoel boundary as there are more TE cells facing the blastocoel , and LIF produced by these cells can diffuse freely until it reaches the PrE layer . Our observations indicate apoptosis is important for the robustness of pattern formation . The model predicts that apoptosis becomes increasingly important as the difference between adhesive properties is reduced . The difference in adhesion properties could vary in different genetic backgrounds and also must vary in time as differentiation progresses . In cases in which there is flexibility in adhesion , there would be a greater requirement for apoptosis in proofreading . While we have shown that differential adhesion in combination with apoptosis are sufficient for proper lineage segregation , a number of other mechanisms may contribute to robustness in the segregation process . First , it has been suggested that cellular movements involve not only passive but also active mechanisms , associated with cell protrusions [42] . In line with this , in silico studies that did not consider apoptosis , showed that failures in segregation can be reduced if differential adhesion is complemented by directed cell movements [59] . Second , PrE differentiation occurs in stages that include an uncommitted and biased state within the ICM and committed , PrE lining the blastocoel . Thus , when tested in heterotrophic grafting experiments , early PrE progenitors within the ICM were competent to make EPI , while PrE progenitors that line the blastocoel cavity are only able to participate in endoderm development [29] . In our simulations—in which the segregation of the PrE and EPI is based on differential adhesion—we often , in about 50% of simulations for wild type , observe PrE forming 2 layers . While this is obviously in contrast with the observed single PrE layer in real embryos , to keep the model simple , we choose to count them as a success as long as the PrE layers seal EPI core from blastocoel . Expanding the model to include polar interactions between PrE cells lining the cavity would ensure a single PrE layer and provide a contiguous barrier between the EPI and the cavity [53] . Scaling presents a fascinating example of the robustness in embryo development , and the experimental manipulation of this phenomena served as an important validation step for the model . The close match between the experimental observations and our simulation predicted that timing of FGF/ERK signaling may be the key parameter for controlling the scaling outcome . With the 4 rules in place , the embryo would scale when divided in half or doubled prior to compaction . However , we only observe this property if cell fate specification and emergence of the salt-and-pepper pattern—attributed to the FGF/ERK signaling—take place at the same time counted from fertilization . This implies that competence for FGF/ERK signaling is primed for activation from fertilization . We have validated this prediction by showing that the ratio of PrE in the ICM decreases upon transient inhibition of FGF4 signaling both in the model and in cultured embryos . The notion that we observe normal development with 4 rules that are largely independent of the initial gene regulatory network is particularly relevant to the current debate about the extent to which stochastic gene expression governs the initiation of blastocyst development . Our model demonstrates that initial differences in stochastic gene expression are not a necessary prerequisite for the generation of 3 distinct lineages . Instead , differentiation emerges based on the responses of a cell to its local environment , as interpreted via differential proliferation , adhesion , and gene expression . The existence of a set of rules that allow for blastocyst formation as long as a few simple conditions are satisfied could be an enabler of stochastic variation . It also could explain how mammalian development can allow for the fundamental changes in the gene regulatory network that have been observed when single-cell sequencing data has been compared between mouse and human [89 , 94 , 95] .
To compare with experimental results from 3D blastocysts , we used simple scaling relationships converting between 2D and 3D . Thus , the number of TE cells , NTE3D , placed on the surface of the sphere would correspond to NTE2D=πNTE3D in 2D . Similarly , for cells in the bulk: NEPI/PrE2D= ( 34NEPI/PrE3Dπ ) 2/3 . The interaction between 2 cells is given by the following potential , V ( see Fig 2c ) : V ( d ) =exp ( −d ) −S exp ( −d/β ) ( 2 ) where d is the distance between the cells , S is the attraction factor given by Eq 1 , and β is the parameter controlling the range of the attraction . This potential assures repulsion at short distances , i . e . , 2 cells separated by a distance less than 1 cell diameter ( d < 2 ) will repulse from each other . On the other hand , if cells are separated by more than 1 cell diameter ( d > 2 ) , they will be attracted to each other . In the simulations , we set a distance cutoff , setting potential to 0 for all cell pairs that are further away than 5 cell radii . S and β are chosen to produce a tight packing of cells with minimal overlap ( see configuration of cells in Fig 2a , with S = 0 . 6 and β = 5 ) . The choice of these parameters as well as the form of the potential are not important for the model outcome as long as the condition above is satisfied . While we keep β = 5 fixed throughout the simulations , the S will capture lineage-specific differences in adhesive properties and is thus a lineage-specific parameter . Prior to first cell-lineage decision at E3 . 0 , cells are assumed to have the same adhesive properties and thus the same strength of attraction , S = 0 . 6 . The motion of the cells is described by the overdamped equation of motion: dx/dt=−dV/dx+η ( 3 ) where dV/dx is a x-projection of the resulting force from all of the pairwise cell—cell interactions . To ensure that system reaches equilibrium , we add the noise term η . In the simulations , this is implemented by adding a random number from a normal distribution with a mean of 0 and a standard deviation of 10−3 at every time step . The equation is integrated numerically using Euler integration scheme . The y-position is determined in a similar way . The polarity of the TE cells is assumed to be affected by the orientation of the polarities in neighboring TE cells such that they tend to point in the same direction . This is well described by a pairwise polarity potential Vp = −cos θi , j , where θi , j = αi − αj measures the angle between the polarities of the i , j TE neighbors . The orientation of the polarity is described by an angle α and , similarly to the equation above , the change in polar orientation is given by dαdt=−0 . 1dVpdα+ηp where the prefactor of 0 . 1 makes the changes in polarities happen slower compared to the changes in the positions . This is necessary for the stability of the system . The noise term is implemented in the same way as above , with the only difference being that the random numbers are multiplied by π . The TE is noise sensitive . With the chosen standard deviation 10−3 on the noise parameter , the simulation is very sensitive to the factor 1 . 4 in Eq 1 . If this factor is increased to 1 . 5 , the TE repulses too much when a new TE cell is added . On the other hand , if the factor is decreased to 1 . 3 , it becomes too weak and cannot keep the TE together at E4 . 5 . Thus , here , we apply the maximum tolerated noise to the system . Less or even no noise is also acceptable , and it allows the factor in Eq 1 to be decreased . We started with 1 cell . In contrast to an earlier model where the growth of the blastocysts was driven by the growing blastocoel , in our simulations , the blastocyst grows as a result of cell division: A cell is randomly selected to undergo division , and the daughter cell is positioned between the mother cell and the nearest neighbor cell . In real blastocysts , prior to division , cells gradually increase in size , allowing other cells to readjust their position such that the system is near equilibrium at all times . For simplicity , we chose to keep cell size constant; however , that results in strong perturbation of the equilibrium during the simulated cell division . To assure that the configurations of the simulated blastocysts are not affected by this , we allowed enough time for the system to relax before the next cell division . In 3D , one can significantly speed up the simulations by introducing a new cell in the center of the 3 nearest neighbors , as this is closer to the minimal energetic configuration and reduces the number of relaxation steps . During implementation of Rule 2 , we first pick up random cells to divide among the undetermined ICM’s . At the division , the likelihood of a cell to convert to PrE is proportional to the fraction of high FGF4 ( EPI or undetermined ICM ) cells in the neighborhood ( P ( PrE ) =# of high FGF4 neighbors# of ICM neighbors ) ; conversely , the likelihood to convert to EPI is P ( EPI ) = 1 − P ( PrE ) . After all cells have specified , a random ICM is chosen to divide , and at the division , the same rule applies as for undetermined ICM’s . Embryos used in this study are inbred C57Bl/6NRj ( Janvier Labs , France ) . Mice were maintained in a 12-hour light/dark cycle in the designated facilities at the University of Copenhagen , Denmark . Embryo donor females underwent super-ovulation treatment following a standard protocol: intraperitoneal injection ( IP ) of 5 IU PMSG ( Sigma ) per female and IP injection of 5 IU hCG ( Chorulon , Intervet ) 47 hours later , followed by overnight mating with C57Bl/6NRj stud males . The following morning , females were monitored for copulation plug formation . Embryos were considered E0 . 5 on the day of plug detection . Animal work was carried in accordance with European legislation and was authorized by and carried out under Project License 2012-15-2934-00743 issued by the Danish Regulatory Authority . Embryos were obtained at 8-cell morula stage by washing E2 . 5 oviducts with M2 medium ( Sigma ) . In order to remove the zona pellucida , morulae were briefly incubated in Acid Tyrode’s solution ( Sigma ) at RT and then washed in M2 medium . To generate aggregates , embryos were placed in pairs or triplets in aggregation microwells made with an aggregation needle ( BLS ) on Petri dishes in KSOM medium ( LifeGlobal ) drops . Drops were overlaid with mineral oil ( Nidoil , Nidacon ) . Single embryos were placed alone as control . KSOM was supplemented with 0 . 1% BSA ( Sigma ) to avoid embryos adhering to the plastic . Embryos were cultured at 37°C , 5% CO2 and 90% relative humidity . For MEK inhibition treatment , 1 μM of PD 0325901 ( PZ0162 , Sigma ) was diluted into KSOM . Wild-type embryos were generated by culture in KSOM . 24-hour and 48-hour treated embryos were generated by culture in KSOM with PD 0325901 for 24 hours and 48 hours , respectively . The data were collected over 4 repeat experiments . Fifty-six hours after aggregation , embryos at E4 . 5 were fixed in 4% PFA solution for 15 minutes at room temperature . Afterwards the embryos were stained as previously described [96] . The primary antibodies used were: anti-Nanog ( eBioscience , 14–5761; 1:200 ) , anti-Cdx2 ( Biogenex , MU392A-UC; 1:200 ) , and anti-Gata6 ( R&D , AF1700; 1:100 ) . Embryos were imaged in an Attofluor chamber ( ThermoFisher ) on a 25-mm glass coverslip using 10x magnification on a Leica TCS SP8 confocal microscope . We used ImageJ to manually track positions of the nuclei in single cells . Positions were saved and intensities for each fluorescence channel at each position were processed by custom-built Matlab scripts ( available upon request ) . For each of the channels , we have used the mean intensity of the 5 x 5 x 3 voxel as a readout for single cell . To filter out the noise , cells with the Dapi intensity below 1 were removed . To differentiate between TE and ICM cells , for each of the cells , we ranked the intensities of Cdx2 , Nanog , and Gata6 . Cells where Cdx2 ranked first , were classified as TE cells . We validated that the identified TE cells localize to the periphery of the embryo . We classify the ICM cells as described in Saiz et al . [71]: First , we performed k-means clustering ( by Squared Euclidean distance metric , Matlab built in function ) on the log ( Gata6 ) and log ( Nanog ) into 3 clusters with 10 repetitions on all data pooled together . Second , we classified high Gata6 and high Nanog cells as DP cells , high Gata6 and low Nanog as PrE cells , and low Gata6 and high Nanog as EPI cells . See S3 Fig for the results of the clustering . As the distributions of the analyzed properties were clearly far from normal , we used nonparametric rank-sum test to estimate the statistical significance of the difference in medians .
|
The first 4 . 5 days of mammalian embryo development proceeds without maternal information and is remarkably robust to perturbations . For example , if an early embryo is cut in half , it produces 2 perfectly patterned , smaller embryos . Where does the information guiding this development come from ? Here , we explore this issue and ask whether a model composed of a simple set of rules governing cell behavior and cell—cell interactions produces in silico embryos . This agent-based computational model demonstrates that 4 rules , in which a cell makes decisions based on its neighbors to adopt polarity , make lineage choices , alter its adhesion , or die , can recapitulate blastocyst development in silico . By manipulating these rules , we could also recapitulate specific phenotypes at similar frequencies to those observed in vivo . One interesting prediction of our model is that the duration of cell—cell communication through fibroblast growth factor ( FGF ) signaling controls scaling of a region of the blastocyst , and we confirmed this experimentally . Taken together , our model specifies a set of rules that provide a framework for self-organization , and it is this self-organizing embryogenesis that may be an enabler of stochastic variation in evolution .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blastocysts",
"cell",
"physiology",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"cell",
"cycle",
"and",
"cell",
"division",
"regulatory",
"proteins",
"cell",
"processes",
"dna-binding",
"proteins",
"endocrine",
"physiology",
"cell",
"polarity",
"developmental",
"biology",
"growth",
"factors",
"transcription",
"factors",
"embryos",
"fibroblast",
"growth",
"factor",
"erk",
"signaling",
"cascade",
"embryology",
"proteins",
"endocrinology",
"gene",
"expression",
"biochemistry",
"signal",
"transduction",
"cell",
"biology",
"apoptosis",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cell",
"signaling",
"signaling",
"cascades"
] |
2017
|
Four simple rules that are sufficient to generate the mammalian blastocyst
|
Nitrogen-fixing rhizobia and arbuscular mycorrhizal fungi ( AMF ) form symbioses with plant roots and these are established by precise regulation of symbiont accommodation within host plant cells . In model legumes such as Lotus japonicus and Medicago truncatula , rhizobia enter into roots through an intracellular invasion system that depends on the formation of a root-hair infection thread ( IT ) . While IT-mediated intracellular rhizobia invasion is thought to be the most evolutionarily derived invasion system , some studies have indicated that a basal intercellular invasion system can replace it when some nodulation-related factors are genetically modified . In addition , intracellular rhizobia accommodation is suggested to have a similar mechanism as AMF accommodation . Nevertheless , our understanding of the underlying genetic mechanisms is incomplete . Here we identify a L . japonicus nodulation-deficient mutant , with a mutation in the LACK OF SYMBIONT ACCOMMODATION ( LAN ) gene , in which root-hair IT formation is strongly reduced , but intercellular rhizobial invasion eventually results in functional nodule formation . LjLAN encodes a protein that is homologous to Arabidopsis MEDIATOR 2/29/32 possibly acting as a subunit of a Mediator complex , a multiprotein complex required for gene transcription . We also show that LjLAN acts in parallel with a signaling pathway including LjCYCLOPS . In addition , the lan mutation drastically reduces the colonization levels of AMF . Taken together , our data provide a new factor that has a common role in symbiont accommodation process during root nodule and AM symbiosis .
Legumes can establish a symbiotic association with nitrogen-fixing bacteria through the formation of symbiotic root nodules . Nodulation is initiated by the rhizobia-derived lipo-chitooligosaccharidic nodulation ( Nod ) factors that trigger transient increases in calcium influx levels accompanied with calcium oscillation in the rhizobia-attached root hair cells , initiating dedifferentiation of the underlying cortical cells [1–3] . Studies using two model legumes , Lotus japonicus and Medicago truncatula , have revealed the basically conserved molecular mechanism that results in the progress of Nod factor signaling . In L . japonicus , Nod factor is recognized by two LysM receptor-like kinases NOD FACTOR RECEPTOR 1 ( LjNFR1 ) and LjNFR5 [4–6] , which induce a downstream signaling cascade . The Nod factor signaling pathway includes SYMBIOSIS RECEPTOR-LIKE KINASE ( LjSYMRK ) , nucleoporins and cation channel proteins [7–11] . While loss-of-function mutations in components of the signaling pathway confer a complete nodulation deficiency phenotype , recent studies show that constitutive activation of either of LjNFR1 , LjNFR5 or LjSYMRK can induce spontaneous nodule formation in the absence of rhizobia [12 , 13] . This indicates that at least these three kinases each possess a necessary and sufficient role for nodulation . Following the calcium oscillation , the L . japonicus CALCIUM CALMODULIN-DEPENDENT PROTEIN KINASE ( LjCCaMK ) /M . truncatula DOES NOT MAKE INFECTIONS 3 ( MtDMI3 ) phosphorylates the transcription factor ( TF ) LjCYCLOPS/M . truncatula INTERACTING PROTEIN OF DMI3 ( MtIPD3 ) [14–18] . Phosphorylated LjCYCLOPS then induces the L . japonicus RWP-RK type TF , NODULE INCEPTION ( LjNIN ) , by directly binding to its promoter region [14 , 19] . A number of nodulation-related genes now have been identified as direct targets of Lj/MtNIN , including genes encoding the NUCLEAR FACTOR ( NF ) -Y subunits [20] . Root cortical proliferation is induced by constitutive expression of either of phosphorylated LjCYCLOPS , LjNIN or LjNF-Y subunits in the absence of rhizobia [14 , 20 , 21] , indicating that induction of the LjCYCLOPS>LjNIN>LjNF-Y hierarchical transcription cascade is sufficient to initiate nodulation . Several data indicate that cytokinin signaling is another essential regulator of nodulation , and that Lj/MtNIN is a downstream component of the cytokinin signaling pathway , as indicated by findings that functional cytokinin receptor is required for rhizobia- and cytokinin-dependent Lj/MtNIN induction [22 , 23] . It was recently shown that MtNIN directly binds to the promoter region of the CYTOKININ RESPONSE 1 ( MtCRE1 ) gene encoding a cytokinin receptor and promotes its expression at root cortex [21 , 24] . This result indicates that there is a positive feedback loop between MtNIN and cytokinin signaling . In addition to its activating role in nodulation , in some contexts LjNIN can negatively regulate nodule organogenesis through direct activation of CLE-ROOT SIGNAL 1 ( LjCLE-RS1 ) and -RS2 that function as putative root-derived signals in long-distance inhibitory signaling of nodulation [23] . Accommodation of rhizobia within host cells is indispensable for the establishment of root nodule symbiosis; therefore , proliferating cortical cells need to be invaded by rhizobia at the appropriate time during nodulation . In L . japonicus and M . truncatula , the rhizobial invasion process starts from the tip of the root hair associated with root hair curing . Rhizobia invade proliferating cortical cells through a plant-derived intracellular tube-like structure called the infection thread ( IT ) , and are finally released into host cells by endocytosis [25–27] . The signaling cascade initiating nodule organogenesis is also essential for the rhizobial invasion process , because in most cases root-hair IT formation is severely retarded if key proteins in the signaling pathway are mutated . A recent study demonstrated that , in addition to Nod factor , rhizobia-derived exopolysaccharides have a crucial role in the rhizobial accommodation process via interactions with the EXOPOLYSACCHARIDE RECEPTOR 3 ( LjEPR3 ) , a LysM receptor-like kinase that is paralogous to LjNFR1 [28] . The Nod factor signaling seems to have a role to induce the LjEPR3 expression at the epidermis . Overall , one signaling pathway achieves two qualitatively and spatially different phenomena , that is , rhizobial root hair accommodation at the epidermis and nodule organogenesis at the cortex . Studies using an epidermal-specific expression system indicated that this can be explained by a difference in the tissue-specific requirements of the genes involved in Nod factor signaling [29 , 30] . In addition , cell-to-cell communication between the epidermis and cortex may be involved [31] . In terms of transcriptional regulation , a direct target of LjNIN , NODULATION PECTATE LYASE ( LjNPL ) has been implicated in the degradation of plant cell walls , and is required for normal root-hair IT formation [32] . Thus , LjNIN may participate in rhizobia accommodation through activation of genes relevant to root-hair IT formation , such as LjNPL . MtNF-Y subunits seem to be involved in rhizobial accommodation processes as well as nodule organogenesis . LjNIN can also directly induce LjEPR3 expression; the LjNIN>LjEPR3 cascade appears to control rhizobia infection process [33] . In particular , ETHYLENE RESPONSIVE FACTOR REQUIRED FOR NODULATION 1 ( MtERN1 ) that encodes a TF involved in root-hair IT formation together with its close homologue MtERN2 , was shown to be a direct target of MtNF-Y subunits [34–36] . Moreover , LjCYCLOPS has a role directly inducing LjERN1 expression [37] . In addition , recent studies show that epidermal cytokinin signaling appears to have a negative role in root hair IT formation [38–40] . Despite these advances in our understanding of the molecular mechanism of nodule organogenesis and the rhizobia accommodation process , our understanding of the mechanism remains incomplete , indicating that further components await discovery . Symbiosis between plants and arbuscular mycorrhizal fungi ( AMF ) is another widely observed plant-microbe mutual relationship known as AM symbiosis . The plant regulatory pathway for AM symbiosis has been shown to share some components , called common symbiosis pathway ( CSP ) genes , of its genetic pathway with root nodule symbiosis [3 , 41] . Based on current data , the role of CSP genes is thought to mostly relate to generate calcium signaling and make a read-out , which occurs commonly during the two symbioses . Both symbioses are strongly impaired by a mutation in the CSP genes such as LjSYMRK , LjCCaMK and LjCYCLOPS . In AM symbiosis the LjCCaMK-LjCYCLOPS module responds to calcium oscillation , transmitting a signal to the downstream pathway , that results in the formation of symbiotic organs such as the arbuscule . LjCYCLOPS/MtIPD3 physically interacts with Lj/MtDELLA to form the LjCCaMK/MtDMI3-LjCYCLOPS/MtIPD3-Lj/MtDELLA complex that directly induces the REDUCED ARBUSCULAR MYCORRHIZA 1 ( Lj/MtRAM1 ) GRAS-type TF during AM symbiosis , which is required for arbuscule branching [42–44] . In the present study , we identify a L . japonicus mutant with a mutation in the LACK OF SYMBIONT ACCOMMODATION ( LjLAN ) gene . Observations of rhizobia infection/invasion patterns together with nodulation foci show that in lan mutant a developmental program of nodulation proceeds in the absence of root-hair IT formation , where rhizobia enter into roots through an intercellular invasion system . The LjLAN gene encodes a protein that is putatively orthologous to Arabidopsis MEDIATOR 2/29/32 ( AtMED2/29/32 ) constituting a Mediator complex . Moreover , the lan mutation reduces symbiosis with AMF . These data suggests LjLAN acts as a putative transcriptional regulatory module required for the establishment of both root nodule and AM symbiosis .
To better understand the molecular mechanisms associated with the control of nodulation , we undertook a screen for nodulation-deficient mutants from EMS-treated L . japonicus wild-type ( WT ) MG-20 plants . From this screen we isolated a mutant with a mutation in the gene that we named lack of symbiont accommodation ( lan ) based on the nodulation-deficient phenotype . F1 plants derived from a cross between lan and the WT MG-20 parental line showed normal nodulation . In the F2 population , normal-nodulation and nodulation-deficient plants segregated in an approximately 3:1 ratio ( 58 normal-nodulation and 18 nodulation-deficient plants ) . Thus , the lan mutation is inherited as a recessive trait . In L . japonicus , mature nodules can be characterized by several morphological and physiological indicators , including nodule size , color , lenticel formation , and nitrogen fixation activity . In WT plants , formation of mature nodules was recognizable at the latest 14 days after inoculation of Mesorhizobium loti ( dai ) ( Fig 1A , 1C , 1E and 1F ) . In contrast , in the lan mutant , no mature nodules were formed at the corresponding stage ( Fig 1B and 1E ) . Formation of mature nodules could be observed at 21 dai , and their number gradually increased over time ( Fig 1D and 1E ) , although the number was consistently lower than WT . Analysis of acetylene reductase activity per plant showed that nodules formed on the mutant roots at a later stage , such as 35 dai , were comparable to those of WT ( Fig 1F ) . Therefore , in terms of nitrogen fixation activity , the mutant nodules formed at the stage appeared to be functional . In order to characterize the effect of the lan mutation on root-hair IT formation and early nodulation , we used two fluorescent-based markers to visualize infection and nodulation foci . A M . loti strain expressing DsRED was used to mark root-hair ITs . During nodule development , a preferential auxin response is observed in proliferating cortical cells and bulge of nodule primordia [45–47] . Thus , we tried to quantify the sites of nodulation foci ( cortical cells proliferation and nodule primordia ) based on the expression of a reporter gene under the control of auxin responsive element DR5 . To visualize the nodulation foci in lan mutant , we produced DR5:GFP-NLS/lan plants by crossing DR5:GFP-NLS/WT transgenic plants [45] with the lan plants . In DR5:GFP-NLS/WT plants , the formation of root-hair ITs was recognizable at 4 dai , and cortical cells located under some of the ITs started to proliferate ( Fig 2A–2D , 2M–2P , 2U and 2V ) . In contrast , root-hair ITs were barely observed in the DR5:GFP-NLS/lan plants during the corresponding time scale ( Fig 2E–2H and 2U ) . In the DR5:GFP-NLS/lan plants , although root-hair ITs were almost undetectable at all time points tested , we found some sites of auxin response , which implied cortical cell proliferation and the formation of nodulation foci ( Fig 2I–2L , 2Q–2T , 2U and 2V ) . The number of nodulation foci gradually increased over time after inoculation ( Fig 2V ) . In most cases , the occurrence of nodulation foci was accompanied with bright DsRED signals suggesting the accumulation of rhizobia at the surface of developing nodules . These results indicate that in the lan mutant the nodulation developmental program can be initiated in the absence of root-hair IT formation . Some mutants impaired in root-hair ITs formation tend to develop an excess number of small uninfected nodule primordia [48–50] . Even in the later nodulation stage such as 45 dai , the formation of such small uninfected nodule primordia were not observed in the DR5:GFP-NLS/lan plants ( S1A Fig ) . In addition , inoculation of M . loti nodC mutants , which could not synthesize functional Nod factors , did not result in making any nodules in the lan mutant as well as WT ( S1B Fig ) . Thus , the nodulation in the lan mutant depends on Nod factor signaling . During nodulation , a series of calcium oscillations , defined as calcium spiking , in responsive cells is induced in response to the rhizobia-derived Nod factor [51 , 52] . A normal calcium spiking pattern could be observed in the lan root hair cells following application of purified Nod factor ( S2 Fig ) , indicating that in the lan mutant , nodulation signaling upstream of the calcium spiking response is unaffected . In L . japonicus DR5:GFP-NLS/WT plants , rhizobia use the root-hair IT-mediated intracellular invasion system to enter into roots ( S3A and S3C Fig ) [53] . In DR5:GFP-NLS/lan plants , despite strongly impaired root-hair ITs formation ( Fig 2U ) , cortical cell proliferation is induced , which results in the formation of nitrogen-fixing nodules ( Figs 1E , 1F and 2V ) , raising the question of how rhizobia enter into roots in the lan mutant . The accumulation of rhizobia on the epidermis of nodule primordia suggested that rhizobia might enter developing nodules through intercellular invasion system as was previously reported in other L . japonicus mutants ( Fig 2Q–2T and S3B and S3D Fig ) [11 , 54] . Thus , in order to clarify rhizobial localization in nodules , we examined sections of nodules . In the mutant nodules , a dense population of rhizobia was observed in some intercellular spaces ( Fig 3A–3D ) . This bacteria localization pattern is reminiscent of that defined as pocket of intercellular bacteria seen in the several L . japonicus mutants , where rhizobia enter nodules predominantly through intercellular invasion system [11 , 50 , 54–56] . In WT nodules rhizobia enter nodule cells through cortical-ITs ( Fig 3C ) [56] . On the other hand , we could not determine the presence of cortical-ITs in the mutant nodules . An observation of lan mutant nodule sections of relatively later stage showed that the number of rhizobia-colonized cells were evidently reduced compared with WT ( Fig 3E and 3F ) . In WT nodules , rhizobia-colonized cells were tightly packed at the inner region of nodules ( Fig 3E ) . On the other hands , in the lan mutant , clusters of uninfected cells were located between rhizobia-colonized cells ( Fig 3F ) . Thus , the lan mutation can affect rhizobia accommodation process throughout nodule development . To understand the molecular function of LjLAN , we first sought to isolate the gene by a positional cloning approach . This mapped the LjLAN locus to a region between the simple sequence repeat ( SSR ) markers TM0216 and TM0135 on chromosome 3 ( S4 Fig ) . Subsequent genome-resequencing of the lan mutant identified an A-to-T nucleotide substitution that occurs in the acceptor site of an intron located upstream of the gene , chr3 . CM0112 . 280 . r2 . d ( S5 Fig ) . In the mutant , the nucleotide substitution causes the production of two transcripts smaller than that of WT ( Figs 4A and S5 ) . We sequenced each mutant transcript , and found that in both cases intron mis-splicing spliced out a DNA region encompassing the original initiation codon of the gene . In addition , in the lan mutant no coding sequence was predictable in the locus . Thus , it is reasonable to suppose that the lan mutation causes a complete loss of function of the gene . The mutant two transcripts were detectable all time points tested after inoculation ( S6A Fig ) . The lan mutation reduced the expression of the gene ( S6B Fig ) . To verify if this gene is responsible for the lan mutation , a 5 . 8-kb genomic fragment containing the WT gene was introduced into the mutant by Agrobacterium rhizogenes-mediated hairy root transformation . The introduction of the fragment into the mutant rescued the phenotype , resulting in the formation of normal number of nodules at 14 dai ( Fig 4B–4F ) , and normal root-hair ITs formation ( Fig 4C–4F ) . The LjLAN gene encodes an uncharacterized protein of 145 amino acids that is putatively orthologous to AtMED2/29/32 , a putative subunit of the Mediator complex ( S7 Fig ) . It is generally thought that the Mediator complex , which consists of a large number of subunits , plays a role as a bridge between promoter-bound TFs and RNA polymerase II to activate gene transcription [57–59] . Indeed AtMED2 was shown to be required for the recruitment of RNA polymerase II [60] . AtMED2 could rescue the lan mutation when it was constitutively expressed by LjUBQ promoter ( S8 Fig ) , suggesting that the LjLAN has a function similar to AtMED2 . The phylogenetic analysis identified a homologue of LjLAN in L . japonicus , which was designated as LjLAN LIKE ( S7 Fig ) ; the similarity and identity values are respectively 94 . 6% and 81 . 7% . The expression pattern of the LjLAN and LjLAN LIKE gene remained constant in some vegetative and reproductive organs investigated ( S9A and S9B Fig ) . To gain insights into the role of the LjLAN gene during nodulation , we examined the time course expression pattern after inoculation of M . loti . LjLAN expression was largely constant during nodulation when whole roots were assayed by RT-qPCR ( S9C Fig ) . However , an approximately 2-fold induction of LjLAN expression was detected in root segments where proliferating cortical cells were enriched ( S9C Fig ) . Furthermore , reporter gene analysis using ProLjLAN:GUS plus construct showed that during nodulation the GUS activity was detectable at epidermis with curled root hairs , proliferating cortical cells and nodule primordia ( S10 Fig ) . The GUS activity was also observed at lateral roots . The lan mutant used for above-mentioned analyses has Miyakojima MG-20 genetic background . We obtained a plant with Gifu B-129 genetic background in which a retrotransposon , LOTUS RETROTRANSPOSON 1 ( LORE1 ) [61 , 62] , was inserted in the middle region of coding sequence of LjLAN gene , causing an occurrence of premature stop codon in the mutant ( S11A and S11B Fig ) . Consequently , we found that the plants have the truncated protein of LjLAN lacking C-terminal part of it ( S11B Fig ) . Unexpectedly , the LORE1-tagged mutant showed normal nodulation phenotypes ( S11C and S11D Fig ) . In order to interpret the observation , we raised two possibilities . First , the effects of lan mutation was observable in an ecotype-specific manner . The second possibility was that the truncated LjLAN that was produced in the LORE1-tagged mutant was functional . To verify them , we introduced modified LjLAN ( LjLANΔC ) , in which amino acid residues constituting C-terminal part of LjLAN were deleted ( S11B Fig ) , into lan mutant . LjLANΔC could rescue the lan mutation to the extent of same level of the introduction of control intact LjLAN ( S8 Fig ) . We then created stable transgenic plants with nucleotide deletions or insertions in the middle region of coding sequence of LjLAN gene by CRISPR-Cas9 genome-editing system . In the transgenic plants , the frame-shifted mutations caused the deletion of amino acid residues constituting C-terminal part of LjLAN ( S11B and S11E Fig ) . The nodulation phenotypes of the transgenic plants were indistinguishable from WT plants ( S11F and S11G Fig ) . These results indicate that C-terminal part of LjLAN is not essential for the LjLAN function . Therefore , the lack of phenotype of the LORE1-tagged mutant can be explained by the retention of LjLAN function rather than an ecotype difference . The LORE1-tagged mutation did not affect the expression of LjLAN ( S6B Fig ) . After decoding calcium spiking followed by rhizobial infection , LjCYCLOPS has an important role in root nodule symbiosis , as it regulates both rhizobial infection and nodule organogenesis through induction of different downstream target genes [14 , 37] . cyclops mutants retain nodulation to some extent [63] , providing an accessible baseline for screen for second mutations influencing the cyclops nodulation defects . We then created lan cyclops double mutant . Of note , the lan cyclops double mutant plants showed a complete non-nodulating phenotype , different from each single mutant ( Fig 5 ) . To gain insight into the potential relationship between LjLAN and LjCYCLOPS with respect to gene expression , we investigated the two nodulation-related genes expression , LjNIN and LjNF-YA . LjNIN , a direct target of LjCYCLOPS , has a pivotal role in the transcriptional cascade that is required for both nodule formation and rhizobial infection [19 , 20] , and LjNF-YA has been shown to be a direct target of LjNIN [20] . Confirming previous reports , we found that expression of LjNIN and LjNF-YA was strongly induced throughout nodulation stages investigated ( Fig 6A and 6B ) [19 , 23 , 31] . We found that in the lan and cyclops mutants the induction level of LjNIN was consistently weaker than that in WT along the time course after inoculation ( Fig 6A ) . However , although the lan and cyclops mutation suppressed LjNF-YA induction at 1 and 7 dai , the induction level in lan and cyclops roots at 14 dai was largely comparable to that in WT roots of the corresponding stage ( Fig 6B ) . Furthermore , in the lan cyclops double mutant , the expression of the two genes were strongly impaired at all time point tested as well as ccamk mutant . Together with lan cyclops nodulation phenotype , these results indicate that LjLAN acts in parallel with LjCYCLOPS for the control of key nodulation-related genes expression . In order to clarify the potential impact of the LjLAN gene on the control of AM symbiosis , the lan mutant were inoculated with Rhizophagus irregularis . The level of AMF colonization of hyphae and arbuscules of the mutant at 21 dai was significantly lower in comparison with that in WT ( Fig 7A–7D ) . The lower level of AMF colonization was maintained even if the plants were grown for a long time such as 28 and 35 dai following inoculation with R . irregularis ( Fig 7A and 7B ) . In the lan mutant , R . irregularis tended to colonize in the lateral roots rather than primary roots ( Fig 7E ) . The introduction of WT LjLAN gene into the mutant by A . rhizogenes-mediated hairy root transformation rescued the phenotype relevant to AM symbiosis ( Fig 7F and 7G ) . In the hairy root system , although the defects in AM symbiosis was rescued compered with empty vector control , the colonization level was lower than normal root system . This may be due to the difference in root system . Overall , these results suggest that LjLAN is required for the establishment of AM symbiosis . LjLAN and LjCYCLOPS appear to have additive role for the control of AM symbiosis , as the double mutation of lan and cyclops had an additive effect on the AM symbiosis ( S12 Fig ) . To gain insight into the phenotype of AM symbiosis from marker genes expression , the expression of LjSbtM1 , LjRAM1 and LjPT4 were next investigated . Similar to previous reports [44 , 64 , 65] , the three genes were specifically and strongly activated by AMF infection in WT plants ( Fig 8A–8C ) . In the lan mutant , induction levels of LjSbtM1 and LjRAM1 were weaker than those in WT , but the LjPT4 level was largely unaffected ( Fig 8A–8C ) . AMF colonization was normal in the LORE1-tagged mutant ( S13 Fig ) . Expression of LjLAN itself seemed to be unaffected by AMF infection ( S9D Fig ) . In addition to the effect on root nodule and AM symbiosis , the LjLAN expression in non-symbiotic organs suggested that the role of LjLAN might not be restricted to the control of plant-microbe symbiosis ( S9A Fig ) . We then examined the effect of the lan mutation on shoot and root growth by growing the plants in the soil that contained enough nutrients in the absence of rhizobia and AMF . The shoot and primary root lengths in the lan mutant was shorter than WT ( S14A–S14C Fig ) . In addition , shoot branching tended to be promoted in the mutant ( S14A Fig ) . These results suggest that LjLAN has a role in the control of overall plant development . The shoot and root phenotypes of LORE1-tagged mutant was indistinguishable from WT plants ( S15 Fig ) .
Mediator is a multiprotein complex that has a fundamental role as an integrator of gene transcription , and governs diverse regulatory processes in plants including development , phytohormone signaling , and responses to biotic and abiotic stress [57–59] . The involvement of Mediator complex in such pleiotropic aspects seems to be achieved by assigning respective Mediator subunits specific functions . In this study , we showed that a nodulation-deficiency phenotype was caused by the mutation of a gene encoding a protein putatively homologous to AtMED2/29/32 subunit of Mediator complex . AtMED2 is required for the recruitment of RNA polymerase II , indicating that AtMED2 has an actual component of the complex [60] . We also demonstrated that AtMED2 could rescue the lan mutation . Thus , the functions of LjLAN and AtMED2 seem to be conserved . To the best of our knowledge , this is the first report describing the identification of a Mediator subunit that is involved in plant-microbe symbiosis . Mediator complex subunits are arranged into four modules; the head , middle and tail modules form the core part of Mediator complex , and the kinase module is separable . AtMED2/29/32 is considered as a tail module-type Mediator subunit . The function of AtMED2/29/32 appears to be pleiotropic and it has a role in abiotic stress signaling related to cold and redox , and phenylpropanoid biosynthesis [60 , 66 , 67] . Arabidopsis MED25/PHYTOCHROME AND FLOWERING TIME 1 ( PFT1 ) , which is a member of the tail module , is one of the best characterized Mediator subunits . AtMED25/PFT1 mediates pleiotropic phenomena , including flower and root development , jasmonate signaling , and salinity and water stress by interacting with key TFs acting in specific regulatory processes [58 , 68] . Upon stress or developmental stimuli , plants synthesize jasmonate isoleucine , which enables interaction between AtMED25/PFT1 and AtMYC TFs , achieving transcription of jasmonate-responsive genes [69] . In auxin signaling a compositional change in Mediator complex , that includes AtMED13 and AtMED25 , upon auxin stimuli enables Arabidopsis AUXIN RESPONSE FACTOR 7 ( AtARF7 ) and AtARF19 to activate expression of downstream genes [70] . As LjLAN is a putative orthologue of AtMED2/29/32 , an expected molecular function of LjLAN may be related to mediate gene transcription through interactions predominantly with TF in response to an environmental cue . Then what kind of TF and environmental cue can be involved in this machinery ? To date , studies using L . japonicus and M . truncatula have identified several TFs involved in nodulation , such as LjCYCLOPS/MtIPD3 , Lj/MtNIN , Lj/MtNF-Y subunits , Lj/MtNODULATION SIGNALING PATHWAY 1/2 and Lj/MtERN1/2 [3 , 37 , 71 , 72] . However , the largely severe nodulation phenotype of mutants of these TFs , does not resemble the lan nodulation phenotype , although we cannot rue out the possibility that relatively milder lan nodulation phenotype may be explained by partial functional redundancy of other Mediator subunits with LjLAN . The arrested nodulation phenotype of cyclops is partly similar to the lan nodulation phenotype [63] , but the analysis of lan cyclops double mutant suggests that LjLAN and LjCYCLOPS act in a parallel rather than in a same genetic pathway . Given the normal calcium spiking in the lan mutant , LjLAN-mediated transcriptional machinery may act downstream of calcium signaling in parallel with CSP pathway including LjCYCLOPS for the control of nodulation-related gene expression ( S16 Fig ) . Thus , the data so far obtained suggest that LjLAN may interact with unidentified TF ( s ) rather than known ones . However , we cannot rule out the possibility that lan phenotype is due to overall low transcription of key symbiotic genes . An identification of interacting proteins of LjLAN based on the analysis of protein-protein interactions will be undoubtedly needed to verify the possibilities . With respect to the potential environmental cue in this machinery , it seems reasonable to propose that rhizobia infection may be a preferential cue . As the pattern of symbiotic calcium spiking is normal in the lan mutant , a more specific cue may be produced downstream of this signal . In an example of a plant-pathogen interaction , oomycete downy mildew pathogen can attenuate salicylic acid-triggered immunity in Arabidopsis by imposing the interaction between its effector and AtMED19a [73] . Hence , it is possible that a rhizobia-derived factor may directly affect plant Mediator complex to control plant gene transcription relevant to nodulation . As described above , the Mediator complex is involved in different aspects of plant development and environmental responses . Although in this study we put particular emphasis on the role of LjLAN in plant-microbe symbiosis , it is possible that LjLAN is involved in overall plant development because shoot and root growth were affected by the lan mutation under nutrient sufficient conditions . In L . japonicus stable transformation , we use an A . tumefaciens-medited transformation , where tissue cultures undergo callus formation and shoot regeneration processes . While we were successful in making the transgenic plants with deletion in C-terminal part of LjLAN , we failed to create complete knockout plants of lan by aiming to mutate N-terminal part of LjLAN . In addition , in a stable transformation to complement non-symbiotic phenotype of lan , no regenerated plants were obtained . Therefore , based on these findings , we reason that null mutations of LjLAN are likely to affect callus formation and/or shoot regeneration processes . The non-symbiotic phenotype of lan may provide an intriguing scenario , where a general component of transcriptional machinery had been recruited to the specific functional context during the evolution of plant-microbe symbiosis . To verify this , detailed molecular function and non-symbiotic role of LjLAN need to be elucidated as an important next study . In L . japonicus WT plants , rhizobia enter into roots through the intracellular invasion system , that principally depends on the formation of root-hair ITs . The Nod factor signaling pathway has a crucial role in this process by regulating root-hair ITs formation . Generally , defects in the signaling pathway cause complete loss of root-hair ITs formation that is accompanied by no nodule formation . While L . japonicus has adopted root-hair ITs-mediated intracellular rhizobia accommodation system , the intercellular invasion system can be used in the case where some nodulation-related factors are mutated [11 , 50 , 54–56] . For example , in nfr1 nfr5 symrk spontaneous nodule formation 1 ( snf1 ) quadruple mutants , intercellular rhizobial invasion takes place despite apparently no root-hair ITs formation , which leads to the formation of functional nodules [56] . The snf1 plant is a gain-of-function mutant of LjCCaMK , in which spontaneous cortical cell proliferation occurs [16] . This observation indicates that Nod-factor receptors ( LjNFR1/5 ) and LjSYMRK may not be essential to the intercellular invasion process . Furthermore , proliferating cortical cells may need to preexist in order to allow rhizobia to intercellularly enter into roots . In the lan mutant , formation of root-hair ITs is strongly compromised , but it is likely that rhizobia can intercellularly enter into roots , as functional nodules are formed . As we could not determine if the lan mutation affects cortical-ITs formation , it remains unknown how rhizobia are finally released into nodule cells in the mutant . Due to the delay in nitrogen-fixing nodules , the lan mutant exhibit growth defects in a nitrogen-depleted condition until they obtain benefit from symbiotic nitrogen fixation . The delayed nodulation phenotype is thought to be a common feature of some L . japonicus plants , where the intercellular rhizobial invasion is used to accommodate rhizobia in roots [11 , 50 , 54–56] . Based on the lan phenotype , we propose that the predominant role of LjLAN is to initiate swift and efficient production of nitrogen-fixing nodules by promoting root-hair IT-mediated intracellular rhizobial accommodation . In other words , LjLAN may have a role in preventing protracted and less effective nodulation caused by intercellular rhizobial invasion . Analysis of L . japonicus root hairless mutants indicates that the intercellular invasion system can be adopted in the plants lacking root hairs [54] . It is unlikely that the intercellular invasion phenotype of lan is caused by such physical defects , because root hairs are normally formed in the mutant ( S14D and S14E Fig ) . It is hypothesized that root-hair IT-mediated intracellular invasion is an evolutionarily advanced invasion system , whereas intercellular rhizobial invasion is the basal pathway [56] . Among the various plants that have an ability to perform root nodule symbiosis , it is estimated that 75% of plants use intracellular invasion and the remaining 25% of plants use the root-hair independent intercellular invasion system [74] . Interestingly , a plant such as Sesbania rostrata has a dual mode invasion system where both a root hair-independent intercellular invasion and a root hair-dependent invasion can be used , depending on whether the soil is flooded or dry [75] . Future detailed analysis of LjLAN may contribute to our understanding of the genetic basis and the evolution and diversity of the rhizobial invasion system . In addition to root nodule symbiosis , the phenotype of the lan mutant during AM symbiosis suggests that LjLAN is also required for symbiosis with AMF . In contrast to the lan nodulation phenotype in which formation of functional nodules eventually takes place , the lan mutation continues suppressing the establishment of AM symbiosis . As the pattern of symbiotic calcium oscillation was normal in the lan mutant , the lan mutation seems to affect the progression of AM symbiosis downstream of calcium signaling . The expression of the AM-inducible genes , LjSbtM1 , LjRAM1 and LjPT4 is generally suppressed by mutation of the CSP genes so far identified [44] . In contrast , while the induction levels of LjSbtM1 and LjRAM1 are reduced by the lan mutation , that of LjPT4 is largely unaffected . Currently , it remains almost completely unknown why these genes show different expression patterns in the lan mutant . However , based on the loss-of-function phenotypes of each gene , LjSbtM1 and LjRAM1 are required for initiation and/or growth of arbuscules [44 , 64] . In contrast , a major role of legume PT4 seems to be associated with phosphate transport [76] . It is unclear if PT4 is directly involved in arbuscules development . Such differences in the molecular function of three genes might underlie different gene expression patterns depending on the context; while canonical CSP pathway regulates both arbuscules developmental program and phosphate transport by inducing the three genes , the LjLAN-mediated pathway may only regulate arbuscules developmental program by inducing LjSbtM1 and LjRAM1 . AMF accommodation can employ both the intercellular and intracellular dual invasion system [77] . A specialized structure called the prepenetration apparatus ( PPA ) mediates the intracellular invasion of AMF and has been suggested to share structural similarities with IT [78–80] . Given that LjLAN has a conserved role between root nodule and AM symbiosis , the predominant role of LjLAN in IT formation indicates that LjLAN also may be involved in intracellular AMF accommodation by mediating PPA formation . Future studies investigating the role of LjLAN during AM symbiosis should place particular emphasis on investigating if PPA formation is involved in AMF accommodation . Because of the lack of evidences , it is currently difficult to integratedly interpret the molecular function of LjLAN in root nodule and AM symbiosis . However , based on lan phenotype , LjLAN-mediated transcriptional regulatory system could be associated with regulation of genes acting symbiont infection processes . As cell cycle activation such as nuclear enlargement and endreduplication commonly occurs during both symbiont infections , the genes involved in this process may be target genes of LjLAN-mediated regulatory system .
The Miyakojima MG-20 and Gifu B-129 ecotype of L . japonicus was used as the WT in this study . The lan mutant was isolated as a result of screen for nodulation-deficient mutants using the M2 generation of WT plants that had been mutagenized with 0 . 4% ethylmethane sulfonate ( EMS ) for 6 hours . The LORE1-tagged line of lan ( Plant ID: 30008618 ) was obtained from Lotus Base ( https://lotus . au . dk ) . A description of the DR5:GFP-NLS plants and cyclops-6 has been published previously [45] . ccamk-14 mutant with MG-20 background was newly identified in this study . For the analysis of root nodule symbiosis , plants were grown with or without M . loti MAFF 303099 as previously described [81] . M . loti nodC mutant was obtained from LegumeBase ( https://www . legumebase . brc . miyazaki-u . ac . jp/top . jsp ) . For the analysis of the AM symbiosis , plants were grown with or without R . irregularis ( DAOM197198; PremierTech ) as previously described [65] . The nitrogenase activity of nodules was indirectly determined by measuring acetylene reductase activity ( nmol/ h per plant ) as previously described [82] . The leaves of the lan mutant were ground with liquid nitrogen using a mortar and pestle . Genomic DNA was isolated using a DNeasy Plant Mini Kit ( Qiagen ) . The quality of purified genomic DNA was evaluated by a Quant-iT dsDNA BR Assay Kit ( Invitrogen ) . For whole-genome shotgun sequencing of the lan mutant , we performed paired-end sequencing with HiSeq 2000 ( Illumina ) . After fragmentation of the isolated genomic DNA , an Illumina library with a mean insertion length of 350-bp was constructed using TruSeq Nano DNA LT Sample Preparation Kit ( Illumina ) following the manufacturer’s instructions . These libraries were subsequently sequenced 101 bp from both ends , yielding 6 . 25 gigabase ( Gb ) of raw data . After the removal of adaptor sequences and low quality reads ( Phred quality score ≥ 20 in < 90% of the bases ) , 5 . 97 Gb of high quality sequences remained . The remaining reads were mapped against L . japonicus genome assembly build 2 . 5 using the Bowtie software [83] . The median value of per-base sequence depth was 18 . 3 and the genome coverage was 90 . 2% . The resulting data in the sam format were converted into bam format using Samtools [84] . Genome-wide SNPs were called from the bam files using Samtools and Bedtools [85] . A SNP that is specific to the lan mutant was found by examining the mapped region harboring the LjLAN locus with Integrative Genomics Viewer program ( https://www . broadinstitute . org/igv/ ) . The primers used for PCR are listed in S1 Table . For the complementation analysis , a 5 . 8-kb genomic DNA fragment including the LjLAN candidate gene was amplified by PCR from WT genomic DNA . This fragment including 4 . 4 kb of sequence directly upstream of the initiation codon , was cloned into pCAMBIA1300-GFP-LjLTI6b [45] . The coding sequences ( cds ) of LjLAN and LjLANΔC were , respectively , amplified by PCR from template cDNA prepared from WT L . japonicus . The cds of AtMED2 was amplified by PCR from template cDNA prepared from Arabidopsis Col-0 plants . They were cloned into the pENTR/D-TOPO vector ( Invitrogen ) . The insert was transferred into pUB-GW-GFP [86] by the LR recombination reaction . To obtain the ProLjLAN:GUS plus construct , first an artificially-synthesized GUS plus gene was cloned into pENTR/D-TOPO vector to create the vector pENTR-gus plus . The GUS plus gene in pENTR-gus plus was introduced into a vector pCAMBIA1300-GW-GFP-LjLTI6b [87] by the LR recombination reaction to create the vector pCAMBIA1300-GUS plus-GFP-LjLTI6b . Next , 4 . 4 kb of sequence directly upstream of the initiation codon of LjLAN was amplified by PCR and cloned upstream of GUS plus gene of pCAMBIA1300-GUS plus-GFP-LjLTI6b to create the vector pCAMBIA1300-pLjLAN-GUS plus-GFP-LjLTI6b . For the analysis of calcium spiking , we used a construct in which nuclear-localized yellow-chameleon ( YC2 . 60 ) was expressed under the control of the LjUBQ promoter [81] . The recombinant plasmids were introduced into A . rhizogenes strain AR1193 [88] and were transformed into roots of L . japonicus plants by a hairy-root transformation method as previously described [45] . To create CRISPR-Cas9 construct of LjLAN , targeting site in the gene was designed using the CRISPR-P program ( http://cbi . hzau . edu . cn/crispr/ ) [89] . Oligonucleotide pairs ( S1 Table ) were annealed and cloned into a single guide RNA ( sgRNA ) cloning vector , pUC19_AtU6oligo , as previously described [90] . Then , the sgRNA expression cassette prepared in pUC19_AtU6oligo was excised and replaced with OsU3:gYSA in pZH_gYSA_FFCas9 , an all-in-one binary vector harboring a sgRNA , Cas9 , and an HPT expression construct , as previously described [90] . The recombinant plasmid was introduced into A . tumefaciens strain AGL1 and was transformed into WT L . japonicus MG-20 plants by a stable transformation method as previously described [82] . The primers used for PCR are listed in S1 Table . Total RNA was isolated from each plant tissue using the RNeasy Plant Mini Kit ( Qiagen ) or the PureLink Plant RNA Reagent ( Invitrogen ) . First-strand cDNA was prepared using the ReverTra Ace qPCR RT Master Mix with gDNA Remover ( Toyobo ) . Real-time RT-PCR was performed using a Light Cycler 96 System ( Roche ) or a 7900HT Real-Time PCR system ( Applied Biosystems ) with a THUNDERBIRD SYBR qPCR Mix ( Toyobo ) according to the manufacturer’s protocol . The expression of LjUBQ was used as the reference . Data are shown as mean±SD of 3–4 biological replicates . Sequence data from this article can be found in the GenBank/EMBL data libraries under the following accession numbers: LjLAN , LC171403; LjLAN LIKE , LC194237 . Data of short reads from the lan genomic DNA has been deposited in the DNA Data Bank of Japan Sequence Read Archive under the accession number DRA004948 .
|
Symbiosis between plants and beneficial microbes such as nitrogen-fixing bacteria and arbuscular mycorrhizal fungi has enabled plant colonization of new environments . Root nodule symbiosis with nitrogen-fixing rhizobia enables sessile plants to survive in a nitrogen-deficient environment . To establish the symbiosis , host plant cells need to accommodate rhizobia during nodule development , a process mediated by a plant-derived intracellular structure called the infection thread ( IT ) . In this study , we show that LACK OF SYMBIONT ACCOMMODATION ( LAN ) is involved in intracellular rhizobia accommodation in the model leguminous plant Lotus japonicus . LjLAN encodes a putative subunit of Mediator complex , a multiprotein complex that has a fundamental role as an activator of gene transcription . Mutation analysis suggests that LjLAN is required for root hair IT formation , which enables swift and efficient rhizobial accommodation . Moreover , we show that LjLAN is required for symbiosis with arbuscular mycorrhizal fungi . These data add a new component to the molecular mechanism relevant to the establishment of root nodule and arbuscular mycorrhizal symbiosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"biotechnology",
"reverse",
"transcriptase-polymerase",
"chain",
"reaction",
"plant",
"anatomy",
"engineering",
"and",
"technology",
"symbiosis",
"root",
"nodules",
"cell",
"signaling",
"root",
"hairs",
"plant",
"science",
"genetically",
"modified",
"plants",
"molecular",
"biology",
"techniques",
"plants",
"calcium",
"signaling",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"bioengineering",
"genetically",
"modified",
"organisms",
"artificial",
"gene",
"amplification",
"and",
"extension",
"gene",
"expression",
"molecular",
"biology",
"signal",
"transduction",
"eukaryota",
"plant",
"roots",
"cell",
"biology",
"phenotypes",
"polymerase",
"chain",
"reaction",
"genetics",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"plant",
"biotechnology",
"organisms"
] |
2019
|
LACK OF SYMBIONT ACCOMMODATION controls intracellular symbiont accommodation in root nodule and arbuscular mycorrhizal symbiosis in Lotus japonicus
|
Sulphur is an essential element that all pathogens have to absorb from their surroundings in order to grow inside their infected host . Despite its importance , the relevance of sulphur assimilation in fungal virulence is largely unexplored . Here we report a role of the bZIP transcription factor MetR in sulphur assimilation and virulence of the human pathogen Aspergillus fumigatus . The MetR regulator is essential for growth on a variety of sulphur sources; remarkably , it is fundamental for assimilation of inorganic S-sources but dispensable for utilization of methionine . Accordingly , it strongly supports expression of genes directly related to inorganic sulphur assimilation but not of genes connected to methionine metabolism . On a broader scale , MetR orchestrates the comprehensive transcriptional adaptation to sulphur-starving conditions as demonstrated by digital gene expression analysis . Surprisingly , A . fumigatus is able to utilize volatile sulphur compounds produced by its methionine catabolism , a process that has not been described before and that is MetR-dependent . The A . fumigatus MetR transcriptional activator is important for virulence in both leukopenic mice and an alternative mini-host model of aspergillosis , as it was essential for the development of pulmonary aspergillosis and supported the systemic dissemination of the fungus . MetR action under sulphur-starving conditions is further required for proper iron regulation , which links regulation of sulphur metabolism to iron homeostasis and demonstrates an unprecedented regulatory crosstalk . Taken together , this study provides evidence that regulation of sulphur assimilation is not only crucial for A . fumigatus virulence but also affects the balance of iron in this prime opportunistic pathogen .
Aspergillus fumigatus is an opportunistic fungal pathogen that may cause invasive infections in immunocompromised patients . During the last decades the incidence rate of invasive pulmonary aspergillosis ( IPA ) , the most severe infection caused by this mould [1] , has dramatically increased which mainly results from the rise in immune-suppressive medical therapies . Furthermore , IPA is accompanied by a high mortality rate which may reach up to 90% depending on the immune status of the host [2] , [3] , [4] . This pronounced lethality is primarily attributed to the relative inefficiency of current chemotherapies [5] , which are based on disrupting the integrity of the fungal cell membrane or cell wall [6] . Therefore , an urgent need to identify targets for the development of novel antifungal substances is evident . Nutritional supply is an essential prerequisite for the onset and manifestation of infection by any pathogen [7] . In opportunistic fungi like A . fumigatus , which does not seem to express host-specific virulence factors [8] , [9] , nutrient uptake and metabolic versatility have to be considered as non-specific virulence determinants ( for a recent review see [10] ) that , however , might represent promising antifungal targets [11] . To date , several metabolic routes fundamental for IPA manifestation have been identified: de novo UMP biosynthesis [12] , the folate synthesis route [13] , siderophore-mediated iron assimilation [14] , or the methylcitrate cycle [15] are essential metabolic processes supporting in vivo growth and virulence of A . fumigatus [16] . Nevertheless , detailed knowledge about the metabolic status of the fungus during intrapulmonary growth is still scarce due to the complexity of the pathogen-host system . In a seminal study , preliminary insights into the A . fumigatus-host adaptation transcriptome were gained employing extensive transcriptional profiling under various culture conditions and during an early phase of pulmonary infection , revealing the main stressors encountered by the fungal pathogen when germinating in a susceptible , leukopenic mammalian host [17] . Among these , starvation for nitrogen became evident . Yet , while numerous studies have focused on A . fumigatus nitrogen metabolism [16] , [18] , [19] , [20] , [21] , [22] , neither the exact source ( s ) of this macroelement nor specific metabolic routes of nitrogen assimilation during pulmonary infection have been identified so far . Sulphur ( S ) is another essential nutrient that the fungus needs to acquire from the surrounding tissue during intrapulmonary growth , as it is a constituent of the proteinogenic amino acids cysteine and methionine as well as of essential organic molecules like coenzyme-A , glutathione and , particularly , iron-sulphur ( Fe-S ) clusters . Only a few studies have addressed its relevance for fungal virulence so far , focusing on synthesis and utilization of the sulphur-containing molecule glutathione in Candida albicans or C . glabrata [23] , [24] . The importance of sulphur metabolism for aspergillosis has not been addressed to date . Aspergillus species can utilize a variety of sulphur-containing molecules , such as inorganic S-sources , e . g . sulphate , or sulphate esters of organic compounds [25] . The assimilation of sulphate is performed via a well-defined pathway that comprises its uptake by specialised permeases , activation by ATP-driven phosphorylation in two steps , and reduction to sulphite and further to sulphide [26] . The latter is condensed with O-acetyl serine to yield cysteine , from which methionine and also S-adenosylmethionine can be formed . Sulphur homeostasis is tightly regulated in filamentous fungi by catabolite repression [27] . In the model filamentous fungus Neurospora crassa for instance , CYS-3 , a positive-acting transcriptional factor of the bZIP family , was identified that activates expression of a set of genes whose products are required to acquire and utilize secondary S-sources under sulphur-starved conditions [28] , [29] , [30] . Furthermore , the role of so-called sulphur controller ( scon ) genes has been addressed that influence methionine repression of sulphur-containing amino acid biosynthesis [31] . Regulation of sulphur metabolism has also been extensively studied in the bakers' yeast Saccharomyces cerevisiae ( for detailed information consult reviews [32] or [26] ) . While in N . crassa CYS-3 is sufficient to bind DNA and to activate transcription , in S . cerevisiae a heteromeric complex of three proteins ( Met4-Met28-Cbf1 ) is required for proper sulphur-dependent regulation [33] , [34] , [35] . The CYS-3 orthologue of A . nidulans , MetR , is required to activate transcription of several genes related to sulphur acquisition and to allow growth on a variety of sulphur sources other than methionine [36] . Successful complementation of an A . nidulans metRΔ strain with its MET1 orthologue from the dimorphic human-pathogenic fungus Paracoccidioides brasiliensis demonstrated functional conservation of this transcriptional regulator [37] . Adaptation to environmental stressors is usually based on reprogramming of the cellular expression pattern triggered by specific transcription factors . Consequently , the relevance of several cellular processes for A . fumigatus virulence has been investigated by targeting the corresponding regulators ( for review see ref . [38] ) . Deletion of genes coding for zinc or iron responsive factors demonstrated the importance of both elements for virulence [39] , [40] . Also , a positive role of amino acid homeostasis in A . fumigatus virulence was established by deletion of the respective cross-pathway control regulator [41] . A substantial benefit from targeting wide-domain regulators of a given cellular aspect lies in its comprehensive outcome , in contrast to particular gene deletions affecting activities that may be encoded redundantly in the fungal genome . Based on the hypothesis that flexible regulation of sulphur homeostasis supports growth in a susceptible host and therefore might affect virulence of A . fumigatus , we became interested in the cellular function of the MetR orthologue in this opportunistic pathogen . Our results demonstrate that this transcription factor is a key regulator of sulphur assimilation and that it is crucial for A . fumigatus pathogenicity . Moreover , we describe an unprecedented regulatory crosstalk of S-metabolism and iron homeostasis .
The A . fumigatus MetR transcription factor was identified by BLAST search [42] , [43] , [44] on the NCBI server ( http://www . ncbi . nlm . nih . gov/ ) using the A . nidulans MetR protein sequence ( PubMed acc . no . AAD38380 ) as a query , revealing 64% identity and 75% similarity between both proteins ( Fig . 1A ) . Notably , the leucine zipper ( bZIP ) domains of the factors are virtually identical with 97% identity and 100% similarity , which suggests that both could recognize a similar DNA target sequence . In the CADRE database [45] the A . fumigatus metR gene ( AFUA_4G06530 ) had been automatically annotated to contain 1517 base pairs ( bp ) , resulting in a predicted coding region of 918 nucleotides ( nt ) based on the presence of an unusually long intron 599 nt in size . The observable size difference in a 1% agarose gel between a metR cDNA ( obtained by reverse transcription from mRNA ) and gDNA ( amplified from genomic DNA ) matched with the presence of this intron ( not shown ) , and sequencing of a complete cDNA insert confirmed the overall architecture of the metR gene as annotated . The deduced protein sequence comprises 305 amino acids with a predicted molecular weight of 33 kDa . The MetR transcription factor displays a high degree of conservation among ascomycetous fungi ( Fig . 1B ) , being present in all genera analysed . With respect to the A . fumigatus MetR sequence , identities/similarities range from 25/33% to the N . crassa orthologue and up to 95/100% to the Neosartorya fischeri one . Also , the unusually long intron appears to be conserved among ascomycota , suggesting a common origin of the gene . With the aim to probe localisation as well as any cellular function of the metR gene product , an A . fumigatus strain that would express a functional GFP-tagged version from the endogenous metR promoter was constructed . This strain AfS171 was shifted from sulphate-containing medium for 90 minutes to medium lacking any source of sulphur ( Fig . 1C ) . Under S-rich conditions a faint fluorescent signal was observed that was uniformly distributed in the hyphal cytoplasm . Upon S-depletion , however , translocation of the MetR-GFP protein to the nuclei became evident , as deduced from co-localisation of the fluorescent signals with the nuclear stain 4′ , 6-diamidino-2-phenylindole ( DAPI ) . Accordingly , the presumed DNA binding domain of MetR and its cytoplasmic-nuclear shuttling under sulphur-limiting conditions imply that this gene product acts as a transcriptional regulator of A . fumigatus sulphur metabolism . To gain insights into the cellular function of the metR gene product , a full deletion strain of A . fumigatus was constructed by homologous gene replacement employing a self-excising recyclable cassette that contains a hygromycin B resistance gene as selectable marker [46] , [47] . Southern analysis of the resulting strains AfS166 [metR::six-β-rec/hygroR-six] and AfS167 [metR::six] confirmed the homologous replacement and the excising event , respectively ( Fig . S1A ) . A preliminary phenotypic analysis revealed that a metR deletant is unable to grow in the presence of sulphate as sole source of sulphur ( Fig . 2A ) . This allowed us to reintroduce the metR gene at its original locus without using any selection marker but the presence of sulphate ( SO42− ) as the only source of sulphur . In order to differentiate between the desired reconstituted strain and its wild-type progenitor , a silent mutation was introduced in the gene's coding sequence to create an additional BstEII restriction site . Southern analysis confirmed the correct integration event for a representative isolate ( Fig . S1B ) . To address the role of the MetR factor for the ability of A . fumigatus to utilize different sulphur sources , the resulting strains were subjected to phenotypic inspection on various S-containing media ( Fig . 2A ) . With respect to inorganic S-sources , the metR::six ( syn . metRΔ ) mutant strain AfS167 was unable to grow in the presence of any of the tested substrates . Among the organic sources , the metRΔ mutant grew on methionine as well as homocysteine , which contrasts with a corresponding A . nidulans metRΔ mutant that was described to grow only poorly on the latter compound [32] . It was noticed , however , that homocysteine did not suffice as source of sulphur at alkaline pH while methionine was perfectly assimilated ( not shown ) . Furthermore , AfS167 was able to utilize cysteine or glutathione as combined nitrogen and sulphur sources when the culture media were additionally depleted for nitrogen . Importantly , the metRΔ strain barely grew on porcine lung agar ( PLA ) , suggesting that this mutant could suffer from a growth defect within the pulmonary tissue of a susceptible host and , therefore , might have reduced virulence capacities . When methionine was added to the PLA medium growth of the mutant was restored , which indicates that this tissue contains insufficient free levels of this amino acid to support growth of the deletant . In order to test for any cross-talk of utilisation pathways for other macro-elements , growth of the metRΔ mutant was monitored on various N- , C- , or P-sources in the presence of methionine and sulphate , respectively ( Fig . S2 ) . With one exception , no obvious phenotype emerged in dependency of the various supplements , i . e . the mutant was able to grow only when methionine was present as source of sulphur . However when galactose served as C-source , growth of the deletant appeared repressed even in the presence of methionine , which indicates that this particular sugar interferes with methionine utilisation in A . fumigatus . To better understand the growth defect of AfS167 on inorganic S-sources , germination of the deletant was investigated ( Fig . 2B ) . The strain was unable to swell or germinate in the presence of SO42− up to eleven hours after inoculation; on the contrary , its germination on methionine was only slightly delayed in comparison with the wild-type strain and reached approximately the same rate after prolonged incubation . Therefore , it appears that sulphur assimilation is an indispensable prerequisite for proper germ tube emergence and its absence results in a pronounced germination and growth defect as displayed by the metR::six mutant strain . To further analyse the S-assimilation capacity of the deletion strain we made use of Biolog Phenotype MicroArrays ( Fig . 3 ) . This technology allows the evaluation of a variety of microbial phenotypes in a parallel fashion . From a variety of conditions and stressors available , the PM4 MicroPlate plate containing 35 different sulphur sources was selected . Wild-type and mutant strains were incubated in parallel in the wells of this microtiter plate and their growth ability in the presence of different S-sources was monitored via the optical density ( O . D . ) at 630 nm . The recommended Biolog Redox Dye to monitor growth by measuring the respiration process was also tested; however , as previously described for filamentous fungi [48] , the results obtained using this dye turned out to be inconsistent ( data not shown ) . Therefore , one well lacking any sulphur source was used to normalize O . D . values to define a growth threshold of 0 . 2 , which was further confirmed through microscopic inspection . Since the use of phenotypic microarray plates is a technology that has not been extensively used for filamentous fungi yet , some of the compounds present in the PM4 MicroPlate were re-tested in regular phenotypic assays on solid culture media to support validity of the results ( not shown ) . Growth phenotypes deduced from this Biolog plate confirmed that the metRΔ strain is unable to utilise any oxidized inorganic sulphur source . Indeed , it only grew on methionine or derivatives thereof ( i . e . N-acetylmethionine or methionine sulphoxide ) . Notably , the mutant did not grow in the presence of methionine sulphone , although the wild-type strain grew perfectly well , but it grew better than its progenitor on N-acetylmethionine . The central conclusions reached from these phenotypic analyses are that the metRΔ mutant is unable to grow on inorganic sulphur sources and that utilization of methionine and its derivatives is independent from the presence of the transcriptional activator MetR . Because the AfS167 strain is unable to grow in the presence of S2− ( Fig . 2A ) and as it is known that several fungi , including Aspergillus species , produce volatile sulphur compounds ( VSCs ) like hydrogen sulphide ( H2S ) , dimethylsulphide ( H3C-S-CH3 ) , or methanethiol ( CH3-SH ) as a result of methionine catabolism [49] , [50] , [51] , we became interested in studying whether A . fumigatus would be able to utilize such volatile compounds as S-source and if generation of such VSCs is MetR-dependent . The wild-type and metRΔ strains were cultured in small petri dishes with minimal medium containing methionine as sulphur source . These plates were placed inside larger petri dishes with medium lacking any S-source . Neither the wild-type isolate nor the mutant was able to grow in the absence of any sulphur source ( not shown ) . However , when either strain was grown on the methionine-containing petri dish , growth on the outside sulphur-depleted medium was observed only for the wild-type strain ( Fig . 4 ) . Accordingly , A . fumigatus is able to take up VSCs produced from methionine catabolism and to use them as S-source . Production of VSCs appears to be independent from the presence of the MetR regulator but their utilization as S-source requires the presence of this regulatory factor . In the light of these results , it is valid to conclude that the MetR factor represents a master regulator of sulphur assimilation in A . fumigatus acting predominantly on utilization of inorganic S-sources , while consumption of methionine does not depend on this transcriptional regulator . The aforementioned phenotypic results suggest that the MetR factor might activate the transcription of genes required for the uptake and utilization of different sulphur sources . To investigate this , the short-term transcriptional responses of a wild-type strain and its metRΔ derivative to variations in the available sulphur source were monitored: Mycelia grown overnight in Aspergillus minimal medium ( AMM ) with methionine were shifted to media containing diverse S-sources , incubated for one additional hour , and steady-state levels of several transcripts of genes related to sulphur uptake and utilization were monitored by Northern blot hybridisation . Initially , transcript levels of the metR gene itself were checked to observe that its transcription is apparently not regulated by the nature of the sulphur source ( Fig . 5A ) , resembling the situation in A . nidulans [32] but contrasting findings in N . crassa [24] , [25] . Interestingly , a second hybridising signal was detected for the metR transcript under S-starvation conditions , indicating alternative processing of the encoding transcript . In order to further understand why the metRΔ mutant is unable to grow on oxidized inorganic sulphur sources , expression of all genes encoding enzymes of the sulphate assimilation pathway , which are sulphate permease ( sB ) , ATP-sulphurylase ( sC ) , APS-kinase ( sD ) , PAPS-reductase ( sA ) and sulphite reductase , as well as expression of one arylsulphatase-encoding gene ( required for utilization of sulphur esters , i . e . nitrophenyl sulphate ) was checked ( Fig . 5A ) . Upregulation of the sulphate permease- , ATP-sulphurylase- and APS-kinase-encoding genes under sulphur-starving conditions depended on the presence of MetR factor . For the sB gene , a second , longer transcript became evident under sulphur starvation . Transcription levels of the genes coding for PAPS reductase and sulphite reductase were decreased in the absence of the MetR factor on all S-sources in comparison to the wild-type . Furthermore , expression of the arylsulphatase-encoding gene was completely shut down in the mutant . This transcriptional pattern agrees with and partially explains the incapacity of the metRΔ deletant to grow on inorganic sulphur sources . In contrast to inorganic sulphur sources , the AfS167 deletion mutant grows on methionine independently of other conditions such as varying pH ( not shown ) or the availability of nitrogen . To address the reason for this phenotype , transcription of several methionine-related genes was investigated ( Fig . 5B ) . A BLAST search against the A . fumigatus genome sequence identified three putative transporters involved in methionine uptake , which we named mup genes . Transcript levels of mupA and mupC were constitutive with respect to the sulphur source and seemed to be mostly independent of MetR , although mupA expression appeared somewhat elevated in the wild-type . Transcription of the related mupB gene could not be detected under any condition tested ( not shown ) . Transcription of the putative methionine synthase-encoding gene metH was also not regulated by MetR . Interestingly , its expression was reduced in the presence of methionine and increased under sulphur-starving conditions , suggesting that the intracellular pool of methionine is constantly maintained and emphasizing the importance of this particular amino acid . Importantly , the methionine aminotransferase-encoding gene metAT , whose product is probably responsible for methionine degradation [52] , [53] and consequently for its utilization as sulphur source , was highly expressed in the presence of methionine in a MetR-independent manner . Therefore , the expression pattern of these genes perfectly agrees with the ability of the deletion strain to utilize methionine as S-source . Interestingly in the metRΔ mutant , metAT was highly expressed under sulphur-depleted conditions . We hypothesized that this is due to the strong and rapid sulphur starvation affecting the mutant under such sulphur-restricted conditions . To verify this assumption , the wild-type and mutant strains were incubated up to eight hours on sulphate-containing and sulphur-depleted media ( Fig . 5B ) , which results in substantial sulphur-limiting conditions for the deletant but not for the wild-type . Accordingly , metAT expression was upregulated in the metRΔ mutant but not in the wild-type , demonstrating that after prolonged incubation in the presence of sulphate the mutant strain becomes depleted for sulphur . Surprisingly , the wild-type did not upregulate expression of metAT even after eight hours of incubation under sulphur-depleted conditions , implying that the resulting sulphur starvation in the wild-type is not that severe , probably due to MetR-dependent recycling processes and mobilization of reserves . In conclusion , the metAT expression profile suggests that other mechanisms apart from MetR-mediated regulation must exist to orchestrate gene expression depending on the availability and source of sulphur . With respect to cysteine assimilation no clear candidates for its degradation and utilization as sulphur source have been identified so far and , consequently , the incapacity of the deletion strain AfS167 to grow on cysteine could not be addressed properly . Nevertheless , expression of the putative cysteine transporter-encoding gene cynA , an orthologue to a C . glabrata-specific cysteine transporter [54] , was not upregulated in the metRΔ strain under sulphur starving conditions ( Fig . 5C ) , which would partially explain the observed phenotype . Furthermore , transcription of the cysteine synthase-encoding gene cysB was slightly increased in the absence of cysteine in a MetR-dependent manner , which might translate into a slightly reduced level of cysteine in the deletion strain . To demonstrate a direct effect of MetR on the transcription of sulphur assimilation genes , we performed chromatin immuneprecipitation ( ChIP ) analyses making use of strain AfS171 expressing a functional GFP-tagged version of this transcription factor . Fixed chromatin samples isolated from fungal cultures that had been starved for sulphur were sheared and precipitated with a nano-trap ( see Supporting Information for details ) . Interrogating the output fractions by semi-quantitative PCR revealed a pronounced and reproducible enrichment of fragments spanning the promoters of several candidate genes , such as the ones encoding the APS-kinase , the arylsulphatase , the ATP-sulphurylase , as well as the sulphate permease ( Fig . 5D ) . Following the observation that the expression of genes whose products are required for sulphur assimilation is regulated by MetR , we became interested in understanding to what extent any transcriptional remodeling that takes place under sulphur-limiting conditions is MetR-dependent . For this purpose , overnight-grown mycelia of the wild-type and metRΔ strains were shifted from cultures containing sufficient levels ( 5 mM ) of methionine serving as sole S-source to media containing low methionine levels ( 0 . 2 mM ) over a time frame of eight hours before RNA was harvested . Previous culturing experiments had shown that methionine depletion became manifest within this time frame , so this experimental set-up allows assessment of any MetR contribution to the transcriptional response upon mild S-depletion . Nucleic acid samples were prepared from two biological replicates each to perform digital transcriptome analyses by the RNA-seq approach ( see Materials and Methods for details ) . Comparison of both transcriptomes under this specific condition revealed that 288 genes were downregulated and 349 were upregulated in the metRΔ strain with respect to its wild-type progenitor ( >1 . 5-fold change , p-value<0 . 05 ) ( Table 1 and Table S1 ) . Categorisation via the FungiFun suite [55] revealed that the main cellular functions affected by the absence of MetR are membrane transport , metabolism , carbohydrate metabolism , and oxidation/reduction ( Fig . 6 ) . Therefore , MetR action is required for the correct remodeling of these processes to counteract conditions of sulphur depletion . To further understand this adaptation , we performed a deeper functional categorisation . Various genes assigned to cation homeostasis were less abundant in the metRΔ mutant ( Table 2 ) , suggesting a strong dysfunction in the regulation of the metabolism of these ions . In addition , several genes whose products participate in sugar , glucoside , polyol and carboxylate metabolism were downregulated what highlights the greatly different metabolic status of the mutant under sulphur starvation . Furthermore , genes related to cellular export and secretion were also identified , hinting a distinct interaction with the environment . Surprisingly , several genes related to mRNA synthesis and were also found to be downregulated , which indicates a interplay of MetR regulation with other transcription factors and cell cycle regulation . Among the transcripts that are more abundant in the mutant ( Table 3 ) , several genes whose products participate in amino acid transport were of special interest as this implies a link between nitrogen and sulphur assimilation . In addition , various genes related to DNA conformation and repair were found to be upregulated , which might reflect the severe stress situation for the mutant under sulphur starving conditions . Intriguingly , several genes related to cation transport , and especially siderophore transport and reductive iron assimilation , were found to be upregulated , suggesting a connection of MetR-mediated regulation to iron homeostasis . Indeed , among the 20 genes that showed a higher expression in the mutant compared to the wild-type ( Table 1 ) , 11 genes that have previously been found to be upregulated during iron starvation dependent on the iron regulator SreA were identified , including five genes of proven function in siderophore biosynthesis [56] , [57] . Further inspection of the entire list of upregulated genes revealed 29 of the known 49 SreA target genes with 13 of proven function in siderophore metabolism , reductive iron assimilation and iron regulation ( not shown ) . In conclusion , the MetR-dependent adaptation to sulphur starvation conditions is a complex process that involves broad transcriptional remodeling to achieve altered expression of genes belonging to various functional categories . In order to address the role of fungal sulphur utilization for growth and therefore virulence in a susceptible host , the involvement of the sulphur-related transcriptional regulator MetR in A . fumigatus virulence was assessed in different animal models . Initially , the alternative host model of the wax moth larvae Galleria mellonella was assayed ( Fig . S3 ) , where the metRΔ mutant displayed a significantly reduced virulence ( p-value of <0 . 001 ) similar to the reduction observed for the control strain , an avirulent pabaAΔ mutant . This decrease in virulence was specifically attributed to the absence of metR , since the reconstituted strain recovered full virulence . Interestingly , when injected in a solution containing 5 mM methionine , the metRΔ strain was able to kill larvae as the wild-type , suggesting that the decrease in virulence is due to the absence of a proper source of sulphur . The results obtained in the wax moth model encouraged us to perform infections in established mouse models of aspergillosis . When challenging immunosuppressed , leukopenic mice intranasally with conidia of the metRΔ mutant to induce invasive pulmonary aspergillosis , a highly significant ( p<0 . 001 ) reduction in the virulence capacity of this strain was observed ( Fig . 7A ) with more than 80% of the cohort surviving the infection . This virulence attenuation was once again specifically ascribed to the absence of metR , since the reconstituted strain regained full virulence . Histological inspection of lung sections from infected animals revealed that infectious propagules of the metRΔ strain had been cleared in the course of infection by the residual immune system , in contrast to invasive tissue penetration of the wild-type progenitor strain . Accordingly , all lungs inspected from the metRΔ-infected cohort appeared as normal , while invasive growth of hyphal elements became evident for those infected with the wild-type strain . Fungal burdens assessed from lungs of infected mice ( n = 5 animals per group ) indicated a 50-fold reduction in colony forming units per gram tissue ( 500±141 vs . 23 849±3 770 ) for the deletion strain in comparison to its wild-type progenitor ( Fig . 7B ) . To further corroborate the differences in virulence , competitive infection experiments were performed in order to obtain a competitive index ( CI ) [13] . In this assay , a cohort of four animals was infected with an input ratio of 1∶1 for wild-type and the deletion mutant and , four days later , output ratios were determined by assessing the number of colony forming units on permissive and selective media from homogenised pulmonary tissue . A mean CI value smaller than 0 . 1 was calculated for the metRΔ mutant ( Fig . 7C ) , meaning that this strain is virtually avirulent . To finally analyse the dissemination capacity of the metRΔ mutant in the bloodstream of leukopenic mice and the relevance of sulphur utilization for this process , a systemic infection model was applied ( Fig . 7D ) . Animals infected intravenously with the mutant strain showed significantly delayed mortality ( p<0 . 001 ) . In the light of these results we conclude that regulation of sulphur assimilation is essential for manifestation of pulmonary aspergillosis as well as relevant for haematogenous dissemination after angioinvasion of A . fumigatus . It is well established that defects in mitochondrial Fe-S cluster biogenesis or transport induce transcription of the iron regulon [58] , [59] and that Fe-S cluster-containing proteins participate directly in sensing iron availability in S . cerevisiae [60] , [61] . Accordingly , we expected a regulatory cross-talk between sulphur assimilation and iron homeostasis in A . fumigatus , which was further indicated by the transcriptional profiling data ( see above ) , however , which had not been tested in eukaryotes so far . We took advantage of our metRΔ mutant strain , which can be rapidly depleted for sulphur , to test this hypothesis . Overnight grown strains were shifted from culture media containing methionine to media depleted for this amino acid but containing sulphate . These media pose sulphur starving-conditions for the deletion strain but not for the wild-type or the reconstituted strain . After eight hours of incubation , although there was sufficient iron in the medium , the mutant strain increased transcription of several genes encoding proteins that participate in iron acquisition that are known to be upregulated under iron starvation , i . e . genes involved in siderophore biosynthesis ( sidA ) , siderophore transport ( mirB ) , mitochondrial ornithine export ( amcA ) and iron regulation ( hapX ) [57] . In contrast , the mutant decreased transcription of genes whose products participate in iron-consuming processes that are known to be downregulated under iron depleted conditions [40] , i . e . genes encoding aconitase ( acoA ) , cytochrome c ( cycA ) or components of the mitochondrial iron-sulfur-cluster biosynthetic machinery ( isa1 ) ( Fig . 8A ) . To analyse whether the cells were indeed depleted for iron , levels of iron chelated by ferricrocin ( FC ) , the intracellular siderophore used for iron storage and transport [62] , [63] , were measured ( Fig . 8B ) . Despite an expression pattern resembling that of iron starvation , the AfS167 mutant showed a nearly fivefold increased FC content . Combining ferricrocin analysis with total intracellular iron level measurements further underscored this imbalance in iron homeostasis ( Table 4 ) : MetR deficiency raises the cellular iron content 1 . 6-fold in the presence of methionine , which increases to 2 . 8-fold in its absence . The metRΔ mutant furthermore displays a 5-fold increased FC-chelated iron content under +Met conditions that is further enhanced to 7 . 5-fold when this S-sources is withdrawn . In conclusion , the cells indeed contain sufficient amounts of iron but display a defect in iron sensing and/or regulation . This dysregulation , causing an enhanced expression of iron uptake-related genes under sulphur starving conditions , translates into a phenotype of hypersensitivity to iron ( Fig . 8C ) . At a low concentration of iron and methionine the metRΔ mutant was able to grow , although poorly due to the shortage of sulphur . This phenotype was recovered with higher availability of methionine . However , at higher iron concentrations , the wild-type strain could grow while the AfS167 mutant did not , unless a high amount of methionine was present in the medium . This might be the consequence of both the inability to shut down expression of iron uptake-related genes and of the lower cccA gene expression , encoding a recently described vacuolar transporter that has a prominent role in iron detoxification [64] , in the metRΔ mutant especially under sulphur starvation ( Fig . 8A ) , which results in iron accumulation ( see Table 4: 2 . 4-fold increase for the metRΔ deletant under –Met conditions ) to presumably toxic levels in the cytosol . Dysregulation of iron homeostasis was also tested in the wild-type strain by shifting its mycelium to a medium completely depleted for sulphur . However , expression of the iron regulon was not observed even after 24 hours of incubation ( not shown ) , most likely because of the fact that the wild-type does not face such severe sulphur starving conditions as its metRΔ derivative apparently does .
Fulfillment of nutritional and metabolic requirements is essential for all pathogenic microorganisms to be able to grow inside the host and , thus , to cause infection and disease [7] . For opportunistic fungal pathogens this is highly relevant , since these commonly lack specific virulence factors that would provoke host damage [9] . In recent years more and more evidence has been provided that fungal metabolism is a critical component of fungal virulence [65] . Accordingly , it has been proposed that based on this knowledge novel antifungal targets might be identified [10] . Sulphur metabolism is directly related to virulence of several pathogenic microorganisms , such as Mycobacterium tuberculosis , Salmonella enterica , or protozoan parasites [66] , [67] , [68] , [69] , [70] , [71] . Among fungi , sulphur metabolism has been extensively studied in the bakers' yeast S . cerevisiae [30] , [32] and in the non-pathogenic filamentous fungi N . crassa [24] , [28] , [29] and A . nidulans [36] . However , our knowledge on the role of sulphur metabolism for fungal virulence has remained scarce . Only two studies have specifically addressed the importance of the sulphur-containing molecule glutathione in C . albicans and C . glabrata to demonstrate that glutathione biosynthesis , but not its uptake or degradation , is essential for virulence [23] . Accordingly , glutathione appears not to be the sulphur source these Candida species exploit in vivo , and its relevance for pathogenesis is probably due to its impact on iron metabolism [72] . For the human pathogen Paracoccidioides brasiliensis it was demonstrated that growth of the yeast form , which is the pathogenic state of this dimorphic fungus , strictly relies on inorganic sulphur sources and that the mycelial-to-yeast switch requires an organic source of sulphur [37] , [73] . Here , we report first evidence that proper regulation of sulphur metabolism is crucial for A . fumigatus virulence . This result can possibly be extrapolated to other fungal pathogens and , therefore , might constitute a novel field for the identification of new targets in fighting fungal infections . The information gathered in this work suggests that the MetR regulon directly affects genes whose products are related to assimilation processes of sulphur ( especially inorganic sulphur ) , rather than metabolic processes ( Fig . 9 ) . Presence of the MetR transcription factor is essential for growth on several sulphur sources , specifically on those containing inorganic sulphur sources . Accordingly , MetR is required for activation of transcription of genes encoding enzymes of the sulphate assimilation pathway and an arylsulphatase activity , which demonstrates a direct role of MetR in inorganic sulphur acquisition . Remarkably , the metRΔ mutant was able to use cysteine and the Glu-Cys-Gly tripeptide glutathione as a source of sulphur only under nitrogen-starving conditions , implicating a link between S- and N-acquisition . One possible explanation is that under nitrogen-limiting conditions increased expression of amino acid permeases and oligopeptide transporters facilitates uptake of these particular sulphur-containing compounds , which then can be exploited as S-source . Since no specific enzymes for cysteine catabolism have been identified so far , this scenario could not be investigated further . The fact that the expression of the methionine aminotransferase-encoding gene metAT is elevated in the presence of cysteine suggests that this amino acid might be transformed into methionine rather than being catabolized directly . Accordingly , catabolism of cysteine as sulphur source appears to be MetR-independent and , therefore , its uptake might represent a bottle neck that prevents the mutant to utilize cysteine . This notion is further supported by the fact that in the RNA-seq data set expression of the oligopeptide transporter OptG ( AFUA_6G03140 ) , the orthologue to the C . albicans glutathione transporter OPT7 [23] , was observed to be expressed higher in the wild-type than in the metRΔ mutant . Thus , cynA and optG are candidate genes to support A . fumigatus growth in the presence of cysteine and glutathione , but further studies are needed to elucidate whether they encode specific A . fumigatus transporters and whether cysteine acid is catabolized directly . The ability of A . fumigatus to produce volatile sulphur compounds derived from methionine catabolism has been demonstrated previously [51] . Here we demonstrate for the first time that A . fumigatus is furthermore able to utilize such VSCs as S-source . Importantly , we could show that utilization but not production of VSCs is MetR-dependent . This agrees with the inability of the metRΔ mutant to grow on S2− and also with the fact that all studied genes that participate in methionine metabolism are MetR-independent in their expression . Nevertheless , further studies are necessary to unravel the methionine catabolism pathway in order to understand the capacity of a metRΔ mutant to utilize it as S-source , the VSCs production process , and also to identify any specific VSCs that can be utilized . As an important part of our current study we could show that MetR is important for virulence of A . fumigatus in G . mellonella larvae as well as in leukopenic mice . In the wax moth larvae , metRΔ regained virulence when supplemented with methionine , suggesting that the main reason for the decrease in virulence is the absence of a suitable source of sulphur in the larval hemocoel . In the same way one might speculate that the decrease in virulence observed in the mouse model is also due to insufficient levels of methionine in the murine lung or blood and , thus , that this amino acid is not the primary source of sulphur exploited by A . fumigatus within the pulmonary tissue or in the bloodstream . In line with this are data from Purnell ( 1973 ) on a methionine-requiring mutant of A . nidulans that displayed unaltered virulence in systemic infections of mice [74] . Inorganic compounds that cannot be assimilated by the metRΔ mutant may accordingly serve as initial S-source during infection . Taking into account that the mammalian lung probably constitutes a nitrogen-limiting environment [17] , cysteine and glutathione also likely do not serve as sources of sulphur during pulmonary infection , since the mutant is able to utilize these compounds in vitro under nitrogen-starving conditions . However , since several other sulphur-related processes are deregulated in the metRΔ mutant , we cannot conclude that the mere absence of a suitable sulphur source impairs the growth of the metRΔ mutant within the murine lung . Defects in iron regulation or sensing that are characteristic for the metRΔ mutant may also account for its attenuated virulence , since iron relates to fundamental cellular processes such as respiration or oxidative stress resistance [40] , [75] . In the bakers' yeast the Aft1p and Aft2p transcription factors mediate upregulation of the so-called iron regulon under iron limiting conditions [76] , [77] , [78] . Iron sensing by Aft1p and Aft2p requires proper mitochondrial Fe-S cluster biosynthesis as well as a functional export to the cytoplasm and , consequently , disturbance of these processes provokes upregulation of the iron regulon [58] , [59] , [79] . Impairment of Fe-S cluster biogenesis can be achieved by disruption of the cysteine desulphurase-encoding gene nfs1 that is required for sulphide supply in Fe-S cluster biosynthesis [80] , [81] . In addition , it was shown that glutathione participates in Fe-S cluster translocation to the cytoplasm and thus its depletion activates the iron regulon [79] , [82] . All these relations between sulphur-containing molecules and iron-dependent transcriptional regulation strongly suggest a connection between sulphur metabolism and iron homeostasis . Other fungi , such as Aspergillus species or Schizosaccharomyces pombe , do not express Aft1/2 orthologues . Here , iron regulation is mediated by the interplay of the unrelated HapX/Php4 and SreA/Fep1 proteins [40] , [83] , [84] . Yet , the strategy for iron sensing is likely to be conserved in the fungal kingdom , which might link iron homeostasis to sulphur metabolism in general . Here , by virtue of a deletion mutant , we clearly demonstrate this relationship: starving the A . fumigatus metRΔ mutant strain for sulphur results in increased expression of the iron regulon . We hypothesize that the pronounced sulphur starvation of the metRΔ mutant impairs Fe-S cluster biogenesis and/or glutathione biosynthesis , which , in turn , activates the iron regulon . The corresponding wild-type isolate would not act on iron homeostasis under sulphur-depleting conditions as it apparently does not face such a severe starvation , most likely due to the utilization of reserve pools and salvage pathways . We cannot rule out the possibility that MetR directly regulates transcription of an unidentified gene whose product is required for iron sensing or proper iron regulation under sulphur-starving conditions . But given the fact that MetR deficiency strikingly phenocopies a deficiency for the negative iron regulator SreA ( increased cellular iron as well as ferricrocin contents accompanied by transcriptional derepression of genes involved in iron acquisition such as siderophore biosynthesis and uptake as well as reductive iron assimilation ) , such a direct action on the sreA and hapX genes encoding the main players of iron homeostasis in A . fumigatus is unlikely . Along that line , we could not detect binding of the MetR-GFP protein to promoter regions of either gene by chromatin immune-precipitation . The disclosed link between sulphur metabolism and iron homeostasis represents an appealing crosstalk between two fundamental cellular regulatory circuits that calls for further investigation . In summary , we show for the first time that regulation of sulphur metabolism is important for the ability of A . fumigatus to cause disease . Given the conserved nature of sulphur assimilation in the fungal kingdom , its relevance in virulence is likely to be a general feature among pathogenic fungi . Considering that many of these routes are absent in mammals , some of these processes might represent suitable novel targets for antifungal drug development .
Mice were cared for in accordance with the principles outlined by the European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes ( European Treaty Series , no . 123; http://conventions . coe . int/ Treaty/en/Treaties/Html/123 . htm ) . All infection experiments were carried out in compliance with the German animal protection law in a protocol approved by the Government of Lower Franconia ( file number: 55 . 2-2531 . 01-90/09 ) . The Escherichia coli strain DH5α [85] was used for cloning procedures . Plasmid-carrying E . coli strains were routinely grown at 37°C in LB liquid medium ( 1% peptone , 0 . 5% yeast extract , 0 . 5% NaCl ) under selective conditions ( 100 µg•ml−1 ampicillin or 50 µg•ml−1 kanamycin ) ; for growth on plates , 1 . 5% agar was added to solidify the medium . All plasmid constructs used in the course of this study are listed in Table S2 and were generated using the Seamless Cloning ( Invitrogen ) technology as described in the Supplementary Material ( Text S1 ) . The wild-type A . fumigatus strain ATCC 46645 served as common reference [86] , derivatives of this isolate generated in the course of this study are described in the Supplementary Material ( Text S2 ) . A . fumigatus strains were basically cultured in nitrate-based minimal medium [87] containing 1% glucose , 70 mM NaNO3 , 7 mM KCl , 11 mM KH2PO4 ( pH 5 . 5 ) , 0 . 25 mM MgSO4 , trace elements solution , and 2% agar ( Serva ) for solid media at 37°C . In case of selection for the presence of the hygromycin B resistance marker , 50 µg•ml−1 of this antibiotic ( InvivoGen ) were applied . In sulphur-free medium , MgCl2 was substituted for MgSO4 and a modified mixture of trace elements lacking any sulphate salt was used . For preparation of porcine lung agar ( PLA ) culture medium , 5 g of fresh tissue was snap-frozen in liquid nitrogen and pulverized using a pre-cooled mortar . The resulting powder was put into a 50 ml reaction tube , filled up with an equal amount of sterile saline and briefly incubated in a 50°C warm water bath , followed by addition of an equal amount of 50°C warm , liquid water agar ( 1 . 5% agarose in water ) . To suppress bacterial growth , PLA media was supplemented with 50 µg•ml−1 tetracycline . The suspension was finally vortexed for 10 s and poured onto solidified water agar . For all growth assays on solid media , the culture medium was inoculated with 10 µl of a freshly prepared A . fumigatus spore suspension ( 105 conidia•ml−1 in water supplemented with 0 . 9% NaCl and 0 . 02% Tween 80 ) and incubated at 37°C for three days . Phenotypic microarray plates ( Biolog PM 4 ) were inoculated as follows: a suspension containing 1 . 5×106 spores•ml−1 in sulphur-free minimal medium was prepared and 100 µl aliquots of this were added to each well . Fungal growth was measured at 48 hours by optical density ( O . D . ) at 630 nm in a Multiskan Ascent microplate photometer ( Thermo Electron ) . For the measurement of germination percentages , 107 conidia were inoculated in 200 ml sulphur-free minimal media supplemented with 2 mM sulphate or 5 mM methionine and incubated at 37°C and 150 r . p . m for 11 hours . Each hour a 1 ml aliquot was taken , sonicated , and the germination percentage calculated as the ratio of germinated conidia with respect to the total number of spores . A . fumigatus liquid media shifts were performed according to Narendja et al . [88] with adjusted media to modify the sources of sulphur: 200 ml minimal medium lacking sulphur and supplemented with 5 mM methionine were inoculated with 108 freshly harvested , 5-day-old A . fumigatus ATCC 46645 , metRΔ , or metR+ conidia and propagated at 37°C and 150 r . p . m . for 16 to 22 hours . Mycelia from such pre-cultures were then harvested , washed extensively with water , and split into similar aliquots on a sterile surface . These were then added to 100 ml of minimal medium base without sulphur source or supplemented with 2 mM SO42− , 5 mM methionine , or 5 mM cysteine , respectively , and incubated at 37°C and 150 r . p . m for one to eight hours . Standard protocols of recombinant DNA technology were carried out [89] . Phusion high-fidelity DNA polymerase ( Fermentas ) was generally used in polymerase chain reactions and essential cloning steps were verified by sequencing . Fungal genomic DNA was prepared following the protocol of Kolar et al . [90] and Southern analyses were carried out as described [91] , [92] . Probes for non-radioactive hybridizations were generated and detected using the Gene Images AlkPhos Direct Labelling and Detection System from GE Healthcare . Samples of total RNA were isolated with the TRIzol reagent ( Sigma ) and cleaned with peqGOLD phase-trapA ( peqlab ) . RNA samples for RNA-seq were further purified with RNeasy Plant Mini Kit columns ( Qiagen ) . For Northern hybridisation analyses , 10 µg of total RNA were separated in formaldehyde-containing agarose gels , blotted onto Hybond-N+ membranes ( Amersham Biosciences ) , and hybridized with digoxigenin-labeled probes prepared as recommended by the manufacturer ( Boehringer Mannheim ) . Templates for hybridization probes were generated by PCR amplification using oligonucleotides listed in Table S3 . Autoradiographies were produced by exposing washed membranes to Fujifilm RX films . Binding of MetR to promoter regions of selected candidate targets was interrogated following the chromatin immunoprecipitation ( ChIP ) approach together with the GFP-Trap technology ( ChromoTek ) and using strain AfS171 that carries a codon-optimised version of the gfp2-5 allele [93] from pSK494 [94] preceded by a ( GA ) 5 linker region [95] fused C-terminally to the metR coding sequence . Essential steps were carried out following the protocol of [96] with modifications ( see Supporting Information for details ) , enrichment for distinct fragments was probed by semi-quantitative PCRs with specific primer pairs covering respective promoter regions . Digital transcriptomes of A . fumigatus strains ATCC 46645 and AfS167 [metRΔ] were produced by Eurofins MWG Operon GmbH from two independent biological replicates , which underlay the high reproducibility of this experimental approach [97] . For this purpose , 3′-fragment-specific cDNA libraries were prepared from poly ( A ) -fragment selected mRNA and processed on the Illumina HiSeq 2000 sequencing system using v3 . 0 chemistry and the 1×100 bp single read module , which ensures a high significance in respect to the copy number of each transcript . Mapping of reads on the most recent reference genome sequence ( http://www . aspergillusgenome . org/ ) was performed using BMA , SamTools and Picard software . To enable the direct comparison between the samples , the read count per reference has been normalized as follows: ( total_mapped_reads_per_refseq/number_of_reads_in_sample ) *lowest_total_sample_read_count and differential expression analyses were carried out using DESeq [98] . A 1 . 5-fold change between the average number of reads in the wild-type and metRΔ strain was used as threshold to define genes which are expressed higher or lower , respectively , in the mutant . Functional characterization of these regulated genes was performed on the FungiFun website ( https://sbi . hki-jena . de/FungiFun/FungiFun . cgi ) [55] based on both the FunCat method and the Gene Ontology ( GO ) classification [99] . Infections of larvae from the greater waxmoth Galleria mellonella were performed according to Kavanagh & Fallon [100] . Larvae were injected with varying doses of conidia in a saline solution supplemented with 0 . 02% Tween 80 and 10 µl•ml−1 rifampicin , to avoid bacterial infections , and incubated at 30°C . Female mice ( CD1 or BALB/c from Charles Rivers Breeding Laboratories , Sulzfeld , Germany ) of 20 to 24 g were used for infection experiments . Immunosuppression was carried out by subcutaneous injection of 112 mg•kg−1 hydrocortisone acetate and intraperitoneal injection of 150 mg•kg−1 cyclophosphamide following a sequential protocol as previously described [101] , with the modification that two doses of cortisone on days −3 and −1 were applied . Bacterial infections were prevented by adding 2 g•l−1 neomycin to the drinking water . Inocula were prepared by harvesting conidia from 5-day-old slants of solid medium followed by filtration through Miracloth tissue and washing with saline . Mice were anesthetized by intraperitoneal injection of a ketamine ( 1% ) /xylazine ( 0 , 2% ) solution and either infected intranasally by instillation of 2•105 conidiospores suspended in 40 µl of saline or intravenously by injection of a 50 µl suspension of 1•105 conidia into the lateral tail vein . Disease progression was followed twice daily by tabulating weight profiles and following the animals' behaviour . Signs of respiratory distress , hunched posture or poor mobility , as well as severe weight loss of more than 20% determined the experimental end point for each animal . To evaluate mortality rates in single-strain infection experiments , the log rank method was applied using the GraphPad Prism software . For competitive index ( CI ) assessment [13] , mice were intranasally infected with a 1∶1 mixture of 2•104 conidia from ATCC 46645 and the metRΔ strain AfS167 and sacrificed after four days . The lungs were explanted and aliquots from homogenized tissue were spread onto media containing or lacking methionine as sulphur source to differentiate between the wild-type progenitor and its metRΔ deletant . The CI is defined as the output ratio of mutant to wild-type fungal colonies divided by the input ratio of mutant to wild-type fungal colonies [102] . Histological cryo-sectioning was performed on 4% formaldehyde-fixed lungs , staining procedures with hemotoxylin and eosin together with Grocott's Methenamine Silver were carried out according to standard protocols . Five sections each from four lungs of mice infected with the wild-type and metRΔ strain , respectively , were inspected to yield representative images . Quantification of the intracellular siderophore ferricrocin ( FC ) was carried out as described earlier [40] , [103] . Samples for total iron content measurements were lyophilized and 50 mg digested in 500 µl 60% HNO3 ( Ultrapure , Merck ) for 4 h at 110°C and diluted thereafter with ultrapure water ( Milli-Q ) . Iron was quantified by graphite furnace atomic absorption spectrometry ( M6 Zeeman GFAA-Spectrometer , Thermo Scientific ) at 248 . 3 nm and D2-Quadline background correction using 1000°C ash temperature and 2100°C atomization temperature under argon atmosphere . The iron content was calculated by interpolation from an appropriate standard curve ( 0 . 5 to 12 . 5 µg•l−1 ) using TraceCERT ( Sigma-Aldrich ) standard solution . The accuracy of analysis was assessed by simultaneous analysis of a standard reference human serum sample ( ClinChek , Recipe ) .
|
Invasive pulmonary aspergillosis ( IPA ) is a life-threatening disease that affects primarily immunosuppressed patients . During the last decades the incidence of this disease that is accompanied by high mortality rates has increased . Since opportunistic pathogenic fungi , unlike other pathogens , do not express specific virulence factors , it is becoming more and more clear that the elucidation of fungal metabolism is an essential task to understand fungal pathogenicity and to identify novel antifungal targets . In this work we report genetic inactivation of the sulphur transcription regulator MetR in Aspergillus fumigatus and subsequent study of the resulting phenotypes and transcriptional deregulation of the mutant . Here we show that regulation of sulphur assimilation is an essential process for the manifestation of IPA . Moreover , a regulatory connection between sulphur metabolism and iron homeostasis , a further essential virulence determinant of A . fumigatus , is demonstrated in this study for the first time . A deeper knowledge of sulphur metabolism holds the promise of increasing our understanding of fungal virulence and might lead to improved antifungal therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"microbial",
"metabolism",
"mycology",
"medical",
"microbiology",
"biology",
"microbiology",
"pathogenesis",
"fungal",
"physiology"
] |
2013
|
Regulation of Sulphur Assimilation Is Essential for Virulence and Affects Iron Homeostasis of the Human-Pathogenic Mould Aspergillus fumigatus
|
Iron is an essential nutrient for most bacterial pathogens , but is restricted by the host immune system . Mycobacterium tuberculosis ( Mtb ) utilizes two classes of small molecules , mycobactins and carboxymycobactins , to capture iron from the human host . Here , we show that an Mtb mutant lacking the mmpS4 and mmpS5 genes did not grow under low iron conditions . A cytoplasmic iron reporter indicated that the double mutant experienced iron starvation even under high-iron conditions . Loss of mmpS4 and mmpS5 did not change uptake of carboxymycobactin by Mtb . Thin layer chromatography showed that the ΔmmpS4/S5 mutant was strongly impaired in biosynthesis and secretion of siderophores . Pull-down experiments with purified proteins demonstrated that MmpS4 binds to a periplasmic loop of the associated transporter protein MmpL4 . This interaction was corroborated by genetic experiments . While MmpS5 interacted only with MmpL5 , MmpS4 interacted with both MmpL4 and MmpL5 . These results identified MmpS4/MmpL4 and MmpS5/MmpL5 as siderophore export systems in Mtb and revealed that the MmpL proteins transport small molecules other than lipids . MmpS4 and MmpS5 resemble periplasmic adapter proteins of tripartite efflux pumps of Gram-negative bacteria , however , they are not only required for export but also for efficient siderophore synthesis . Membrane association of MbtG suggests a link between siderophore synthesis and transport . The structure of the soluble domain of MmpS4 ( residues 52–140 ) was solved by NMR and indicates that mycobacterial MmpS proteins constitute a novel class of transport accessory proteins . The bacterial burden of the mmpS4/S5 deletion mutant in mouse lungs was lower by 10 , 000-fold and none of the infected mice died within 180 days compared to wild-type Mtb . This is the strongest attenuation observed so far for Mtb mutants lacking genes involved in iron utilization . In conclusion , this study identified the first components of novel siderophore export systems which are essential for virulence of Mtb .
Iron is an essential micronutrient for most forms of life on earth because of its vital role as a redox cofactor of proteins required for critical cellular processes . Pathogenic bacteria have evolved an array of intricate mechanisms to scavenge limited iron from the host [1] . Mycobacterium tuberculosis ( Mtb ) , one of the most successful human bacterial pathogens , is no exception . Mtb meets its iron demands by stripping host iron stores employing two hydroxyphenyloxazoline siderophores , mycobactin ( MBT ) and carboxymycobactin ( cMBT ) . To counteract these bacterial iron acquisition processes , the alveolar macrophage in which Mtb thrives , keeps phagosomal iron levels extremely low by the natural resistance-associated macrophage protein Nramp1 in particular after activation by interferon-γ [2] , [3] . MBT and cMBT increase the biologically available iron within the phagosomal compartment almost by 20-fold indicating that the Mtb siderophores can overcome these host defense mechanisms [4] . Furthermore , studies using siderophore biosynthesis and uptake mutants underpin the importance of siderophore-mediated iron acquisition to the virulence of Mtb [5] , [6] , [7] . In Mtb , siderophore biosynthesis is induced under low-iron conditions . When sufficient iron is available the regulator IdeR represses expression of MBT biosynthesis genes mbtA-N . The inner membrane transporter IrtAB and the Esx-3 secretion machinery are required for utilization and uptake of siderophores [6] , [7] , [8] . In M . smegmatis , export of the siderophore exochelin was shown to be mediated by ABC-like exporter ExiT [9] . Given mycobacteria's unique outer membrane [10] , it is likely that a siderophore secretion system of Mtb requires both inner and outer membrane components [11] , similarly to the EntS-TolC system of E . coli [12] , [13] . In this study , we examined two iron-regulated genes encoding predicted outer membrane proteins MmpS4 and MmpS5 . We show that either MmpS4 or MmpS5 is required for growth of Mtb under low iron conditions . While single mmpS4 or mmpS5 deletion mutants do not exhibit a low iron growth phenotype , they have diminished virulence compared to the wild-type strain . Deletion of both mmpS4 and mmpS5 drastically decreases synthesis and secretion of siderophores in Mtb and greatly reduces its virulence in mice . Subcellular fractionation reveals that MmpS4 and MmpS5 are membrane associated . This study identifies MmpS4 and MmpS5 as the first components of a novel siderophore export system that is crucial for survival of Mtb in its host .
To investigate whether MmpS4 and MmpS5 are important for growth under iron-deplete conditions , mutants with in-frame deletions of the mmpS4 and mmpS5 genes were constructed using homologous recombination in both virulent Mtb H37Rv and avirulent Mtb mc26230 ( ΔRD1 , ΔpanCD; Table S1 ) . Since no low-iron growth defect was observed with the single ΔmmpS4 and ΔmmpS5 mutants ( Figs . 1 , S1–2 ) , and expression of both mmpS4 and mmpS5 is induced under iron-limited conditions [14] we suspected that they might have redundant functions . Therefore , we constructed a double mmpS4/mmpS5 mutant using the single ΔmmpS5 mutant as the parent strain in both virulent and avirulent Mtb . The mutant strains were unmarked by site-specific recombination and confirmed by Southern blot analysis ( Fig . S3 ) . Western blot experiments demonstrated the absence of MmpS4 and MmpS5 in the double mutant and in the respective single mutants , while wild-type levels of both proteins were observed in the complemented strains ( Figs . 1A , S4 ) . No differences in growth of the single ΔmmpS4 and ΔmmpS5 strains were observed on self-made low iron glycerol-alanine salts ( GAS ) agar plates ( Fig . 1B ) . By contrast , the ΔmmpS4/S5 mutant did not grow on GAS agar plates ( Fig . 1B ) . Growth of the double mutant was partially rescued when GAS agar plates were supplemented with 5 µM hemoglobin that was previously shown to function as an iron source for Mtb ( Fig . 1C ) [15] . However , in liquid medium , the addition of hemin completely rescued the growth of the ΔmmpS4/S5 mutant ( Fig . S1 ) verifying that the growth defect of this strain is indeed iron dependent . Complementation of the ΔmmpS4/S5 mutant with mmpS4 and mmpS5 restored growth on low iron plates to wt levels ( Fig . 1B ) . Interestingly , blocking siderophore biosynthesis by insertion of a resistance cassette into mbtD in the ΔmmpS4/S5 double deletion mutant ( ΔmmpS4/S5/ΔmbtD::hyg strain ) also restored growth on hemoglobin plates to the level of the wt strain ( Fig . 1C ) . These results indicate that siderophore biosynthesis impairs the growth of this mutant despite the availability of an alternative iron source . To provide further evidence for the iron growth defect of the ΔmmpS4/S5 double mutant , growth experiments were conducted in various low iron conditions that included self-made low iron 7H9 medium , or the addition of 2 , 2′-dipyridyl ( DIP ) or desferrioxamine ( DFO ) as ferrous and ferric specific chelators , respectively , to standard 7H9 medium ( Figs . S1–2 , S5–6 ) . Under each low iron growth condition , ΔmmpS4/S5 failed to grow . Unlike on solid media , in iron-replete liquid media , ΔmmpS4/S5 had only a slightly delayed growth phenotype and eventually reached optical densities equal to the wt strain . It is concluded that deletion of mmpS4 and mmpS5 confers a low-iron growth defect phenotype in Mtb . The inability of Mtb ΔmmpS4/S5 to grow under iron-limiting conditions may be due to defects in siderophore biosynthesis , iron sensing , uptake or secretion of siderophores . Recently , a biosynthetic pathway has been proposed based on the substrate specificities of enzymes encoded by the mbt gene cluster [16] which accounts for all enzymatic activities required for MBT biosynthesis . Therefore , it is unlikely that MmpS4 and MmpS5 play a direct role in biosynthesis of MBT and cMBT . To test whether the ability of the ΔmmpS4/S5 mutant is impaired in sensing low iron conditions , we utilized a gfp-based iron-regulated reporter construct [17] . Under low iron conditions , transcription from IdeR-regulated promoters was induced in wt Mtb containing the reporter construct as indicated by a strongly increased GFP fluorescence ( Fig . S7 ) . However , when wt Mtb was grown under high iron conditions , only background fluorescence was observed confirming that Mtb senses iron availability ( Fig . 2A ) . To examine the iron sensing capability of the ΔmmpS4/S5 mutant we exploited the observation that removal of the antibiotic resistance cassette from the MBT biosynthesis mutant ML1600 ( ΔmbtD::hyg ) [18] resulted in the strain ML1610 ( ΔmbtD::loxP ) ( Table S1 ) with only a partial low-iron growth defect in vitro ( Fig . S8 ) . This result suggests that replacing mbtD with the hyg cassette inhibits expression of downstream genes thereby completely eliminating siderophore production . IdeR-regulated promoters are induced under high-iron conditions in Mtb ΔmbtD::loxP , but the addition of exogenous cMBT to this mutant repressed these promoters , demonstrating that this mutant is capable of sensing iron availability and is suitable as a control strain ( Fig . 2A ) . Likewise , IdeR-regulated promoters were induced in the ΔmmpS4/S5 mutant under high iron conditions , but were repressed after addition of cMBT ( Fig . 2A ) , demonstrating that the ΔmmpS4/S5 mutant is capable of sensing iron availability . To test whether MmpS4 and MmpS5 are involved in siderophore uptake we monitored the accumulation of 55Fe-loaded cMBT . Iron-loaded cMBT has been shown to donate its iron to MBT in the cell envelope of Mtb in addition to being taken up via the inner membrane ABC-transporter IrtAB [8] , [19] . To rule out the possibility that differences in MBT levels affected the measured iron uptake rates , we examined 55Fe-cMBT uptake at 37°C in the siderophore biosynthesis mutant ΔmbtD::hyg and the triple mutant ΔmmpS4/S5/ΔmbtD::hyg ( Fig . 2B ) . Despite the absence of MBTs/cMTBs no differences were observed in the amount of iron accumulated by these strains . Another control experiment showed that only background 55Fe levels were associated with cells at 4°C , indicating that the cell-associated 55Fe observed at 37°C was indeed transported inside the cell and not adsorbed at the cell surface ( not shown ) . Taken together , these results demonstrate that MmpS4 and MmpS5 are not involved in uptake of cMBT . The low iron growth defect of the ΔmmpS4/S5 mutant is not caused by an iron sensing defect or by lack of cMBT uptake . An alternative explanation might be a defect in secretion of cMBT . To this end , cMBTs in wt Mtb and in the ΔmmpS4/S5 mutant were radioactively labeled by feeding the bacteria the biosynthetic precursor 7-[14C]-salicylic acid . Cell-associated and secreted siderophores were extracted using chloroform from cell pellets and from the culture filtrate and analyzed by thin-layer chromatography ( TLC ) . As controls , purified and deferrated MBTs/cMBTs from M . bovis BCG were loaded with 55Fe and used to visualize siderophore spots . TLC analysis demonstrated that purified siderophores from M . bovis BCG had the same Rf values as siderophores from Mtb validating their use as controls ( Fig . 3 ) . The extracts of cell pellets and of culture supernatants showed that the single deletion mutants Mtb ΔmmpS4 and ΔmmpS5 synthesized and secreted siderophores as wt Mtb ( Fig . 3 ) . By contrast , the double deletion mutant ΔmmpS4/S5 produced much less cell-associated and secreted siderophores compared to wt Mtb , but was still capable of synthesizing siderophores in contrast to the ΔmbtD::hyg mutant ( Fig . 3 ) . It should be noted that MBT was detected in the culture supernatants of Mtb in addition to cMBT . This is most likely caused by partitioning of cell surface-associated MBT with the medium in the presence of detergents . Taken together , these results suggest that MmpS4 and MmpS5 are part of siderophore export system of Mtb . MmpS4 in M . smegmatis was previously shown to be involved in biosynthesis and export of glycopeptidolipids ( GPLs ) [20] which Mtb does not synthesize . To examine whether the deletion of mmpS4 and mmpS5 caused an altered lipid profile , we performed a complete lipid analysis by TLC ( Figs . S9 , S10 ) . All major lipids of Mtb were identified in wt Mtb and the ΔmmpS4/S5 mutant ( Fig . S9 , S10 ) indicating that MmpS4 and MmpS5 are not involved in lipid biosynthesis . However , a lipid which was shown to be produced by Mtb under iron limiting conditions [21] was not identified in the ΔmmpS4/S5 mutant ( Figs . S10A–C ) . Bacon et al . [20] showed by 1H-NMR that this lipid consists of a long alkyl chain with a cis double bond and an ester unit . It is unclear whether the absence of this lipid is a direct consequence of the lack of MmpS4/S5 , or might be caused indirectly by the slow growth of the double mutant under iron-limiting conditions . Our data suggests that MmpS4 and MmpS5 are involved in siderophore export , but it is unclear how MmpS4 and MmpS5 contribute to MBT transport . Proteomic analysis of subcellular fractions of Mtb yielded contradictory results regarding the localization of MmpS4 and MmpS5 [22] , [23] . In order to determine the subcellular localization of MmpS4 and MmpS5 , the culture filtrate containing secreted proteins was prepared . Membrane and cytoplasmic fractions were obtained by ultracentrifugation of cell lysates of Mtb . Both MmpS4 and MmpS5 were present in the membrane , but not in the cytoplasmic or secreted fractions ( Fig . 4A ) . All fractions were well separated as indicated by the control proteins , the membrane-associated OmpATb ( Rv0899 ) , the cytoplasmic regulator IdeR and the secreted Ag85 protein ( Fig . 4A ) . Thus , MmpS4 and MmpS5 are the first examples of membrane-associated proteins that are required for export of siderophores in Mtb . The strongly reduced MBT/cMBT level is a striking phenotype considering the intact biosynthesis capacity of the Mtb ΔmmpS4/mmpS5 mutant . Based on the previous observation that MmpS4 connects glycopeptidolipid biosynthesis enzymes with the MmpL4 transporter in M . smegmatis [20] we hypothesized that MmpS4 might provide a link between MBT/cMBT biosynthesis and export in Mtb . However , in vivo crosslinking experiments with formaldehyde in the avirulent Mtb strain mc26230 ( Table S1 ) expressing a chromosomal copy of a gene encoding hexahistidine- and HA-tagged MbtG did not show direct binding of MbtG to MmpS4 . Next , we examined the subcellular localization of MbtG , the lysine monooxygenase which activates MBT/cMBT by hydroxylating dideoxymycobactins as the predicted last step in MBT biosynthesis [24] . In order to catalyze this reaction MbtG has to be in the cytoplasm because it requires access to the cytoplasmic co-factors NADPH and FAD+ . Subcellular fractionation experiments in wt Mtb mc26230 revealed that MbtG is membrane-associated although no transmembrane helices and no signal peptide are apparent ( Fig . 4B ) . This result indicates that MbtG might fractionate with membranes due to interactions with another protein and provides the first hint how MBT/cMBT biosynthesis and export might be coupled in Mtb . The mmpS genes are located in operons with mmpL genes [25] . In order to genetically determine if MmpS4 and MmpS5 interact with their cognate MmpL proteins , the triple mutants ΔmmpS4/L4/S5 and ΔmmpS4/S5/L5 were constructed from the ΔmmpS5 and ΔmmpS4 strains , respectively , by deleting the respective mmpSL operon ( Fig . S11 ) . Similar to the double deletion mutant ΔmmpS4/S5 , these triple mutants failed to grow in iron-deplete media ( Fig . 5A ) . To this end , each triple mutant was complemented with either an empty integrative vector or integrative vectors containing either mmpS4 or mmpS5 . The mmpL5 containing strain ( ΔmmpS4/L4/S5 ) complemented with either mmpS4 or mmpS5 grew in low iron medium ( Middlebrook 7H9 supplemented with 50 µM DIP ) ( Fig . 5A ) . However , the mmpL4 containing strain ( ΔmmpS4/S5/L5 ) was only complemented with mmpS4 but not with mmpS5 . These results indicate that MmpL4 only interacts with its cognate MmpS4 protein , while MmpL5 is capable of interacting with MmpS4 and MmpS5 to mediate siderophore export by Mtb . To confirm and further define the interaction between MmpS4 and MmpL4 , an in vitro pull-down assay was employed . According to the topology predictions MmpS4 possesses an N-terminal transmembrane ( TM ) helix and a C-terminal soluble domain , while MmpL4 contains eleven TM helices and two long loops—L1 between TM1 and TM2 , and L2 between TM6 and TM7 ( Fig . S12 ) . We tested the interaction between the purified soluble domains of MmpS4 ( residues 52–140 ) and the predicted loops L1 ( 58–199 ) and L2 ( 416–763 ) of MmpL4 . The soluble domain of MmpS4 formed a complex with loop L1 ( Fig . 5B ) , but not with loop L2 ( data not shown ) of MmpL4 . The in vitro interaction of the soluble domain of MmpS4 with loop L1 of MmpL4 also shows that both peptides form independently folding domains . The mmpS4 gene encoding an N-terminally truncated MmpS4 protein lacking the predicted transmembrane helix was expressed in E . coli and the water-soluble domain of MmpS4 ( residues 52–140 ) was purified by chromatography . The structure of MmpS452–140 was solved by NMR using 762 nuclear Overhauser effect ( NOE ) and 127 paramagnetic relaxation enhancement ( PRE ) distance restraints , and 122 dihedral angle restraints ( Table S4 ) . The 20 lowest energy structures were selected out of 200 accepted structures . The statistics about the quality and precision of these structures is summarized in Table S4 . The backbone superimposition of the final 20 conformers and the representative structure are presented in Fig . 6A . The MmpS4 structure shows seven consecutive β-strands and an unstructured C-terminus ( residues 131–140 ) ( Fig . 6B ) which might be due to the lack of resonance assignment in this region . The seven β-strands are arranged in two layers , with β4-β1-β6-β7 in one layer and β3-β2-β5 in the other layer . To assess the role of MmpS4 and MmpS5 for virulence of Mtb , BALB/c mice were infected with low dose aerosols containing the Mtb H37Rv parent strain ML617 , the ΔmmpS4/S5 mutant ( ML618 ) , and the double deletion mutant complemented with mmpS5 ( ML619 ) , mmpS4 ( ML620 ) , and mmpS4/S5 ( ML624 ) . The growth kinetics of the parent Mtb H37Rv strain in lungs showed the expected logarithmic increase during the acute phase followed by a plateau during the chronic phase of infection . Similar growth kinetics in spleens demonstrated that this strain is competent for dissemination . Loss of the single mmpS4 and mmpS5 genes also compromised the ability of Mtb to survive in the lungs as the number of viable bacteria decreased by 100-fold from the initial burden compared to wt Mtb . However , loss of these genes alone did not alter the ability of Mtb to disseminate to and proliferate in the spleen . The ΔmmpS4/S5 mutant failed to proliferate in lungs and spleen as reflected by a 24 , 000- and 1 , 800- fold , respectively , decreased bacterial burden compared to wt Mtb after 16 weeks of infection ( Fig . 7 ) . Loss of these genes resembles the “severe growth in vivo” ( sgiv ) phenotype [26] and , to our knowledge , is the strongest in vivo phenotype observed so far for genes involved in iron utilization by Mtb . The single mmpS4 or mmpS5 genes partially complemented the virulence defect of the double mutant . Full complementation of the double mutant by both genes confirmed that the mmpS4 and mmpS5 genes are essential for virulence of Mtb ( Fig . 7 ) . Gross mouse lung examination and histological assessment in mice infected with the ΔmmpS4/S5 double deletion strain showed almost no signs of infection ( Figs . 8 , S13–15 ) . However , lungs of mice infected with either Mtb H37Rv wt or the fully complemented Δmmps4/S5 strain exhibited extensive lesions ( Figs . 8 , S13 ) and displayed significant lymphocytic infiltrates ( Fig . S14 ) . Lungs of mice infected with the mmpS4 or mmpS5 singly complemented strains showed lesions and lymphocytic infiltrates , but to a much lesser degree than lungs of mice infected with wt or the fully complemented strain . In survival experiments loss of mmpS4 and mmpS5 severely attenuated virulence of Mtb as none of the mice infected with the ΔmmpS4/S5 double deletion mutant died within 180 days ( Fig . 9 ) . Similarly , loss of either mmpS4 or mmpS5 alone resulted in attenuation of virulence . The difference in mean survival time between the groups of mice infected with wt and the fully complemented strain was longer than expected and could partly be explained by a lower bacterial burden of the fully complemented strain in the lungs . In conclusion , the infection experiments revealed that mmpS4 and mmpS5 are essential for virulence of Mtb in mice .
In this study , we showed that the lack of MmpS4 and MmpS5 strongly reduced siderophore secretion and caused a growth defect of Mtb under low iron conditions . Pull-down experiments demonstrated that the MmpS4 protein forms a complex with the inner membrane transporter MmpL4 in vitro . This observation was corroborated by genetic complementation experiments demonstrating that MmpS4 and MmpS5 interact with their respective MmpL proteins to restore growth of Mtb under iron-limiting conditions . Considering that siderophore uptake is not altered in Mtb lacking mmpS4/S5 and that the MmpL proteins are inner membrane transporter proteins , it is concluded that the respective MmpS/MmpL complexes translocate siderophores across the inner membrane into the periplasmic space . Such a transport process is defined as export [27] . Proteins which enable siderophore export in Mtb have been unknown so far [11] , largely because Mtb does not have any proteins resembling known siderophore export systems such as EntS of Escherichia coli [13] or PvdE of Pseudomonas aeruginosa [28] . Previously , MmpS5 and MmpL5 have been implicated in drug efflux due to weak similarity with RND efflux pumps of E . coli [29] . MmpL3 , MmpL7 and MmpL8 were shown to export lipids such as trehalose monomycolate [30] , [31] , [32] , phthiocerol dimycocerosate [33] , [34] , and sulfolipid 1 [35] leading to the hypothesis that the MmpL proteins are lipid transporters . Since carboxymycobactins and in particular mycobactins are quite hydrophobic molecules and have similar chemical properties as lipids , this finding is rather an expansion than a deviation from the rule . Taken together , we conclude that MmpS4/MmpL4 and MmpS5/L5 constitute novel bacterial siderophore export systems . The lack of the MmpS4/5 proteins also reduced the amount of detectable carboxy/mycobactins suggesting a role in biosynthesis of these siderophores in Mtb . Recently , a biosynthetic pathway comprising all enzymatic activities required for MBT/cMBT biosynthesis has been proposed based on the substrate specificities of enzymes encoded by the mbt operons [16] . Modifying enzymes to generate nonmethylated or α-methylated MBT derivatives have not been identified yet , but they are not expected to alter the total MBT amount . Thus , it is concluded that the strongly reduced siderophore levels in the Mtb ΔmmpS4/S5 mutant likely result from an indirect effect of these proteins on biosynthesis . Indeed , such a mechanism has been proposed for the MmpS4 protein which is required for efficient synthesis and export of surface-exposed glycopeptidolipids ( GPL ) in M . smegmatis [20] . Co-localization of MmpS4 with FadD23 and MbtH indicated that the GPL biosynthesis enzymes form a multi-protein complex with the membrane proteins MmpS4 and MmpL4a/b in M . smegmatis . Since lack of MmpS4 resulted in enzyme diffusion in the cytoplasm , biosynthesis of GPLs was much less efficient in M . smegmatis [20] . This phenotype was complemented by the Mtb mmpS4 gene indicating that Mtb MmpS4 also enables formation of a biosynthetic multi-enzyme complex at the inner membrane of M . smegmatis . In this study we show , that MmpS4 is involved in siderophore export in Mtb . The fact that the siderophore biosynthesis enzyme MbtG is located at the inner membrane , as shown in this study , supports the hypothesis that a similar multi-enzyme complex for efficient siderophore synthesis and transport exists in Mtb . In principle , block of transport caused by the mmpS4/mmpS5 deletions and degradation of siderophores as is observed for the ferric enterobactin esterase IroD , IroE , and Fes of E . coli and Salmonella [36] would also explain the low level of MBTs/cMBTs in the Mtb mmpS4/mmpS5 double mutant . However , there are no enzymes in Mtb with similarities to known siderophore esterases . In addition , degradation of imported siderophores to release iron in the cytoplasm is rare and has only been observed for trilactone siderophores such as enterobactin of E . coli and Salmonella [37] . The high energy cost of MBT/cMBT production [38] , [39] also argues in favor of a synthesis tightly regulated by the requirement for iron and the capacity to export newly synthesized siderophores . Coupled synthesis and export would also prevent toxic accumulation of siderophores in Mtb as has been observed in other bacteria [40] , [41] , [42] . An interesting question is whether the MmpL4/MmpS4 or the MmpL5/MmpS5 systems are specific for MBTs or cMBTs . The single mutants clearly produce and secrete both siderophores indicating that the MmpS4/MmpS5 proteins are not specific for either substrate . This conclusion is supported by the observation that the Mtb mmpS4 gene complements the GPL synthesis and transport defect of the M . smegmatis mmpS4 mutant , although GPLs do not exist in Mtb [20] . It is more likely that the transporters themselves , namely MmpL4 and MmpL5 , confer specificity for MBTs or cMBTs . This hypothesis is currently under investigation . Interestingly , we observed that both MmpS4 and MmpS5 interact with MmpL5 , while MmpS5 cannot restore growth of an Mtb triple mutant ΔmmpS4/S5/L5 expressing only mmpL4 under iron-limiting conditions . Thus , MmpS4 seems to be more promiscuous in its interactions with MmpL proteins . In this regard , it should be noted that the genetic complementation experiments indicate that the MmpS5/MmpL5 pair is more efficient in restoring wt growth of Mtb under low iron conditions . In conclusion , it appears that Mtb ensures efficient siderophore export by employing at least two partially redundant transporters . The NMR structure of MmpS452–140 revealed no similarity to any protein of known function , but was similar to an uncharacterized protein from Parabacteroides distasonis ( PDB: 2LGE ) . The superimposition of the MmpS4 structure with that of this putative calcium-binding protein showed a root-mean-square deviation of 3 . 6 Å over the Cα atoms of 75 aligned residues ( Fig . S15A ) with similar secondary and tertiary structures ( Fig . S15B ) . Secondary structure prediction [43] indicated an eighth β-strand including the residues 131–137 of MmpS4 . This gave rise to the hypothesis that the C-terminus of MmpS4 might be unordered in its unbound state , but may form a more stable structure with two 4-stranded sheets when bound to MmpL4 . Further experiments are required to provide evidence for this hypothesis . Importantly , the structure of MmpS4 shows no similarity to AcrA [44] or other periplasmic adapter proteins from drug efflux systems of Gram-negative bacteria [45] indicating that the mycobacterial MmpS proteins constitute a novel class of accessory proteins in complex transporter systems . In this study , we identified a novel siderophore export system of Mtb which is composed of the transporters MmpL4 and MmpL5 and their associated MmpS proteins . Previously , it was proposed that the MmpS proteins function as periplasmic adapter proteins [29] which was based on the low sequence similarities between the transporters of tripartite efflux pumps of Gram-negative bacteria [46] with MmpL proteins [47] . The localization of MmpS4 in the periplasm of M . smegmatis [20] and our observation that the MmpS4/5 proteins interact with their respective MmpL transporters support their role as accessory transport proteins . In addition , we show that , in contrast to their Gram-negative counterparts , the MmpS4/MmpS5 proteins are not only required for export , but also for biosynthesis of cMBT/MBT . Therefore , we hypothesize that MmpS4 functions as a scaffolding protein to couple synthesis and export of MBT/cMBT in Mtb as has been proposed for GPLs in M . smegmatis [20] . The surprising result that the MBT/cMBT activating enzyme MbtG is membrane-associated despite the absence of any recognizable membrane anchor domain suggests that MbtG might interact directly or indirectly with membrane proteins such as MmpL4/L5 in Mtb . These findings are summarized in the model depicted in Fig . 10 . Hitherto unknown are the hypothetical outer membrane proteins required for cMBT/MBT secretion to the extracellular medium and for uptake of cMBT . The role of the Esx-3 system in cMBT/MBT-mediated iron acquisition is also unknown [7] . An interesting observation was that growth of the Mtb ΔmmpS4/S5 mutant under low iron conditions was not fully restored by adding hemoglobin as an iron source ( Fig . 1 ) . This result is in contrast to the Mtb ΔmbtD::hyg mutant which is unable to synthesize mycobactins [18] . However , the ΔmmpS4/S5 mutant grew like wt Mtb with hemoglobin as the sole iron source when MBT/cMBT biosynthesis was additionally eliminated . These results indicate that low level synthesis of siderophores and their intracellular accumulation due to the lack of export inhibits growth of the ΔmmpS4/S5 mutant , e . g . by chelating iron or other cations from essential proteins of Mtb . The mechanism of this peculiar type of growth inhibition is currently under investigation . Deletion of both mmpS4 and mmpS5 drastically reduced the virulence of Mtb in mice . Considering the strong growth defect of the ΔmmpS4/S5 mutant under low iron conditions in vitro and the known requirement of siderophore biosynthesis and utilization for growth of Mtb in vivo [5] , [6] , [7] , it is likely that the attenuation of the ΔmmpS4/S5 mutant is due to its inability to take up sufficient iron in the absence of siderophores . However , it cannot be excluded that other functions of MmpS4 and MmpS5 contribute to the virulence defect of the ΔmmpS4/S5 mutant . In favor of this conclusion is the observation that expression of mmpS5 fully restores siderophore export and growth of the Mtb ΔmmpS4/S5 mutant under low iron conditions in vitro , but still has a significant virulence defect in mice . This result indicated that other functions of MmpS4 , which are not present in MmpS5 , may contribute to the virulence defect of the ΔmmpS4/S5 mutant . Interestingly , the Mtb mmpL4 mutant showed a 10-fold reduced bacterial burden in the lungs of mice [47] . This is consistent with our finding that the number of bacteria of an Mtb strain which lacks only mmpS4 was between 10- and 100-fold lower in the lungs of mice after the acute phase of infection . Similarly , the slight attenuation of an Mtb strain which lacks only mmpS5 is consistent with the in vivo growth defect of an Mtb mmpS5 mutant in transposon site hybridization ( TraSH ) studies [48] . The loss of virulence of Mtb ΔmmpS4/S5 mutant in mice is much more pronounced than that observed for the irtAB mutant which lacks an ABC transporter required for cMBT uptake [6] . This correlates with the different magnitude of their in vitro phenotypes: While the irtAB mutant showed only a minor growth defect under low iron conditions [6] , loss of mmpS4 and mmpS5 completely abolished growth of Mtb under those conditions ( Figs . 1 , S1 , S2 ) . In this study , we identified that interaction of the membrane proteins MmpS4 and MmpS5 with their cognate MmpL transporters is required for siderophore export in Mtb and propose a model for siderophore secretion . These novel siderophore transport systems are essential for virulence of Mtb in mice . Considering the almost universal requirement of bacterial pathogens for iron [49] it is tempting to speculate that these systems might be good drug targets . However , more work is required to determine whether these two partially redundant transporters can be poisoned by a single drug .
BALB/c mice were obtained from the Charles River Laboratories and were housed and cared for in a pathogen-free biosafety level 3 vivarium facility at Johns Hopkins University . Mice were provided food and water ad libitum as well as appropriate monitoring and clinical care . The protocols used in this study were reviewed and approved by the Johns Hopkins Institutional Animal Care and Use Committee and are described in protocol MO09M101 . The Johns Hopkins Animal Care and Use Committee complies with Animal Welfare Act regulations and Public Health Service Policy . Johns Hopkins University also maintains accreditations with the Association for the Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) International . The strains used in this study are listed in the supplement ( Table S1 ) . Media , growth conditions and construction of plasmids are described in detail in the supplement ( Text S1 ) . The mmpS4 ( rv0451c ) , mmpS5 ( rv0677c ) , mmpS4/S5 , mbtD ( rv2831c ) , and mmpS4/S5/mbtD deletion mutants of Mtb H37Rv and Mtb mc26230 were constructed using a two-step selection strategy as described in SI . Complementation of ΔmmpS4/S5 double deletion mutant of Mtb H37Rv and Mtb mc26230 were performed using L5 and Ms6 phage integration systems as described in the supplement ( Text S1 ) . Low-iron GAS plates were prepared by dissolving 150 mg Bacto Casitone , 2 g K2HPO4 , 1 g citric acid , 0 . 5 g L-alanine , 0 . 6 g MgCl2•6H20 , 0 . 3 g K2SO4 , 1 g NH4Cl , and 8 . 3 ml 60% glycerol in 450 ml high grade Millipore water ( Barnstead Nanopure Diamond; 18 . 2 MΩ-cm ) , the pH was adjusted to 6 . 6 with NaOH and 5 g Agar Noble ( BD Biosciencese ) was added . The volume was brought up to 500 ml in an acid washed glass bottle ( 6 M HCl ) , autoclaved , supplemented with pantothenate , hygromycin , and split into acid washed bottles to which 5 µM human hemoglobin was added when required . Pre-cultures were grown in 7H9 Middlebrook medium supplemented with 10% OADC , 0 . 2% casamino acids , 24 µg/ml pantothenate , 50 µg/ml hygromycin , 0 . 02% tyloxapol ( 7H9 MR ) and 20 µM hemin . Once in mid-logarithmic phase ( OD600 = 0 . 5–2 . 0 ) cells were filtered through a filter with 5 µm pores and washed once in low-iron GAS medium . Cells were diluted to an OD600 = 0 . 01 and 10-fold serial dilutions were prepared in low-iron GAS medium . 3 µl of each dilution was deposited on low-iron GAS plates or low-iron + hemoglobin plates using a multi-channel pipette . Plates were incubated for nine weeks at 37°C . Strains were grown in 7H9 MR medium . Cultures were inoculated in biological triplicate , grown at 37°C and split at mid-logarithmic phase ( OD600 = 0 . 2–0 . 3 ) . Purified Fe-cMBT-BCG ( 1 µg/ml ) was added to one set of triplicates , while other triplicates were left untreated . Optical densities were determined in 1 cm path length cuvettes by diluting cells to OD600 = 0 . 1–1 in the above media . Readings were taken every day until stationary phase was reached . Fluorescence intensities reported in Fig . 2A were two days after the addition of Fe-cMBT . Green fluorescent protein ( GFP ) fluorescence intensities were determined using a Biotek Synergy HT plate reader with a 485 nm excitation and a 528−/+20 nm emission filter . Fluorescence intensities were normalized to the optical density of the same samples according to the following equation: Fe-cMBT-BCG ( 93 µg ) was deferrated as previously described by incubation in the presence of 50 mM EDTA pH = 4 . 0 at 37°C for 18 hours [8] . Precipitated EDTA was pelleted by centrifugation , supernatant was extracted twice with chloroform , washed twice with water and evaporated to dryness . Deferrated residue was suspended in a 1∶1 mixture of EtOH and 50 mM KH2PO4 buffer pH = 7 . 0 . 55Iron- ( 396 µCi ) was added to the mixture and incubated for 1 hour at room temperature ( at which point , the solution developed a brown hue ) . One ml of water was added to the mixture and extracted twice with 2 volumes of chloroform . The chloroform extract was washed twice with water and evaporate to dryness . The material was resuspended in warm ethanol . This preparation yielded 16 . 2 µM 55Fe-cMBT-BCG with the radioactive concentration of 47 . 5 µCi/ml . The strains ΔmbtD::hyg and ΔmmpS4/S5/ΔmbtD::hyg were grown in 7H9 MR medium , and 20 µM hemin to OD600 = 1 . 0 . Cells were washed on ice with a low iron media consisting of 500 µM MgCl2•6H20 , 7 µM CaCl2•2H2O , 1 µM NaMoO4•2H2O , 2 µM CoCl2•6H2O , 6 µM MnCl2•4H2O , 7 µM ZnSO4•7H2O , 1 µM CuSO4•5H2O , 15 mM ( NH4 ) 2SO4 , 12 mM KH2PO4 pH = 6 . 8 , 1% ( w/v ) glucose , which was supplemented with 10% OADC , and 0 . 2% casamino acids . Cells were resuspended in the same media to an OD600 of approximately 3 . 0 on ice . For uptake experiments , 2 ml of cell suspensions were equilibrated at 37°C for 15 min and shaken at approximately 400 rpm . 55Fe-labeled cMBT was added to the cells at a final concentration of 0 . 25 µM cMBT , 0 . 45 µCi 55Fe . 200 µl samples were removed at 1 , 2 , 4 , 8 , and 16 minutes and added to 400 µl of a killing buffer consisting of 100 mM LiCl , 50 mM EDTA in 4% formaldehyde in Spin-X filter microcentrifuge tubes . Cells were immediately centrifuged and washed twice in killing buffer . The radioactivity of the cells was quantified using liquid scintillation counting ( Beckman Coulter LS6500 ) . 55Fe counts were converted to total iron by the use of a standard curve and normalized to dry weight of cells by determining the dry mass of 4 ml of the washed cell suspensions . All experiments were done in triplicate . Radiolabelling of siderophores was performed in a similar manner as previously described with modifications [5] , [50] . Iron free self-made 7H9 media supplemented with 0 . 2% glucose and 0 . 01% Tyloxapol was deferrated using Chelex-100 to remove any trace contaminants of iron . Pre-cultures were grown under iron rich conditions to OD600 of 1–2 . To deplete intracellular iron stores , iron free media was inoculated with pre-culture to an OD600 = 0 . 05 , and allowed to grow to OD600 of 1–2 . Only ML618 ( ΔmmpS4/S5 ) and ML1424 ( ΔmbtD::hyg ) were not grown in iron free media because these strains do not grow under low iron conditions; however , IdeR derepression occurs even under iron-replete conditions in these strains ( this study ) . Five ml of cell cultures were adjusted to OD600 = 0 . 2 using iron free media and incubated with 1 µCi/ml [7-14C]-salicylic acid ( 21 . 3 µM final concentration ) for 11 days while shaking at 37°C . Cultures were centrifuged at 4 , 000× g for 7 min and supernatants and cell pellets were collected . Ferric chloride ( 20 mg/ml FeCl3•6H2O in ethanol ) was added to the supernatants at a final concentration of 0 . 6 mM and allowed to incubate at room temperature for one hour . The supernatants were extracted twice with 5 ml CHCl3 and the organic fraction was retained . The cell pellets were resuspended in 2 . 5 ml of ethanol and incubated with shaking for 12 hours at 37°C . After centrifuging for 7 min at 4 , 000× g , the ethanol supernatant was retained and 2 . 5 ml of water and FeCl3 ( to 2 . 2 mM ) was added . This mixture was allowed to incubate at room temperature for one hour , after which it was extracted twice with 5 ml CHCl3 and the organic fraction retained . The cell and supernatant extracts were then evaporated using a Vacuufuge ( Eppendorf ) and resuspended in 500 µl CHCl3 . After having normalized based on CPMs , extracts were then subjected to TLC on 10 cm×10 cm , 250-µm-thick silica gel 60 ( Sigma ) developed in ethanol/hexanes/water/ethyl acetate/acetic acid , 5∶25∶2 . 5∶35 . 5 [51] . Plates were allowed to dry and then exposed to a phosphoimager screen for 60 hours and analyzed with a Storm Phosphoimager ( Molecular Dynamics ) . 55Fe-loaded MBT and cMBT , as well as radiolabelled salicylic acid substrate , were also subjected to TLC alongside the extracts . Rf values for MBT ( 0 . 42 ) and cMBT ( 0 . 16 ) were the same as those previously reported and the salicylic acid , MBT , and cMBT loading controls ran the same as their extracted radiolabelled counterparts . Prior to virulence studies , all strains were demonstrated to have PDIMs and positive neutral red assessments . Mid-log phase cultures of wt Mtb ( ML617 ) , the mmpS4/S5 double deletion mutant ( ML618 ) , the mmpS5 singly complemented mutant ( ML619 ) , the mmps4 singly complemented mutant ( ML620 ) , and the doubly mmpS4/S5 complemented mutant ( ML624 ) were diluted to OD600 ∼0 . 1 to implant ∼1 , 000 bacilli in the lungs of mice using a Middlebrook inhalation exposure system ( Glas-Col ) . Twenty four 4- to 5-week old female BALB/c mice ( Charles River ) were infected with ML617 , ML618 , ML619 , ML620 , or ML624 . Four mice from each group were weighed and sacrificed at days 1 , 14 , 28 , 56 , 84 , and 112 post-infection to determine the number of bacilli in the lung and spleen . Mouse organs were aseptically removed , homogenized , and serially diluted . Appropriate dilutions were plated onto Middlebrook 7H11 agar plates to determine the colony forming units . For histological analysis , representative tissue samples from each group at days 1 , 28 , 56 , 112 post-infection were fixed in 10% formaldehyde , embedded in paraffin , sectioned , and stained with hematoxylin and eosin using standard procedures . Thirteen 4- to 5-week old female BALB/c mice were infected with 7 , 500–10 , 000 bacilli using the five strains mentioned above using the same method already described . Time to death was followed and survival proportions of mice infected with high dose aerosol were calculated . The experiment was stopped after 180 days post-infection . Detailed protocols for other experiments are provided in the supplement ( Text S1 ) .
|
In the late 19th century the French physician Armand Trousseau recognized that treating anemic tuberculosis patients with iron salts exacerbated the disease . In 1911 Twort postulated that mycobacteria produce an essential growth factor which was identified in 1953 as mycobactin . The hydrophobic mycobactin and its more water-soluble cousin carboxymycobactin are small molecules made by Mycobacterium tuberculosis to scavenge iron from its human host . While the biosynthesis of these siderophores has been decoded , it was unknown how M . tuberculosis secretes these molecules . In this study , we identified two similar transport systems , MmpS4/MmpL4 and MmpS5/MmpL5 , which are required for biosynthesis and export of siderophores by M . tuberculosis . The lack of these transport systems drastically decreased the number of M . tuberculosis cells in the lungs and spleens of infected mice . Lung examination and histological assessment in mice infected with the mmpS4/S5 deletion strain showed almost no signs of infection . Further , none of the mice infected with this strain died within 180 days in contrast to wild-type M . tuberculosis . In this study , we identified the first components of a novel siderophore export system in M . tuberculosis and showed the importance of siderophore export for virulence of M . tuberculosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"lipids",
"microbial",
"physiology",
"proteins",
"protein",
"structure",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens"
] |
2013
|
Discovery of a Siderophore Export System Essential for Virulence of Mycobacterium tuberculosis
|
Dengue is becoming an increasing threat to non-endemic countries . In Japan , the reported number of imported cases has been rising , and the first domestic dengue outbreak in nearly 70 years was confirmed in 2014 , highlighting the need for greater situational awareness and better-informed risk assessment . Using national disease surveillance data and publically available traveler statistics , we compared monthly and yearly trends in the destination country-specific dengue notification rate per 100 , 000 Japanese travelers with those of domestic dengue cases in the respective country visited during 2006–2014 . Comparisons were made for countries accounting for the majority of importations; yearly comparisons were restricted to countries where respective national surveillance data were publicly available . There were 1007 imported Japanese dengue cases ( Bali , Indonesia ( n = 202 ) , the Philippines ( n = 230 ) , Thailand ( n = 160 ) , and India ( n = 152 ) ) . Consistent with historic local dengue seasonality , monthly notification rate among travelers peaked in August in Thailand , September in the Philippines , and in Bali during April with a smaller peak in August . While the number of travelers to Bali was greatest in August , the notification rate was highest in April . Annually , trends in the notification rate among travelers to the Philippines and Thailand also closely reflected local notification trends . Travelers to dengue-endemic countries appear to serve as reliable “sentinels” , with the trends in estimated risk of dengue infection among Japanese travelers closely reflecting local dengue trends , both seasonally and annually . Sentinel traveler surveillance can contribute to evidence-based pretravel advice , and help inform risk assessments and decision-making for importation and potentially for subsequent secondary transmission . As our approach takes advantage of traveler data that are readily available as a proxy denominator , sentinel traveler surveillance can be a practical surveillance tool that other countries could consider for implementation .
Dengue is a mosquito-borne febrile viral disease that is endemic in over 100 countries in tropical and subtropical regions with notifications increasing 30-fold over the last 50 years , affecting an estimated 100 million people annually [1–3] . Over these decades , it has become apparent that dengue shows fairly consistent seasonality with periodic epidemic years in many endemic areas [4–5] . Importantly , dengue is becoming an emerging threat to non-endemic countries . The number of people traveling to dengue-endemic areas has been rising , along with reports of dengue cases among travelers . In fact , dengue is currently the second most common specific etiologic diagnosis among returned travelers with febrile illness after malaria globally , and the most common febrile illness among travelers returning from Southeast Asia and South Central Asia [6–8] . Imported dengue cases can contribute to further spread of dengue in non-endemic areas when competent vectors such as Aedes albopictus and Ae . aegypti are present . In recent years , autochthonous dengue outbreaks following importation have been reported in several non-endemic countries such as France , Croatia , Portugal , and the United States [9–12] . Japan has been no exception . While dengue is not endemic in Japan , approximately 50 to 200 imported cases have been reported annually in the last 10 years with an upward trend , causing growing concern of domestic spread as Ae . albopictus is active in much of Japan from spring to fall [13] . Such concern became a reality in summer 2013 , when a German dengue case suspected to have been infected while traveling in Japan was reported [14] , and in 2014 Japan experienced the first confirmed domestic dengue outbreak in nearly 70 years with 162 autochthonous cases [15 , 16] . The autochthonous dengue strain was shown to be similar to those circulating in Southeast Asia , the region associated with the majority of Japanese imported cases [15] . This recent chain of events of autochthonous dengue transmission in Japan , arising from importation of dengue , highlighted the need for greater situational awareness , risk assessment , and evidence-based communication . Since the domestic outbreak of 2014 , interest in situational awareness of dengue in endemic areas frequented by travelers has increased , and such information , together with existing sentinel traveler case data , may feed into better informed risk assessment . Previously , Nakamura et al . demonstrated that the risk of dengue among Japanese travelers appeared to be greater during historic dengue high season relative to low season in the endemic countries that they had visited; these findings supported the idea that the local dengue situation may directly affect the risk of infection among travelers [13] . However , the association between yearly trends in imported cases and yearly domestic fluctuations in the associated destination countries was not assessed , although periodic epidemic years occur in many endemic areas [4 , 5] . Such assessment became increasingly important after the domestic outbreak—if yearly trends in endemic destination areas are correlated with those among travelers , then the former can directly inform risk assessment for secondary domestic transmission following importation , assuming that a higher number of imported cases would increase the risk of secondary transmission and accounting for activity level of competent mosquitoes . Therefore , building on our previous work and the experience from others [5–8] , our objectives were to describe and compare the trends in notifications between sentinel traveler dengue cases among Japanese travelers with those of domestic dengue cases in the country visited , by month and by year . The monthly assessment updates the seasonality findings reported by Nakamura et al . [13] , while the yearly assessment approach highlights additional benefits from sentinel traveler surveillance .
Dengue has been a notifiable disease in Japan since April 1999 . Physicians are required to report demographic , clinical and exposure history information of laboratory-confirmed cases through the Japanese national infectious disease surveillance system ( National Epidemiological Surveillance of Infectious Diseases ( NESID ) ) . To report a dengue case in Japan , at least one of the following laboratory confirmation methods is required: 1 ) virus isolation; 2 ) detection of virus-specific nucleic acid sequences by polymerase chain reaction ( PCR ) method; 3 ) detection of dengue nonstructural protein 1 ( NS1 ) antigen in serum; or 4 ) detection of anti-dengue IgM antibody in serum; or seroconversion or 4-fold rise in IgG or IgM antibody titers in paired serum samples in neutralization test or hemagglutination inhibition test [13 , 16] . Dengue cases reported during 2006–2014 were extracted from NESID . Cases were excluded from analyses if any of the following criteria were met: 1 ) missing information on country traveled; 2 ) traveled to multiple countries during the potential incubation period ( 2–14 days ) ; 3 ) home address outside of Japan ( i . e . , potentially not included in the proxy denominator of Japanese travelers ) ; 4 ) known to be a non-Japanese national ( i . e . , potentially not included in the proxy denominator of Japanese travelers ) ; or 5 ) traveled to a dengue-endemic country that had a total of 30 or fewer reported imported cases during the nine-year study period ( due to sample size constraints ) . Comparisons of temporal trends in notifications between sentinel traveler dengue cases among Japanese travelers and domestic dengue cases in the country visited were made for India , Indonesia , the Philippines , and Thailand because these countries accounted for the majority of imported dengue cases to Japan . Annual international traveler data for Japanese travelers that visited India , Indonesia , the Philippines , and Thailand were obtained from the Japan National Tourist Organization ( http://www . jnto . go . jp/jpn/reference/tourism_data/departure_trends/index . html ) [17] . Data for Japanese travelers visiting India and Indonesia were restricted to those from 2006–2013 , as 2014 data were not yet available as of July 2015 . Japanese travelers to the Philippines and Indonesia were based on Japanese residency rather than Japanese nationality , as the latter were unavailable . Monthly traveler data for Japanese travelers that visited Bali province , Indonesia , the Philippines , and Thailand were obtained from the Japanese Tourism Marketing Company ( http://www . tourism . jp/statistics/outbound/ ) [18] . For Indonesia , monthly traveler data were only available for Bali province . Philippines data were restricted to those from 2006–2013 , as monthly traveler data were not yet available for 2014 as of July 2015 . Monthly Japanese traveler data were not available for India . For dengue cases reported domestically in the select endemic countries , annually reported numbers of dengue cases for the Philippines and Thailand were available and obtained from the World Health Organization’s Regional Office for the Western Pacific’s dengue website ( http://www . wpro . who . int/topics/dengue/en/ ) [19] and the national surveillance reports for dengue published by the Bureau of Epidemiology , Department of Disease Control Ministry of Public Health , Thailand ( http://www . boe . moph . go . th/ ) [20] , respectively . Reported imported dengue cases in Japan were described by age , sex , and country visited . To compare the country-specific trends in imported dengue cases among Japanese travelers with the number of cases reported in the country visited 1 ) by month and 2 ) by year , the following methods were used: Only the extracted NESID data contained individual level data , which did not include identifiable information . None of the other data sources included individual unit data . Thus , ethical approval and informed consent were not required .
A total of 1512 dengue cases were reported during the study period from 2006–2014 . 162 were autochthonous cases associated with the domestic outbreak in 2014 , while 343 cases were excluded based on the exclusion criteria ( including 95 ( 6% ) cases excluded for visiting multiple countries during the potential incubation period ) . Thus , a total of 1007 ( 67% ) cases remained for analysis . Among those , 625 ( 62% ) were male and the median age was 31 years ( range: 0–90 ) . For both sexes , the age group most reported was 20–29 years ( 326/1007 [32%] ) ( Fig 1 ) . The countries associated with the largest number of imported dengue cases were Indonesia ( n = 317 [31%] ) , the Philippines ( n = 230 [23%] ) , Thailand ( n = 160 [16%] ) , and India ( n = 152 [15%] ) , comprising approximately 85% of the 1007 cases ( Fig 2 ) . The monthly number of reported imported cases from 2006–2014 was highest during August through October , although the distribution varied by country . For example , the number of cases associated with visiting Thailand peaked in August , while the Philippines peaked in September and India was highest in October ( Fig 3; Table 1 ) . In Indonesia , the highest numbers of cases were observed during February to April and in August ( Fig 3 ) ; similarly , among those that visited Bali province , Indonesia , the trend was bimodal , peaking during February to April and again in August ( Table 1 ) . When the monthly number of imported cases from 2006–2014 was calculated as the monthly notification of country-specific imported cases per 100 , 000 Japanese travelers , the overall monthly trend was similar to the number of imported cases ( Fig 4; Table 1 ) . Notification rates for Thailand and the Philippines peaked in the same months as the number of imported cases . For Bali , the notification rate was highest during February to May , peaking in April , and with a smaller peak in August . While the number of travelers to Bali was greatest in August , the notification rate was highest in April ( Table 1 ) . The annual number of reported imported cases from 2006–2014 was highest in 2013 ( n = 192 ) , followed by 2010 ( n = 188 ) . Although the distribution varied by country visited , there was an increase from 2009 to 2010 followed by a decrease in 2011 for all four countries ( Fig 2 ) . When calculated as the annual country-specific notification rate ( i . e . , estimated from number of reported imported cases per 100 , 000 Japanese travelers per year ) , differences by country visited and by year were observed . For instance , the annual notification rate for Thailand was consistently lower than those of the other three countries . Notably , rates among travelers also saw an increase from 2009 to 2010 followed by a decrease in 2011 for all four countries ( Fig 5 ) . The trend in the annual number of dengue cases per 100 , 000 travelers visiting the Philippines closely mirrored the trend in the annual number of dengue cases reported in the Philippines ( Fig 6A ) . Similarly , the trend in the annual number of dengue cases per 100 , 000 travelers visiting Thailand reflected the annual number of dengue cases reported in Thailand ( Fig 6B ) .
In the present study , we found that travelers to dengue-endemic countries appear to serve as “sentinels” . Based on retrospective analysis of national Japanese surveillance data from 2006–2014 , the estimated risk of dengue infection among Japanese travelers closely reflected local dengue trends , seasonally and annually . The overall trend in Japanese dengue traveler cases was the same as previously reported [13] . Male cases remained dominant , with case distribution skewed towards young adults , although the age distribution of travelers is unknown . Moreover , despite yearly variation , the majority continued to be associated with travel to India , Indonesia , Thailand , and the Philippines , comprising 70–90% of imported cases . Similarly , as previously reported [13] , country-specific seasonality in notification rate among Japanese travelers was consistent with historic dengue seasonality in the endemic countries visited . For instance , the notification rate among travelers to Thailand and the Philippines was highest during July to September , coinciding with their historic dengue high seasons [4 , 13 , 21] . For Bali province , an area popular for Japanese tourists , we found that the notification rate among travelers peaked in April , which corresponds to local peak activity during 2001–2010 , ranging from February to June [22] . We also found that the yearly country-specific dengue notification rate among travelers was correlated with the yearly number of reported dengue cases among the respective dengue-endemic countries . In endemic countries , dengue epidemics are known to occur every few years [4 , 5] , and notifications among travelers mirrored such trends . The year 2010 was particularly notable , as an increase was observed in notification rates among travelers , domestically in multiple endemic countries , and in the entire Western Pacific Region [4 , 22] . Similar correlations between trends in imported cases and local trends in dengue-endemic countries have been reported previously . For example , patterns of local incidence of dengue in the Pacific Islands closely reflected patterns of incidence of dengue imported from these areas to New Zealand [23] . The international GeoSentinel Surveillance Network suggested that seasonality of dengue in travelers is similar to that in the local population , and thus such data may benefit those advising prospective travelers or those diagnosing ill-returned travelers [5] . Additionally , an excess of cases among travelers to Southeast Asia in the years 1998 and 2002 was shown to reflect epidemics in Southeast Asia . Thus , the concept of sentinel traveler surveillance is not new; various approaches and networks on several diseases have been established , particularly in Europe and North America ( e . g . , TropNet Europe , GeoSentinel , Canadian Travel Medicine Network ) [6–8 , 24–26] . However , existing systems have been mostly limited to participating private clinics and trend assessment of sentinel cases has often been based on number of cases or proportionate morbidity , using ill travelers as the denominator . Interpretation of trends based on numerators or proportionate morbidity alone requires caution , as they may not reflect incidence . More rigorous methods have gone further to estimate person-time incidence among cohorts of travelers [27]; however , such special studies can be costly and time-consuming . The strength of our method is that we use publicly available “big data” as a proxy to estimate country-specific dengue incidence among travelers ( strictly speaking a ratio , as crudely comparing number of cases to the number of travelers to a given location and period ) . Inferring trends from the frequency of imported cases alone requires caution , as risk cannot be determined solely from numerator data . As with Bali , while there was a high number of reported cases in August , the number of Japanese travelers tended to be highest in August ( summer vacation season ) , and the estimated risk among travelers to Bali was higher from February to May ( Table 1 ) . It is important to consider separately the risk of infection for travelers and the potential for secondary transmission following importation . While the risk for travelers to Bali may be highest from February to May , risk of further transmission in Japan is minimal from such importations , as the vector is not active during these months in most of Japan . On the contrary , importation of viremic cases during the summer months , even those that are asymptomatic [28] , would indicate a higher risk of secondary transmission and a greater public health concern , as the vector is active and more people are outside with more exposed skin . Additionally , all else being equal , given our findings on annual trends , the risk of importation—and potentially subsequent secondary transmission—may be higher when endemic countries are experiencing an epidemic year . Our approach enables evidence-based risk communication to travelers and help inform risk assessment and decision-making for secondary transmission . Furthermore , sentinel traveler surveillance may provide potentially important information to endemic countries , especially low-capacity/resource-limited countries . In such settings , surveillance for dengue may be limited , and information from sentinel traveler surveillance from import countries may provide earlier awareness [5 , 23] . Additionally , if virologic data such as genotype are lacking or unavailable in a timely fashion in the endemic country , such information may potentially be of value for the endemic country . There are important limitations in this study . Specific human activity/behavior has been reported as major risk factors for dengue infection [5 , 29–32] , but specific travel locations or activity/behavior at destination sites were not available from surveillance data . These aspects may confound the association between country visited and risk of dengue infection . The difference in notification rates among the destination countries could be explained by differing behaviors of Japanese travelers visiting those countries that directly relate to dengue infection risk , and such inter-country comparisons require caution . In addition , as dengue activity has been shown to vary at the local level [22 , 30 , 33–35] , the risk of infection in areas travelers visit may differ from those that are frequented by the local population [36] . Nevertheless , in our study , the monthly and yearly notification rate among sentinel travelers correlated closely with trends observed from national surveillance of the endemic countries . Secondly , the risk of dengue infection for Japanese travelers might be underestimated , as the information is based on reported surveillance data from patients who sought healthcare [36] . Also , as dengue has a short incubation period , many Japanese who developed illness abroad may be missed . However , as such bias likely remains consistent over time , it is unlikely to influence the interpretation of temporal trends observed in sentinel traveler surveillance . Similarly , while our assessments were based on a few select countries ( limited to those comprising the majority of Japanese sentinel traveler cases ) , given sufficient sample sizes , we do not believe that the correlations observed would only apply to these specific countries but not to others . Lastly , 316 ( 31% ) of the cases were confirmed by detection of anti-dengue IgM antibody in serum from a single specimen , and this may pose a concern as other flavivirus ( e . g . yellow fever , Japanese encephalitis , Zika viruses ) infections could result in false detections . However , as a sensitivity analysis restricted to cases confirmed by methods other than IgM resulted in very similar yearly and monthly trends , our interpretations remain qualitatively unchanged ( see Supporting Information ) . Given that our approach takes advantage of readily available data that are often available at little to no cost , we encourage others to explore sentinel traveler surveillance based on their surveillance and traveler data and corroborate our findings . To conclude , we found that trends in dengue notifications among Japanese travelers closely reflected local dengue trends , both seasonally and annually . Sentinel traveler surveillance can be a practical , low-cost , evidence-based tool to help inform risk assessment , situational awareness , and decision-making , in the effort to both reduce infection among travelers and the potential for secondary domestic transmission . Enhanced awareness among travelers and healthcare providers , as well as global information-sharing , will continue to be essential for combating the dengue threat . Given the ever-increasing expansion of dengue and other vector-borne diseases , maximizing the utilization of sentinel traveler surveillance is more important than ever .
|
With increasing globalization , the threat of dengue is rising in areas that were previously unaffected . Japan has been experiencing a rise in notifications of imported cases , and in 2014 confirmed the first domestic outbreak in nearly 70 years . Such events prompted the country to more actively utilize existing imported dengue case data among travelers to inform situational awareness , risk assessment , and evidence-based decision-making . Using both national disease surveillance data and publically available traveler statistics , we compared monthly and yearly trends between reported numbers of dengue cases among Japanese travelers and those of domestic dengue cases in the countries visited . By using the number of Japanese travelers to a dengue-endemic country as an approximate denominator , we estimated the risk of dengue infection among travelers to the country . This method is more appropriate than simply monitoring the number of reported imported cases because it accounts for fluctuating numbers of travelers , such as during vacation periods . This study demonstrated that the trends in dengue notifications among travelers were consistent with local dengue trends , both yearly and seasonally . Our simple approach , which takes advantage of existing data , may be readily adopted elsewhere to help inform risk of importation and potential subsequent domestic transmission .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"decision",
"making",
"japan",
"geographical",
"locations",
"india",
"neuroscience",
"indonesia",
"cognition",
"infectious",
"disease",
"control",
"infectious",
"diseases",
"epidemiology",
"people",
"and",
"places",
"infectious",
"disease",
"surveillance",
"philippines",
"asia",
"oceania",
"disease",
"surveillance",
"biology",
"and",
"life",
"sciences",
"cognitive",
"science",
"thailand"
] |
2016
|
Dengue Sentinel Traveler Surveillance: Monthly and Yearly Notification Trends among Japanese Travelers, 2006–2014
|
Disease progression in response to infection can be strongly influenced by both pathogen burden and infection-induced immunopathology . While current therapeutics focus on augmenting protective immune responses , identifying therapeutics that reduce infection-induced immunopathology are clearly warranted . Despite the apparent protective role for murine CD8+ T cells following infection with the intracellular parasite Leishmania , CD8+ T cells have been paradoxically linked to immunopathological responses in human cutaneous leishmaniasis . Transcriptome analysis of lesions from Leishmania braziliensis patients revealed that genes associated with the cytolytic pathway are highly expressed and CD8+ T cells from lesions exhibited a cytolytic phenotype . To determine if CD8+ T cells play a causal role in disease , we turned to a murine model . These studies revealed that disease progression and metastasis in L . braziliensis infected mice was independent of parasite burden and was instead directly associated with the presence of CD8+ T cells . In mice with severe pathology , we visualized CD8+ T cell degranulation and lysis of L . braziliensis infected cells . Finally , in contrast to wild-type CD8+ T cells , perforin-deficient cells failed to induce disease . Thus , we show for the first time that cytolytic CD8+ T cells mediate immunopathology and drive the development of metastatic lesions in cutaneous leishmaniasis .
CD8+ T cells contribute to the control of pathogens by cytokine production , cytolytic activity or both . In the case of intracellular parasites , the production of IFN-γ by CD8+ T cells is protective , while in viral infections CD8+ T cells provide protection by inducing cytokine production and killing virally infected cells [1] . Nevertheless , these same CD8+ T cell effector functions can also promote increased pathology , and the presence of CD8+ T cells has been associated with increased pathology in several infectious and autoimmune diseases [2] , [3] , [4] , [5] , [6] , [7] , [8] . In some cases the pathology is believed to be associated with IFN-γ or IL-17 production , while in other situations cytolytic activity is linked with disease . Still , the mechanistic basis by which CD8+ T cells could potentially contribute to increased pathology is difficult to determine in humans . Cutaneous leishmaniasis is one of many diseases where the outcome of the infection depends on both the extent of parasite elimination and the relative induction of potentially immunopathologic responses . A great deal is known about how leishmania parasites are eliminated . Thus , control of these intracellular parasites requires a CD4+ Th1 cell response , which leads to IFN-γ production that enhances the killing capacity of infected macrophages and dendritic cells [9] , [10] . CD8+ T cells respond during infection and contribute to the control of Leishmania by producing IFN-γ , which not only activates macrophages to kill the parasites , but also promotes the differentiation of naïve T cells into Th1 cells [11] , [12] . On the other hand , few studies have addressed how immunopathology develops in cutaneous leishmaniasis . Correlations with enhanced immunopathology and lower levels of IL-10 or IL-10 receptor expression have been observed in patients , but the unregulated responses that promote pathology are not defined [13] , [14] . In patients infected with L . braziliensis , the number of CD8+ T cells increases in lesions as the disease worsens , and patients with mucosal disease–where metastatic lesions develop in the nasopharyngeal mucosa due to a destructive inflammatory response–have elevated numbers of cytotoxic cells in the blood [15] , [16] , [17] . Interestingly , it has also been reported that L . major-infected Rag1−/− mice reconstituted with CD8+ T cells develop much larger lesions than unreconstituted Rag1−/− mice [11] . Together , these observations implicate CD8+ T cells as inducers of pathology . As CD8+ T cells can produce IFN-γ in leishmaniasis , it is possible that an overproduction of IFN-γ promotes increased pathology . On the other hand , the severity of disease in patients infected with L . braziliensis is directly associated with increased numbers of granzyme expressing CD8+ T cells [15] . Thus , it remains to be determined whether CD8+ T cells are indeed pathogenic , and if so , whether they increase disease severity by cytokine production and/or enhanced cytolytic activity . Defining the mechanisms that promote the immunopathology observed in cutaneous leishmaniasis is a critical first step in developing an approach to control the disease . Here , we define the pathologic role that CD8+ T cells play in L . braziliensis infections . We found that the most highly expressed genes in leishmanial lesions are associated with the lytic pathway and that CD8+ T cells within the lesions of L . braziliensis patients are functionally cytolytic . Using a murine model we found that CD8+ T cells contribute to increased lesion size following infection with L . braziliensis parasites . Strikingly , we found that the development of metastatic lesions was also promoted by the presence of CD8+ T cells . Mechanistically , we demonstrated that the pathology associated with unregulated CD8 function is not due to enhanced IFN-γ or IL-17 production , but rather is due to excessive perforin-dependent cytolytic activity by CD8+ T cells . Thus , our findings show for the first time that cytolytic CD8+ T cells are not only present during infection , but that they promote increased immunopathology and metastatic lesions in cutaneous leishmaniasis .
To better understand the local immune environment during human leishmaniasis and the extent to which a cytolytic program is associated with the infection , we carried out whole genome expression profiling of lesions from patients infected with L . braziliensis . Over 500 genes were expressed ≥3-fold ( p≤0 . 05 ) in lesions compared to normal skin from uninfected donors ( not shown ) . Strikingly , many known components of cytolytic granules were found amongst the most strongly expressed genes in lesions [18] ( Fig . 1A ) . For example , granzyme B , granzyme A and granulysin , a pore-forming molecule found in the granules of human CTLs and NK cells , were the first , seventh and tenth genes most strongly expressed genes overall in lesions , respectively . Moreover , gene set enrichment analysis of the entire expression data set revealed a significant enrichment for KEGG pathways involved in NK cell mediated cytotoxicity , graft-versus-host diseases and allograft rejection ( not shown ) , all of which involve cytolysis [19] . Next we measured the protein levels of granzyme B and perforin in CD8+ T cells recovered from the peripheral blood or from lesions of L . braziliensis infected patients ( Fig . 1B ) . More CD8+ T cells obtained from lesions expressed granzyme B ( Fig . 1C ) in comparison to cells from the blood , and both populations contained cells expressing perforin ( Fig . 1D ) . These data suggest that both skin and peripheral CD8+ T cells have the capacity to degranulate . To determine if in fact these cells were degranulating , we assessed their surface expression of CD107a ( Fig . 1B , lower panels ) . CD107a is a lysosomal membrane glycoprotein ( also known as Lamp1 ) present in CD8+ T cell granules . During degranulation , this molecule is transiently exposed on the cell surface and thus is a marker for the release of cytotoxic granules by CD8+ T cells [20] . We found a significant increase in CD8+ T cells expressing surface CD107a from lesions , while CD8+ T cells from the blood failed to express CD107a ( Fig . 1E ) . These results not only confirm previous studies showing that CD8+ T cells express granzyme in leishmaniasis [15] , but also extend these findings to show that the CD8+ T cells are cytolytically active within human leishmanial lesions . To determine if our results from L . braziliensis patients could be mechanistically dissected using animal models we first asked if depletion of CD8+ T cells would affect lesion development in BALB/c mice . In contrast to L . major , L . braziliensis infections in BALB/c mice results in the development of an ulcerated lesion that eventually resolves [21] . Thus , BALB/c mice infected with 105 L . braziliensis developed a substantial lesion that healed spontaneously when treated with control isotype antibody ( Fig . 2A , closed circles ) . In contrast , mice depleted of CD8+ T cells with an anti-CD8-specific antibody developed substantially smaller lesions ( Fig . 2A , open circles ) , suggesting that CD8+ T cells contribute to lesion size . The relative decrease in lesion size in CD8 depleted mice was not due to an alteration in parasite number ( Fig . 2B ) , indicating that the change in lesion size was due primarily to differences in the inflammatory response . Thus , these data indicate that CD8+ T cells contribute to the inflammatory response following L . braziliensis infection in BALB/c mice . Two factors seem to be associated with immunopathology in L . braziliensis patients: an increase in CD8+ T cells recruited to the lesions and a decrease in immunoregulatory cytokines . To test if CD8+ T cells could directly enhance disease , L . braziliensis infected Rag1−/− mice were reconstituted with naive CD8+ T cells ( RAG+CD8 ) or CD8+ and CD4+ T cells ( RAG+CD4+CD8 ) , at a 1∶1 ratio , and the course of infection was followed . Unreconstituted Rag1−/− mice infected with L . braziliensis developed minimal lesions . Thus , similar to L . major or L . amazonensis [11] , [22] , lesion development with L . braziliensis likely depends upon generation of a T cell-dependent inflammatory response . As expected , RAG+CD4+CD8 mice developed small nodules that resolved within 7 weeks following infection ( Fig . 2C ) . In contrast , transfer of CD8+ T cells alone to Rag1−/− mice led to the development of an uncontrolled lesion ( Fig . 2C ) . To determine if the increased pathology observed in RAG+CD8 mice was due to uncontrolled parasite growth , parasite loads were assessed within the lesions . We found that Rag1−/− mice and RAG+CD8 mice had similar numbers of parasites at the infection site , in spite of the disparity in lesion size observed in these animals , while transfer of CD4+ and CD8+ T cells into Rag1−/− mice led to significantly better control of the parasite in the infected ear ( Fig . 2D ) . Thus , the exacerbated lesion development in RAG+CD8 mice compared with Rag1−/− mice is due to CD8+ T cell mediated pathology rather than differences in the number of parasites in the ear . Most notably , the enhanced lesion size in RAG+CD8 mice was accompanied by a rampant immunopathologic response . By 5 weeks post infection we observed destruction of the infected ear in RAG+CD8 mice , but minimal pathology in either Rag1−/− or RAG+CD4+CD8 mice ( Fig . 2E ) . This destruction was accompanied by infiltration of many more CD8+ T cells than in Rag1−/− mice that received both CD4+ and CD8+ T cells ( data not shown ) . Histologically , we could observe at lower magnification the substantial differences in lesion thickness in mice without T cells and RAG+CD4+CD8 or RAG+CD8 mice . Higher magnification showed that lesions from Rag1−/− mice were composed of infected macrophages and granulocytes ( Fig . 2E ) . At this time point , lesions from RAG+CD4+CD8 mice were healing and exhibited a mild dermal lymphocytic infiltrate accompanied by epidermal hyperplasia and spongiosis with few leishmania organisms ( Fig . 2E ) . In contrast , the lesions from RAG+CD8 mice showed a dramatic cellular infiltration consisting of lymphocytes , granulocytes and many highly infected cells ( Fig . 2E ) . Moreover , the epidermis in these lesions exhibited substantial hyperplasia and areas of ulceration , and the severe inflammatory response in the dermis led to alterations in the cartilage . Using a pathology score that goes from 0 to 5 , where 0 is mild and 5 is severe disease , at 7 weeks post-infection RAG+CD8 mice had the most severe disease ( 5 ) followed by Rag1−/− mice with moderate disease ( 2 ) and RAG+CD4+CD8 mice with no disease ( 0 ) . We also characterized the myeloid cell composition present in lesions from RAG+CD8 mice , and found that a majority of the myeloid cells within lesions were neutrophils . ( Fig . S1 ) . Together , these observations illustrate the critical role of the inflammatory response as the main factor driving lesion development , further highlighting the importance of identifying the mechanisms that control pathology in cutaneous leishmaniasis . An additional unexpected result observed in RAG+CD8 mice was the development of metastatic lesions . This was particularly notable in the contralateral ear , which developed gross pathology indistinguishable from the primary lesion ( Fig . 2F ) . In addition to the contralateral ear , we observed lesions at other skin sites , including the nose , tail , and footpad . We were able to culture parasites from these regions , confirming the spread of parasites to these additional skin sites ( data not shown ) . In contrast , we did not observe metastatic lesions in Rag1−/− or RAG+CD4+CD8 mice ( Fig . 2F ) . Thus , the development of both primary and metastatic lesions in Rag1−/− mice was dependent upon CD8+ T cells . To characterize the functions of the CD8+ T cells transferred in the presence and absence of CD4+ T cells , cells from lesions of L . braziliensis infected reconstituted Rag1−/− mice were assessed for IFN-γ , IL-17 and granzyme B by flow cytometry ( Fig . 3A ) . A higher percentage of CD8+ T cells made IFN-γ when these cells were transferred together with CD4+ T cells , although CD8+ T cells made IFN-γ in the absence of CD4+ T cells ( Fig . 3B ) . In contrast , CD8+ T cells from RAG+CD8 mice produced significantly more granzyme B in the absence of CD4+ T cells ( Fig . 3C ) . Finally , although only a small percentage of CD8+ T cells from RAG+CD8 mice produced IL-17 , IL-17 production was completely abrogated by the presence of CD4+ T cells in RAG+CD4+CD8 mice ( Fig . 3D ) . Overall , these results suggest that CD8+ T cells could be mediating increased pathology due to cytolytic activity ( indicated by high levels of granzyme B ) , IL-17 , or IFN-γ , and indicate that CD4+ T cells may regulate these responses . As we observed expression of genes associated with cytolysis in leishmanial lesions , we first assessed if the immunopathology observed in RAG+CD8 mice was related to cytolytic activity . Cells were obtained 5 weeks after infection of RAG+CD4+CD8 or RAG+CD8 mice , and were stained for CD107a expression directly ex vivo . In RAG+CD4+CD8 mice analysis of CD107a expression showed a small percentage of degranulating CD8+ T cells at the infection site ( Fig . 4A , 4B ) . On the other hand , a high percentage of CD8+ T cells from lesions of RAG+CD8 mice expressed surface CD107a ( Fig . 4A , 4B ) . To confirm that CD107a expression was indicative of degranulation , we sought to visualize CD107a at the interface between CD8+ T cells and infected cells . For these experiments , eGFP+ CD8+ T cells were transferred into Rag1−/− mice that were subsequently infected with L . braziliensis parasites expressing mCherry . Cells from lesions taken 5 weeks after infection were incubated for 1 hour in the presence of anti-CD107a monoclonal antibody and then run on an ImageStream flow cytometer . We observed the presence of CD107a ( blue ) at the synapse between CD8+ T cells ( green ) and L . braziliensis infected target cells ( red ) providing further support that CD8+ T cells from RAG+CD8 mice were degranulating ( Fig . 4C ) . Analysis of the total doublets that contained eGFP+ CD8+ T cells showed that surface expression of CD107a on CD8+ T cells was more frequent when CD8+ T cells were in contact with infected in comparison to uninfected target cells ( Fig . 4D ) . Finally , to directly show that CD8+ T cells from RAG+CD8 mice induce infected cell death , we visualized the interactions between CD8+ T cells and infected cells using a spinning disk confocal microscope . As above , mice were infected with mCherry L . braziliensis parasites and reconstituted with GFP+ CD8+ T cells , and after 5 weeks cells from the lesions were isolated and immediately visualized . We observed several different types of interactions between CD8+ T cells and infected cells . In some cases , T cells moved around the surface of infected cells and ultimately detached from the target cell ( Video S1 ) . We also detected stable conjugates between infected cells and CD8+ T cells that , after an average of 25 minutes ( ranging from 10 to 60 minutes ) , led to the formation of membrane blebs on the target cell , but with no visible damage to intracellular parasites . This was followed by immediate detachment ( Figure 4E; Video S2 ) in most cases , while in some cases the CD8+ T cells remained in contact for several minutes after the target cell underwent apoptosis ( Video S3 ) . Analysis of 63 movies obtained from 6 individual experimental infections showed that infected cells were killed more frequently than uninfected cells ( Fig . 4F ) , even though the number of uninfected cells was much greater in these preparations ( data not shown ) . Due to the fact that cells can leave the imaging field , we could not always determine the outcome of CD8+ T cell-target interactions ( undetermined ) . Thus , we expect that our analysis may have underestimated the number of times CD8+ T cell interactions with infected cells led to target cell killing . Because CD8+ T cells preferentially lysed L . braziliensis infected cells , we hypothesized that this was a specific CD8+ T cell interaction . To test this , we transferred CD8+ T cells that would not recognize L . braziliensis ( transgenic OT1 cells ) or wild-type ( WT ) CD8+ T cells to Rag1−/− mice and infected them with L . braziliensis parasites . Confirming data shown in Fig . 2C , large lesions were detected in mice that received WT CD8+ T cells ( Fig . 5A ) . In contrast , mice that received OT1 cells exhibited small lesions and no evidence of severe pathology ( Fig . 5A and 5B ) . We found no differences in the number of parasites detected in the lesions of mice reconstituted with WT CD8+ T cells and OT1 cells , providing further evidence that the pathology observed in the RAG+CD8 mice was unrelated to the number of parasites observed in the lesions ( Fig . 5C ) . Finally , we measured OT1 cell degranulation in the lesions of L . braziliensis infected Rag1−/− mice by assessing CD107a expression . No CD107a expression by the OT1 cells was observed , while a significant percentage of WT CD8+ T cells were expressing CD107a ( Fig . 5D , 5E ) . Although the induction of apoptosis by CTLs is often associated with minimal inflammation , cell death can also promote extensive inflammatory responses [23] . To directly assess if CD8+ T cell-mediated pathology is related to their cytolytic capacity , we next determined if CD8+ T cells required perforin to mediate immunopathology in RAG+CD8 mice . Rag1−/− mice received either Prf1−/− or WT CD8+ T cells and were infected with L . braziliensis . While WT CD8+ T cells promoted increased pathology , perforin deficient CD8+ T cells failed to promote lesion development ( Fig . 6A ) or severe pathology ( Fig . 6B ) . Moreover , in contrast to WT CD8+ T cells , mice that received Prf1−/− T cells failed to develop metastatic lesions ( data not shown ) . This difference in lesion development was independent of the number of parasites in the primary lesions as the parasite load was the same in mice that received Prf1−/− and WT CD8+ T cells ( Fig . 6C ) . The absence of severe disease in mice that received Prf1−/− CD8+ T cells was not due to differences in the capacity of CD8+ T cells from Prf1−/− mice to be recruited to the infected skin , since both groups had similar percentages of CD8+ T cells present in the skin ( Fig . 6D , 6E ) . We also tested if IFN-γ or IL-17 were required for immunopathology induced by CD8+ T cells . To do so , we transferred CD8+ T cells deficient in either IFN-γ or IL-17 and both groups of mice developed lesions similar to WT CD8+ T cells , with similar numbers of parasites , suggesting that these cytokines are not required for lesion development ( Fig . 6F–I ) . Taking all of our results together , we conclude that perforin-dependent cytolytic activity is the main mechanism by which CD8+ T cells promote disease . Moreover , we have discovered a previously unappreciated role for CD8+ T cells in promoting the development of metastatic lesions in cutaneous leishmaniasis .
Leishmania are intracellular parasites that infect and multiply within myeloid-lineage cells , such as macrophages and DCs . They are transmitted to humans by sand flies and more than 10 different species can cause disease in humans . Cutaneous leishmaniasis is the most common clinical form of leishmaniasis and it is estimated that 1 . 5 million new cases of cutaneous leishmaniasis occur annually and 12 million individuals are exposed to infection worldwide [24] . The severity of cutaneous disease depends on both the extent of parasite replication and the relative induction of immunopathologic responses . For example , disease induced by Leishmania braziliensis , the leading causal agent of leishmaniasis in South America , has several clinical manifestations , all of which are associated with significant immunopathologic responses . The factors that mediate these immunopathologic responses are poorly defined . Here , by combining clinical data from patients with results from our experimental L . braziliensis infection , we have identified CD8+ T cell-mediated cytotoxicity as a major contributor to increased pathology in cutaneous leishmaniasis . While CD4+ Th1 cells are critical for controlling Leishmania , CD8+ T cells also play a protective role , since mice lacking CD8+ T cells exhibit increased susceptibility to L . major [11] , [12] , [25] . Moreover , in several experimental vaccine models CD8+ T cells contribute to immunity [26] , [27] , [28] , [29] , [30] , [31] . In contrast , we found that depletion of CD8+ T cells in L . braziliensis infected mice decreases inflammation . It is currently unclear why CD8+ T cell depletion has different effects in L . major and L . braziliensis infected mice . However , CD8+ T cell derived IFN-γ blocks the development of a Th2 response in L . major infected mice [12] , and the apparent lack of Th2 induction in L . braziliensis infection may contribute to these differential effects [21] , [32] . Nevertheless , our findings indicate that CD8+ T cells can promote increased disease . Although this may seem paradoxical , our current results suggest that a bifurcation in CD8+ T cell effector function provides a potential explanation . Specifically , CD8+ T cells provide protection by releasing IFN-γ , thereby activating macrophages to kill the parasites , and promoting a stronger CD4+ Th1 response in response to L . major infections [12] . In contrast , our current results with L . braziliensis suggest that cytolytic function , rather than IFN-γ production , promotes increased pathology . These findings are consistent with our own observations , as well as others , that in L . braziliensis patients there is a direct correlation between granzyme expressing CD8+ T cells in lesions and disease severity [15] , [33] . While we found that perforin-expressing CD8+ T cells were required for this pathology , we cannot exclude a role for NK cells . Indeed , NK cells have been linked with pathology in mucocutaneous leishmaniasis patients [17] and NK cells were present in the lesions of L . braziliensis infected RAG+CD8 mice ( data not shown ) . Similarly , we cannot exclude a role for CD8 derived chemokines in promoting infiltration of inflammatory cells into these lesions [34] . Taken together , the human studies suggest that CD8+ T cell cytolytic activity is pathologic , and using a mouse model we are the first to show conclusively that indeed a dysregulated CD8+ T cell response promotes perforin-dependent immunopathology in cutaneous leishmaniasis . Therefore , defining mechanisms that control CD8+ T cell function , including regulation by factors present within the cytokine milieu [35] , will be a crucial step in identifying therapeutic targets to treat immunopathologic sequelae . Surprisingly , the role of CTLs in cutaneous leishmaniasis has not been extensively explored . Nevertheless , studies have shown that CD8+ T cell lines or clones can be lytic for Leishmania infected cells [17] , [34] , [36] , [37] , [38] , [39] , [40] . Here , we extend those studies and show the ability of effector CD8+ T cells from infected mice to kill naturally infected targets . We were able to visualize for the first time CD8+ T cells lysing L . braziliensis infected targets and characterize the interactions that preceded the lytic event . In this regard , we found that CD8+ T cells bind target cells for an average of 25 minutes before target cell lysis , similar to what is seen for CD8+ T cell induced apoptosis of peptide pulsed B cells [41] . This killing appeared to be specific , since OT1 T cells–which cannot recognize leishmanial antigens–failed to degranulate or induce pathology in L . braziliensis infected mice . Additionally , while almost 40% of the interactions between CD8+ T cells and infected cells resulted in target cell lysis , a much smaller fraction ( ∼4% ) of interactions between CD8+ T cells and uninfected cells resulted in apoptosis . Although the mechanisms promoting killing of uninfected targets was not explored , these targets may have internalized leishmanial antigens or phagocytosed infected apoptotic cells , resulting in cross presentation [42] . Alternatively , the killing could be antigen-independent , as occurs when stressed cells express ligands recognized by NKG2D on CD8+ T cells [43] . Finally , it remains to be determined whether parasites are killed when the target cell is killed . We were unable to determine the fate of the parasites following killing of their host cell , primarily because dying cells did not remain in the field of focus . However , since parasite numbers were similar in Rag1−/− and RAG+CD8 mice , it seems unlikely that the parasites are killed , although we cannot disregard the possibility that in an immunocompetent mouse the parasites might be killed . Consistent with this view , a previous study with a CTL clone indicated that parasites survive after their host cell is killed by CD8+ T cells [40] . CD8+ T cells are best known for their ability to protect against viral infections by lysing virus-infected targets . However , a pathologic role for CD8+ T cells has also been observed . For example , virally induced myocarditis is prevented in mice deficient in perforin [5] . Similarly , experimental cerebral malaria is mediated by CD8 CTLs [6] , [7] . Perhaps most analogous to the pathologic CTL responses that we observe in leishmaniasis are data from experimental Trypanosoma cruzi infections in which CD8+ T cell responses that have traditionally been considered protective have more recently been linked to pathology [8] , [44] . Such a finding is again consistent with a bifurcation in CD8+ T cell function in which IFN-γ producing CD8+ T cells are protective , while perforin expressing CD8+ T cells mediate increased pathology in the heart [8] . Leishmania infection is a good model to understand this duality of CD8+ T cell function in vivo , since CD8+ T cells are protective in visceral leishmaniasis [26] , [45] , [46] , [47] , whereas we now understand that this is the opposite for L . braziliensis . Understanding what mechanisms drive CD8+ T cells to become pathogenic or protective is an important goal to design new treatments and is now under investigation in our lab . At present , it is unclear why CTLs induce pathology in cutaneous leishmaniasis , since apoptosis is primarily thought to drive anti-inflammatory responses [48] . However , apoptotic cells can undergo secondary necrosis if not rapidly cleared by phagocytes . In secondary necrosis the integrity of the plasma membrane is lost and intracellular constituents of the cell are released , increasing the inflammatory response [23] , and providing positive feedback to increase the cytolytic activity of CD8+ T cells [49] , [50] . In several clinical forms of leishmaniasis , parasites metastasize to distant cutaneous sites . In mucosal leishmaniasis , metastatic lesions develop in the nasopharyngeal region , which leads to substantial morbidity [16] , while in disseminated leishmaniasis individual nodules develop at multiple sites in the skin [51] , [52] . It has been suggested that an RNA virus present in some South American strains of Leishmania enhances the immune response , potentially promoting metastasis [53] . However , this is not a universal finding , as parasite strains associated with metastatic disease do not all contain this virus ( the strain that we have used here does not contain the virus ) . Nevertheless , the association of enhanced immune responses with metastasis is consistent with our results , and suggests that dysregulated immune responses play an essential role in this process . Although the mechanistic basis for immune-mediated metastasis is unclear , the immune system may simply be responding to disseminated parasites , thereby inducing the inflammatory response required for lesion development at distal sites . Alternatively , the immune response itself may contribute to parasite spread . One attractive possibility is that CTL killing of infected cells enhances the release of parasites , allowing them to metastasize more efficiently to distant skin sites . The relative decrease in immunopathology in mice reconstituted with both CD4+ and CD8+ T cells suggests that CD4+ T cells control the pathogenicity of CD8+ T cells . This may be , in part , by controlling the number of CD8+ T cells or the number of parasites within the lesion . However , we also found that CD8+ T cells transferred in the absence of CD4+ T cells expressed relatively higher levels of GrzB and IL-17 , and less IFN-y , than did CD8+ T cells co-transferred with CD4+ T cells . We predict that this difference is due to the absence of regulatory T cells ( Tregs ) since in tumor models Tregs control the cytotoxicity of CD8+ T cells in a TGF-β dependent manner [41] , [54] , and in leishmaniasis both CD4+ Th1 cells and CD4+ T regulatory cells ( Treg ) dampen the immune response [55] , [56] , [57] . While we have not yet determined how CD4+ T cells control the pathologic CD8+ T cell response , we believe that the ratio of CD4+ and CD8+ T cells may be a critical determinant in disease outcome , since we found that Rag1−/− mice developed disease even in the presence of CD4+ T cells , as long as they are in relatively low numbers ( data not shown ) . This result is consistent with the observed change in ratio of CD4+ and CD8+ T cells present in lesions as disease progresses in human patients [15] . Thus , we suggest that an analysis of the CD4∶CD8 ratio in biopsies taken for diagnostic purposes might be useful in predicting disease outcome . In summary , our results define a new role for CD8+ T cells in leishmaniasis . While CD8+ T cells have previously been thought of as primarily protective , our results demonstrate that they can mediate severe pathologic responses . This finding makes it essential that more effort is directed at delineating the factors that determine CD8+ T cell effector function during leishmaniasis , since such information is critical in considering therapies or vaccines that may impact CD8+ T cells . Moreover , our data highlight the importance of evaluating CD8+ T cell effector function in many infections where CD8+ T cells may be playing dual protective and pathologic roles .
This study was conducted according to the principles specified in the Declaration of Helsinki and under local ethical guidelines ( Ethical Committee of the Maternidade Climerio de Oliveira , Salvador , Bahia , Brazil; and the University of Pennsylvania Institutional Review Board ) . This study was approved by the Ethical Committee of the Federal University of Bahia ( Salvador , Bahia , Brazil ) ( 010/10 ) and the University of Pennsylvania IRB ( Philadelphia , Pa ) ( 813390 ) . All patients provided written informed consent for the collection of samples and subsequent analysis . This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee , University of Pennsylvania Animal Welfare Assurance Number A3079-01 . All cutaneous leishmaniasis patients were seen at the health post in Corte de Pedra , Bahia , Brazil , which is a well-known area of L . braziliensis transmission . The criteria for diagnosis were a clinical picture characteristic of cutaneous leishmaniasis in conjunction with parasite isolation or a positive delayed-type hypersensitivity response to Leishmania antigen , plus histological features of cutaneous leishmaniasis . In all cases , the immunological analysis was performed before therapy . Biopsies were performed using a 4 mm punch , treated with Liberase ( Roche ) for 90 mins at 37°C/5% CO2 . Biopsies were dissociated and passed through a 50 µm Medicon filter ( BD phamingen ) . Peripheral blood mononuclear cells were obtained from heparinized venous blood layered over a Ficoll-Hypaque gradient ( GE Healthcare ) , then washed and resuspended in RPMI1640 and stained for flow cytometry as described below . For whole genome expression microarray , lesion biopsies preserved in RNAlater ( Qiagen ) were homogenized using a rotor-stator and RNA was isolated using the RNeasy Plus kit ( Qiagen ) . Biotin-labeled complementary RNA ( cRNA ) was generated using the Illumina TotalPrep RNA amplification kit ( Ambion ) . RNA and cRNA quality were assessed on a BioAnalyzer ( Agilent ) . Illumina HumanHT-12 version 4 expression beadchips were hybridized with cRNA from ten L . braziliensis lesion biopsies and two biopses collected from uninfected donors . Data was quantile normalized and differential expression analysis was carried out using GenomeStudio v1 . 8 software ( Illumina ) . Genes were considered differentially regulated if expression increased or decreased ≥3-fold with a diffscore of ≥13 or ≤−13 ( equivalent to p≤0 . 05 ) . Data was deposited on the Gene Expression Omnibus ( GEO ) database for public access ( GSE# GSE43880 ) . Heat map tools available on GenePattern [58] were used to graphically display differentially regulated genes in Figure 1 . BALB/c and C57BL/6 mice ( 6 weeks old ) were purchased from NCI , and Rag1−/− ( B6 . 12957-RAG1tm1Mom ) ( N14F12 ) , Ifn-γ−/− ( B6 . 129S7-Ifngtm1Ts ) ( N8+2F23 ) , Prf1−/− ( perforin ) ( C57BL/6-Prf1tm1Sdz ) ( F ? +52 ) were purchased from The Jackson Laboratory . C57BL/6 IL17a−/− mice ( N9 ) were provided by Dr . Yoichiro Iwakura ( University of Tokyo , Japan ) and OT1 ( B6 . 129S7-Rag1tm1MomTg ( TcraTcrb ) ) mice were purchased from Taconic Farms and mice expressing eGFP in all T cells were originally obtained from Ulrich van Andrian ( Harvard University ) . All mice were maintained in a specific pathogen-free environment at the University of Pennsylvania Animal Care Facilities . L . braziliensis parasites ( strain MHOM/BR/01/BA788 ) ( deMoura et al . , 2005 ) were grown in Schneider's insect medium ( GIBCO ) supplemented with 20% heat-inactivated FBS , 2 mM glutamine , 100 U/ml penicillin , and 100 µg/ml streptomycin . Metacyclic enriched promastigotes were used for infection [59] . Transgenic parasites expressing both luciferase and mCherry were generated by transfecting the parental L . braziliensis strain with SwaI linearized pLucCherry , a modified version of pIR1SAT encoding firefly luciferase and mCherry in the SmaI and BglII sites , respectively . Transgenic parasites were selected by plating transfected parasites on M199 agar supplemented with nourseothricin ( 50 µg/ml; Sigma-Aldrich ) . Mice were infected with 105 L . braziliensis in the right ear , and the course of lesion progression was monitored weekly by measuring the diameter of ear induration with digital calipers ( Fisher Scientific ) . Mouse: anti-CD45 . 2 APC-AlexaFluor 750 , anti-CD11b eF450 , anti-CD11c FITC , anti-F4/80 PE-Cy7 , anti-CD3 eFluor 450 , anti-IFN-γ PeCy7 , anti-CD107a eFluor 660 ( all from eBioscience ) . Anti-CD4 APC-Cy7 , anti-IL-17A PE and Ly6C PerCP-Cy5 . 5 ( BD Pharmingen ) , anti-CD8β PerCPCy5 . 5 ( Biolegend ) and anti-granzyme B APC ( Invitrogen ) . Anti-Ly6G APC ( Biolegend ) . Human: anti-CD3 APCCy7 , anti-CD8a PeCy5 . 5 and anti-perforin FITC ( all from eBioscience ) . Anti-CD107a PE ( BD Pharmingen ) and anti-granzyme B APC ( Invitrogen ) . For in vivo CD4 or CD8 depletion , mice received i . p . injections of 250 µg of GK1 . 5 or 53-6 . 72 ( BioXcell ) , respectively . Splenocytes from C57BL/6 , Prf1−/− , Ifn-γ−/− , Il17a−/− and OT1 mice were collected , red blood cells lysed with ACK lysing buffer ( LONZA ) and CD8+ T or CD4+ T cells were purified using a magnetic bead separation kit ( Miltenyi Biotec ) . Three million CD8+ or CD4+ T cells were transferred alone or together to Rag1−/− mice that were subsequently infected with L . braziliensis . Mice reconstituted with CD8+ T cells alone received 4 injections of 250 µg of anti-CD4 within the first 2 weeks . Infected and uninfected ears were harvested , the dorsal and ventral layers of the ear separated , and the ears incubated in RPMI ( Gibco ) with 250 µg/mL of Liberase ( Roche ) for 90 mins at 37°C/5% CO2 . Following incubation , the enzyme reaction was stopped using 1 mL of RPMI media containing 10% FBS . Ears were dissociated using a cell strainer ( 40 µm , BD Pharmingen ) and an aliquot of the cell suspension was used for parasite titration . The parasite burden in the ears was quantified as described previously [12] . Briefly , the homogenate was serially diluted ( 1∶10 ) in 96-well plates and incubated at 26°C . The number of viable parasites was calculated from the highest dilution at which parasites were observed after 7 days . To determine if parasites disseminate , footpad , opposite ear , and nose were cultured in complete Schneider's medium at 26°C and parasite growth was evaluated after 7 days . Cell suspensions from mice were incubated with PMA ( 50 ng/mL ) , ionomycin ( 500 ng/mL ) and Brefeldin A ( 10 µg/mL ) ( all from SIGMA ) for cytokine and granzyme B intracellular staining . For degranulation assays , cells were resuspended in 4×106/mL and incubated for 6 hours at 37°C/5% CO2 with anti-CD107a and monensin ( eBioscience ) . Before surface and intracellular staining , cells were washed and stained for live/dead fixable aqua dead cell stain kit ( Molecular Probes ) , according to manufacturer instructions . For human granzyme B , perforin and CD107a expression , cells were incubated for 6 hours with anti-CD107a antibody and Brefeldin A without stimulation followed by surface and intracellular staining . Rag1−/− mice were reconstituted with 3×106 eGFP CD8+ T cells and infected with 105 metacyclic enriched mCherry L . braziliensis . Six weeks post infection ear tissue was dissociated and cells were incubated with anti-CD107a fluorescent antibody for 1 hour . Total ear cell suspension was acquired on the ImageStream machine and analyzed using the IDEAS software ( Amnis Corporation ) . Rag1−/− mice were reconstituted with 3×106 eGFP CD8+ T cells and infected with 105 metacyclic enriched mCherry L . braziliensis . Five to eight weeks post infection ears were harvested and 2×106 cells used for imaging . Cells were maintained at 37°C at 5% CO2 on the stage of a fully enclosed microscope . Fields were selected randomly where both mCherry+ cells and eGFP+ cells could be detected . Images from multiple fields were acquired every 60 seconds on a Leica DMI 4000 inverted microscope equipped with a Yokagawa spinning disk confocal head and a Hamamatsu EMCCD 510 camera . eGFP+ and mCherry images were taken sequentially with 488 and 561 nm laser excitation , respectively . Image acquisition was controlled by MetaMorph Software . At seven weeks post infection ears were harvested and fixed in 10% formalin . Ears were embedded in paraffin and 5 µm sections were cut and stained with hematoxylin and eosin . Histological sections were blindly scored and the system was pre-defined as the following: the lowest score ( 0 ) was defined by the absence of a lesion; moderate disease was defined as a small lesion with a localized cell infiltration ranging from minimal ( 1 ) , moderate ( 2 ) to intense ( 3 ) ; lesions showing ulceration and intense localized cell infiltration were scored as 4; and the highest score ( 5 ) was defined by severe ulceration , cell infiltration involving the whole ear as well as cartilage destruction . Statistical analysis was performed with the Mann–Whitney test ( two-sided t-test ) using Prism ( GraphPad Software ) .
|
Leishmaniasis is a parasitic disease where the host immune response plays an essential role in pathogenesis . However , the mechanisms promoting immunopathology in patients are still unclear . We performed gene expression profiling of skin lesions from cutaneous leishmaniasis patients and normal skin and the results demonstrated that the most expressed genes in leishmanial lesions were associated with the cytolytic pathway . Using both human samples and mouse models we showed that CD8+ T cells are cytolytic within leishmanial lesions and kill Leishmania infected target cells . We found that the CD8+ T cell cytolytic response was not protective , but rather promoted increased immunopathology , associated with enhanced recruitment of neutrophils to the site of infection . CD8+ T cells also promoted the development of metastatic lesions at distant skin sites . Together , our results clearly demonstrate that activation of CD8+ T cell cytolytic responses is detrimental to the host and that targeting this pathway could be a new approach to treat patients with leishmaniasis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"skin",
"infections",
"leishmaniasis",
"neglected",
"tropical",
"diseases"
] |
2013
|
Cytotoxic T Cells Mediate Pathology and Metastasis in Cutaneous Leishmaniasis
|
Mycolactone is the exotoxin produced by Mycobacterium ulcerans and is the virulence factor behind the neglected tropical disease Buruli ulcer . The toxin has a broad spectrum of biological effects within the host organism , stemming from its interaction with at least two molecular targets and the inhibition of protein uptake into the endoplasmic reticulum . Although it has been shown that the toxin can passively permeate into host cells , it is clearly lipophilic . Association with lipid carriers would have substantial implications for the toxin’s distribution within a host organism , delivery to cellular targets , diagnostic susceptibility , and mechanisms of pathogenicity . Yet the toxin’s interactions with , and distribution in , lipids are unknown . Herein we have used coarse-grained molecular dynamics simulations , guided by all-atom simulations , to study the interaction of mycolactone with pure and mixed lipid membranes . Using established techniques , we calculated the toxin’s preferential localization , membrane translocation , and impact on membrane physical and dynamical properties . The computed water-octanol partition coefficient indicates that mycolactone prefers to be in an organic phase rather than in an aqueous environment . Our results show that in a solvated membrane environment the exotoxin mainly localizes in the water-membrane interface , with a preference for the glycerol moiety of lipids , consistent with the reported studies that found it in lipid extracts of the cell . The calculated association constant to the model membrane is similar to the reported association constant for Wiskott-Aldrich syndrome protein . Mycolactone is shown to modify the physical properties of membranes , lowering the transition temperature , compressibility modulus , and critical line tension at which pores can be stabilized . It also shows a tendency to behave as a linactant , a molecule that localizes at the boundary between different fluid lipid domains in membranes and promotes inter-mixing of domains . This property has implications for the toxin’s cellular access , T-cell immunosuppression , and therapeutic potential .
Buruli ulcer ( BU ) is a cutaneous disease caused by Mycobacterium ulcerans . The presence of necrosis , which is accompanied by surprisingly little inflammation or pain , is considered the most characteristic clinical presentation of BU disease [1–3] . The macrolide exotoxin mycolactone , which is secreted by M . ulcerans , is thought to be the key virulence factor , playing the central role in the pathogenesis of BU [4 , 5] . The exotoxin’s causative effect was brought to light when it was shown that the symptoms of BU could be replicated by injection of mycolactone alone [6] . Mycolactones exist in multiple isomeric forms [7] and have been shown to act both in vivo and in vitro on various mammalian cell types , including fibroblasts [7–11] , adipocytes [12] , keratinocytes [13] , myocytes [6 , 14] , macrophages [15 , 16] , and T cells [17 , 18] . At the cellular level , mycolactones induce apoptosis , cytoskeletal rearrangements , impaired cytokine production , and interference with cellular signaling [19] . Biophysical experiments have shown that the mycolactone hijacks the Wiskott-Aldrich syndrome proteins ( WASP and N-WASP ) , binding to them with 100-fold higher affinity than their natural activator CDC42 [20] . This in turn activates actin branching , leading to defective cell adhesion , uncontrolled directional migration , and eventually to cell death [21] . Recent work has also shown that mycolactone inhibits the Sec61-dependent translocation of proteins into the endoplasmic reticulum , a process that likely explains the absence of immune responses and inflammation in BU [22 , 23] . Additional recent work has shown that mycolactone may exert an analgesic effect by inhibiting signaling by the angiotensin II type 2 receptor ( AT2R ) ; it inhibits AT2R and leads to decreased hyperpolarization in mouse neuronal cells [24] . Collectively , these effects suggest that mycolactone disrupts normal host function via multiple pathways , resulting a suite of consequences and range of cellular cytotoxicity . Multinuclear NMR experiments , combined with chemical synthesis have provided the chemical structure of mycolactone , which is a macrolide [4 , 25–27] . It consists of an 8-undecenolide ( C1-C11 fragment ) substituted at C11 by a nine-carbon atom chain ( C12-C20 , the ‘northern’ fragment ) and at C5 by a pentanoic acid ester ( C1'-C16' , the ‘southern’ fragment ) as shown in Fig 1A . Currently , there are hundreds of synthesized derivatives that retain some of the bioactivity of the native compound and that have helped to shed insight on certain aspects of the structure-activity relationship of the toxin . For instance , it has been shown that the southern fragment of mycolactone is a strong determinant of the molecule’s cytotoxicity [28] . In spite of these advances , much remains unknown about mycolactone’s cytotoxic mechanism of action . Mycolactone has been shown to passively permeate through host cell membranes [29] , and to be delivered from bacterial to host cells via outer membrane vesicles [30] , though the mechanism of vesicle-host cell exchange is unknown . Furthermore , mycolactone’s solubility in aqueous media is poor , yet it is known to bind soluble protein targets [21] . Most importantly , mycolactone evades the host immune response , and it has not yet been possible to elicit antibodies against it with traditional immunization approaches [31] , suggesting its availability in circulation is poor . All of these facets beg the questions: how is mycolactone distributed in the host and what role does its likely association with lipids play in its distribution and pathogenicity ? Mycolactone inhibits both the co- and post-translational pathways for protein translocation across the endoplasmic reticulum ( ER ) [22 , 23] . This action has been associated with the inhibition of activation-induced production of cytokines leading to T-cell immunosuppression . It exerts immune-suppressive effects by impairing the capacity of T cells to produce cytokines [32] and by negatively affecting lymphocyte homing through the downregulation of L-selectin [17] . A recent study found that mycolactone led to a specific blockade of translocation of nascent proteins across the ER membrane , though it found no disruption of organelle membrane structure [22] . Further analysis showed that mycolactone inhibits the Sec61 translocon in many different ways [23] . These studies lead to the question of whether such an effect could be exerted through interaction of mycolactone with the ER membrane and/or ER membrane-bound channels involved in protein translocation . To address this question , we need to know how mycolactone interacts with the ER membrane and how it could affect the conformations of transmembrane channels . It has also been reported that mycolactone interferes with the immune response pathways associated with T-cell activation . The T-cell signaling cascade is triggered by the phosphorylation of ITAM motifs in the T-cell receptor by Lck protein kinase [33] . Then the phosphorylated ITAM recruits ZAP-70 , which leads to phosphorylation of downstream signaling molecules such as LAT . The dysregulation of T-cell activation pathway by mycolactone was specific to Lck kinase [8] . Specifically , mycolactone exposure led to hyperphosphorylation of Lck . Surprisingly , this effect was mediated without the direct interaction of mycolactone with Lck . It was proposed that mycolactone-induced hyperpolarization is mediated by changing the partition of Lck such that the concentration of Lck is enhanced in a lipid-raft like domain after 30 mins of exposure [8] . Given the lipid-like structure of mycolactone , is it possible that mycolactone affects the microdomains ( eg , lipid raft like domains ) in cellular membranes and thereby alters the localization of Lck ? For disease management , rapid point-of-care diagnostics are critical , and mycolactone may provide an avenue to diagnose BU . Interestingly , the development of antibodies against mycolactone by traditional immunization approaches has been challenging [31] , likely because exotoxins have cytotoxic effects on immune cells . More recently , a B cell hybridoma technology has been used to develop immune-sera and monoclonal antibodies against mycolactone [34] . When such antibodies are pre-mixed with synthetic mycolactone in solution , they are indeed effective in protecting cells from toxicity . However , their protective ability within a host environment has yet to be demonstrated . This brings up the question of whether the toxin is buried in host lipid molecules/assemblies , hiding its antigenic domain during infection . The consensus among the above experimental studies has been a lack of details about the association of mycolactone to the cellular membrane , and this deficit has hampered efforts to deduce the mechanisms related to its cytotoxicity , pathogenicity and immunosuppression . Both diagnostic and therapeutic development efforts are also facing technical roadblocks because of the lack of knowledge about how the toxin is distributed and trafficked in the host environment . In this context , and given the toxin’s amphiphilic character , it is critical to address how it interacts with lipids and lipophilic carriers . With the recent advancement in high performance computing and efficient algorithms , numerical simulation approaches such as molecular dynamics ( MD ) simulations are able to fill the gaps in our understanding of molecular mechanisms in spatial and temporal regions which are often inaccessible to experimental techniques . In this study , we used coarse-grained ( CG ) molecular dynamics simulations to study the effects of mycolactone in biomembranes . Our results suggest that mycolactone prefers the membrane over the aqueous environment , localizing predominantly at the glycerol-lipid tail interface just under the membrane surface . We also observe that mycolactone alters several dynamical , physical , and mechanical properties of lipid membranes . Finally , mycolactone is able to decrease the line tension between the ordered and disordered lipid domains , and therefore to potentially interfere with the nano-scale ordering of biological membranes [35] . These findings have substantial implications for the toxin’s distribution in the host environment and mechanisms of pathogenicity .
First we considered the preferential partitioning of the exotoxin mycolactone between the organic ( octanol ) and water phases . It is worth pointing out that proper assignment of the partition coefficient is critical for the accurate representation of the molecule at the CG level , where the right balance is required in order the keep consistency with the different sets of parameters [36] . We computed the water-octanol partition coefficient ( logPow ) of native mycolactone using all-atom molecular dynamics simulations ( see Methods and Panel A in S2 Fig ) . We found that mycolactone has a stronger preference for the organic phase over the aqueous phase . We obtained a logPow of 9 . 0 ± 0 . 2 and 11 . 6 ± 0 . 2 , using Thermodynamic Integration ( TI ) and Bennett’s acceptance ratio ( BAR ) approaches , respectively . Similar evaluation of logPow was carried out with the CG topology using the MARTINI force field [36] . The result from the CG representation provided a logPow of 8 . 8 ± 0 . 1 and 9 . 0 ± 0 . 2 using TI and BAR approaches , respectively . The values obtained at coarse-grained representation are in excellent agreement with their atomistic counterparts ( Panel B in S2 Fig , and S3 Fig ) . The consistency in logPOW between all-atom and CG MD simulations allowed us to use the CG parameters for rapid exploration of the effects of mycolactone in membrane lipid models , at time scales and system sizes not easily attainable with fully atomistic simulations . Next , we study the preferential localization of a single mycolactone molecule in a fully hydrated diC16-PC bilayer by considering two independent all-atom MD simulations ( 1 μs each ) . A single mycolactone was initially placed in the center of the two membrane leaflets ( corresponding to 72 lipids per monolayer ) and its localization was tracked . In one case , mycolactone moved to the lipid-water interface where the polar groups could interact with the glycerol and lipid head groups . This conformation was stable and remained for the duration of the simulation . In the other case , the toxin aligned with the aliphatic tails of the lipids near the middle of lipid bilayer . This configuration was seen only once . Although the atomistic simulations are able to independently capture two different spatial configurations during the MD runs , obtaining an overall conformational probability of mycolactone with such short AA simulations is challenging . Therefore , we resort to CG simulations to quantify overall probability distributions of potential mycolactone configurations in model lipid bilayers . Accordingly , we focus on the study of the preferential localization of the toxin in a fully solvated diC16-PC lipid bilayer using the CG representation . We set up this simulation using the MARTINI force field , which applies a mapping of four heavy atoms to one CG bead ( interaction site ) . The CG representations of mycolactone and diC16-PC are shown in Fig 1A . The initial setup of the system is also shown , with the simulation box represented by a gray square in Fig 1B containing 66 mycolactones ( 5% total lipid composition ) . Fig 1C depicts an equilibrated configuration from a 2 μs CG simulation of the mycolactone-DPPC system . We observe that mycolactone preferably adopts two configurations when embedded in the bilayer . All mycolactone molecules were placed initially in an aligned conformation with the lipid tails of the membrane . Approximately 10% of the mycolactone molecules remained aligned with the aliphatic tails of the lipid membranes ( Fig 1D , red histogram ) , which is consistent with a configuration from the atomistic simulations ( small inset ) . This configuration allows the extended tails of the northern and southern fragments of the mycolactone to closely interact with the glycerol moieties connecting the head group of the diC16-PC lipids . Most of the mycolactones , however , were found in the water-membrane interface , with both tails interacting with the glycerol moieties of the lipid ( Fig 1D , black histogram ) . After the initial 0 . 5 μs , we didn’t observe any further conversion to this state at the water-membrane interface . Considering the better sampling and faster diffusion of the MARTINI based CG representation , the CG simulations can be used to estimate the relative free energy between the two configurations: P ( inter ) P ( alig ) =e− ( ΔGkT ) where P ( alig ) denotes the population in the aligned state and P ( inter ) denotes the population of mycolactones close to the interface . The difference in free energy ( -5 . 7 kJ mol-1 ) suggests that configuration at interface is energetically more favorable . Even though CG simulations are able to populate both configurations , the exchanges between these configurations are still limited by lack of sampling . The potential of mean force calculations as described below provide a more quantitative measure of preference between these two configurations . To gain more insight into the thermodynamic behavior of mycolactone in the lipid membrane , we calculated the potential of mean force ( PMF ) for translocation of the toxin through the membrane . The reaction coordinate for translocation followed the axis parallel to membrane normal , tracking the distance between the center of mass of mycolactone and the center of mass of the membrane . Results for one leaflet are summarized in Fig 2 . As shown , mycolactone is preferentially localized within the boundaries of the glycerol moieties , captured by the minimum at 1 . 5 nm from the bilayer center . The inset shows the mycolactone conformation that populates this minimum in PMF near the membrane-water interface . This conformation is similar to the one obtained using unbiased simulations . A second minimum is found near the center of the bilayer . This configuration is featured by the membrane aligned structure ( inset ) and stabilized by an energy barrier of ~ 10 kJ mol-1 . The full detachment of the toxin requires ~ 50 kJ mol-1 from the most stable configuration . This value is comparable to the energy required for the extraction of a single cholesterol molecule from a lipid bilayer , suggesting that membranes indeed provide a favorable energetic environment for the toxin . In addition , the PMF shows that the configuration at the interface is favored by ~10 kJ mol-1 , which is ~4 . 3 kJ mol-1 higher than the direct observations from unbiased simulations . An effective association constant to the membrane can be calculated by integrating the PMF reported above along the reaction coordinate to the limit of association between the membrane and mycolactone . The association of mycolactone with a lipid membrane is best described as an adsorption or partitioning process . However , one can consider this effective association affinity as a measure of the toxin’s strength of association with membranes , compared to that for its cytosolic targets ( e . g . , WASP ) . Consistent with previous membrane association analysis , we use a mathematical formalism [37] for an non-specific association constant: Kns=[P]mem[P]aq= ( c⊖ ) 1/3∫memexp[−βW ( z ) ]dz∫δ ( z−z* ) exp[−βW ( z ) ]dz Where z corresponds to the reaction coordinate distance between the center of mass of mycolactone and the bilayer during the PMF . Here , β = 1/kBT and [P] is the concentration of the biomolecule either in aqueous ( aq ) solution or in the membrane ( mem ) . z* corresponds to a chosen point on the binding pathway when the biomolecule is far away from the membrane-water interface ( e . g . bulk water ) . The standard state concentration c⊖ of 1/1660 Å-3 , corresponds to the standard concentration of 1 M in bulk water , and is raised to the 1/3 since only one degree of freedom ( z ) is integrated over in the PMF . Notice that the integral over the membrane region in the numerator has units of distance , and must be set relative to the z dimension of freedom in the bulk standard concentration , leaving the association constant unitless , as it should be [38 , 39] . Although they cannot be directly related , the computed nonspecific association constant ( 3 . 9 x 107 ) is on par with the experimentally measured specific association constant of mycolactone for N-WASP ( 5 . 8 x 106 ) , and thus clearly reflects a strong association of mycolactone with our model membrane . The most relevant quantity physiologically is the proportion of toxin buried in membranes relative to that bound to cytosolic targets . One can estimate this proportion based on the approximate size of a cell and concentration of WASP . For a model neutrophil , for example , with a radius of 4 . 15 μm and a 9 μM concentration of WASP , there would be ~238 molecules of mycolactone bound to the membrane for every one bound to WASP ( see S1 Text for details ) . Although this is an approximate estimate for a model membrane ( pure DPPC lipid bilayer ) , it strongly suggests coexistence of the toxin in association with lipids and cytosolic targets . The physiological relevance of this coexistence cannot be overlooked . Next , we performed long timescale comparative CG MD simulations of a mycolactone-diC16-PC system and a pure diC16-PC lipid bilayer system . In addition to our PMF calculations , which suggested a preference of mycolactone towards lipid membranes , the toxin’s influence on the structure and/or dynamics of lipid bilayers was also of interest . This can be characterized through comparative simulations with and without the mycolactone . First , 2 μs simulations are used to investigate the effect of mycolactone on the transition melting ( Tm ) temperature of a pure diC16-PC bilayer . As shown in Fig 3A , the presence of mycolactone reduces the gel-liquid Tm of diC16-PC by ~5 K , stabilizing the liquid phase at lower temperatures . When the mycolactone-diC16-PC system is cooled from 323K , the formation of a gel phase is observed at 287K . When mycolactone was not present , the pure diC16-PC system transitioned towards the gel phase at a slightly higher temperature . The liquid phase is observed at 292K , which is in agreement with the reported value in the original model [40] . In the simulations , the formation of gel phase is captured by the drop in the area per lipid which converges to an averaged area per lipid of ~ 0 . 47 nm2 in both systems . Above these transition temperatures , both systems display similar area per lipid ( 0 . 64 nm2 ) . We also measured the lateral diffusion constants at 323 K and these values are similar for both systems ( 1 . 7 x 10−7 cm2 s-1 and 2 x 10−7 cm2 s-1 for the pure and mixed systems , respectively ) . Another important feature under consideration was the area compressibility modulus KA which can inform on membrane deformations due to stretch . We calculated KA ( see Methods ) considering a large patch of ~ 6000 diC16-PC lipids with either 5% or 10% mycolactone content . Our results suggest that mycolactone strongly influences the membrane resistance to compression . For pure diC16-PC bilayers , the area compressibility modulus is KA = 282 ± 60 mN/m . This value compares reasonably well with the experimental value , which was reported to be KA = 231 ± 20 mN/m [41] . The compressibility modulus is reduced , however , when 5% mycolactone is incorporated in the bilayer ( KA = 208 ± 30 mN/m ) . Addition of more mycolactone ( up to 10% ) decreases the KA even more , to a value close to 99 ± 50 mN/m . Thus , compared to a pure lipid bilayer , lower force can rupture the membrane when mycolactone is present . Then , we examined whether mycolactone can affect the role of biological membranes as barriers by increasing their tendency to be porous . Similar comparative simulations were used to compute the effect of mycolactone on the critical line tension ( see Methods ) of the diC16-PC bilayer . Under low tension , mechanically generated pores are prone to close ( fill up ) . Under high tension , however , pores tend to grow larger , eventually causing rupture of the membrane . Typically , highly ordered lipid bilayers have a high edge energy ( line tension ) compared with more elastic membranes . Fig 3B shows the effect of applying ~ 60 mN/m surface tension to a pure diC16-PC bilayer . After 2 μs , the membrane became thinner , but no pore formation was observed . However , the membrane containing 5% mycolactone , was rapidly stretched , followed by the spontaneous formation of a pore ( Fig 3C ) . Similar behavior was also observed in 10 independent simulations . However , when the applied surface tension was reduced to 55 mN/m , pore formation was not observed . This suggests a decrease of at least ~8% in the critical line tension . Furthermore , in pure diC16-PC bilayer , pre-formed pores are stabilized at ~ 25 mN/m , however we found that this value decreases to ~20 mN/m when the membrane interacts with mycolactone . A proper regime of different surface tensions needs to be sampled to obtain a more quantitative value of line tension and to profile the timeline of formation and evolution of pores . Even though mycolactone could directly alter the surface tension , that effect is very small at the given concentration . However , in an already thinned membrane , 5% mycolactone affects the surface tension enough to form pores . The lateral heterogeneity of biological membranes likely plays an important role in cellular biophysics . For example , the activity of proteins localized within different domains can be influenced by the local membrane properties , and thus by phase properties . In model membranes , ternary mixtures of saturated lipids , unsaturated lipids , and cholesterol are known to segregate into two coexisting fluid lipid domains , liquid-ordered ( Lo ) and liquid-disordered ( Ld ) [42] . A line tension exists at the interface between such domains . Linactants are molecules that can modify the equilibrium in the boundaries by modulating the line tension between these domains and shift the preferential segregation of the molecules embedded in such regions . By comparing simulations of a ternary lipid mixture with and without mycolactone , we evaluated whether mycolactone can act as a linactant . To achieve this , we set up a control system composed of randomly placed diC16-PC , diC18:2-PC and cholesterol ( 4:3:3 lipid ratio ) as described in the Methods section . At the CG level , this system has previously been observed to segregate into liquid ordered and liquid disordered domains on a simulation time scale of 2 μs [43] . In addition , we also set up a mixed ternary system with 5% ( of total lipid ) made up of mycolactone . Both simulations were run for a total of 10 μs . Considering that the MARTINI model speeds up diffusion by a factor of ~4 [36] , processes longer than 10 μs should be captured in these simulations as well . Both systems were started with a random lipid distribution . In the absence of mycolactone , the ternary system undergoes lipid segregation ( Fig 4A ) , where cholesterol rich-domains are surrounded by saturated lipid tails . Cholesterol-poor domains are localized in the Ld region containing unsaturated lipid tails . It took about 5 μs for these domains to equilibrate , and no further transition was observed . We quantified the line tension between the two domains ( see Methods ) , considering the last 5 μs of the trajectory . In line with previous calculations [44] , the membrane patch considered here converged to a line tension of 16 ± 1 . 3 pN . Under similar simulation conditions , the membrane patch containing 5% mycolactone ( Fig 4B ) converged to a value of 7 . 2 ± 0 . 5 pN . This decrease of line tension by ~50% is comparable with the effect observed with different linactants [44] . To show the preferential localization of mycolactone in the membrane , we computed the electron-density profile along the interface of both domains in the membrane . As depicted in Fig 4C , the slightly higher peak corresponding to mycolactone is localized within the boundaries of the Lo and Ld , suggesting that indeed mycolactone co-localizes within the interface of both domains and may act as a linactant . We further quantified the lipid interaction preference , computing the lateral radius distribution ( RDF ) of mycolactone with respect to the center of mass of the saturated and unsaturated lipids , respectively . As shown in Fig 4D , mycolactone preferentially interacts with the unsaturated lipids from the Ld domain . The electron density profile shows a preference of ~70% for the interface region compared to bulk Ld and Lo regions . The preferential lateral interaction free energy ( ΔGPLI ) of mycolactone with the saturated versus unsaturated lipids is ~2 kJ mol-1 . As shown in Fig 4D , energetically , the direct interaction with mycolactone is slightly preferred for unsaturated lipids over saturated lipids . Values within the same range have been reported previously for the preferential partitioning of other lipid species [45] . Fig 4E and 4F show that similar preference is maintained for the head and tail regions of the molecule . Given the low ΔGPLI ( in the range of kT ) , our results suggest that mycolactone is preferentially localized within the interface . Such a mechanism may allow it to decrease the line tension and drop the energy barrier that keeps both domains stable .
The pathogenesis of Buruli ulcers , caused by Mycobacterium ulcerans , is clearly tied to mycolactone , a lipid-like exotoxin capable of permeating the host cell membrane , binding to cytosolic and membrane-bound targets , and ultimately inducing cell death . Although much has been learned about mycolactone’s mechanism of toxicity , many questions remain . In this study , we focused on the interactions of mycolactone with models of biomembranes , hoping to shed light on the toxin’s distribution and mechanisms of host cell penetration . The computed water-octanol partition coefficient indicates that mycolactone prefers to be in an organic solvent rather than an aqueous environment . Furthermore , our free energy calculation reveals that the exotoxin is preferentially buried in a pure lipid bilayer , in agreement with the partitioning data . In addition , both the unbiased and enhanced sampling simulations suggest that once in the bilayer , mycolactone preferentially localizes around the lipid glycerol groups close to the lipid-water interface , but that it can also span the membrane , reaching out to interact with the polar groups on either leaflet . This suggests a plausible mechanism for its translocation from the extracellular to the cytoplasmic side of the plasma membrane in which is localizes on one leaflet’s lipid-water interface , then spans , then flips to localize on the other leaflet’s lipid-water interface . There have been no experimental measurements of mycolactone in lipid versus solvent mediums to directly compare to our findings . Although it was reported that a fluorescent derivative of mycolactone accumulates in the cytosol of murine cells [29] , the reported images suggest localization within the ER membrane . Our simulation results are also indirectly supported by the composition of the native toxin extracted from infected patients . Those studies report that the toxin is heavily bound to the lipid extracts of the cell , suggesting that mycolactone is , in the absence of higher binding affinity host proteins , localized within the non-polar phase of the cell ( e . g . , membranes ) [18 , 46] . Mycolactone clearly needs to translocate across cellular membranes to reach its cytosolic and ER targets [18 , 47] . Given that our study shows that the exotoxin has a preference for membrane relative to aqueous environments , it is suggested that membranes and other lipid carriers facilitate its translocation throughout the host environment . Some of these carriers include cytosolic proteins , such as WASP and N-WASP . The mechanism for exchange from membrane to host cytosolic carriers could be further facilitated by co-localization , since targets like WASP and N-WASP are recruited to the surface of cell membranes by membrane-bound or membrane-associated proteins prior to activation by effector molecules ( e . g . , CDC42 ) . This is supported by the computed association constant between mycolactone and a model lipid membrane being on the same order or slightly stronger than the measured association constant with N-WASP [21] . Importantly , the relative affinity of mycolactone for the membrane will depend on the membrane composition . It is expected that the exotoxin prefers intracellular membranes , such as the endoplasmic reticulum , but this will have to be verified in future studies . Regardless , a relatively strong affinity to the plasma membrane would help to increase the local concentration of mycolactone on the host membrane , directing its flow toward intra-cellular targets ( e . g . , WASP , the endoplasmic reticulum , and Sec61 ) . Thus , our findings support a passive diffusion mechanism of cellular access for mycolactone , as reported [29] , but further suggest that membrane localization could be playing a direct role in its uptake into intracellular membranes and handoff to cytosolic targets . It is interesting that our CG MD simulations suggest a mycolactone-dependent membrane disruptive effect , albeit at reasonably high toxin concentrations . In fact , all of our calculations demonstrate that the addition of the exotoxin perturbs both structural and dynamic properties of the lipid bilayers . The addition of mycolactone lowers the crystalline-gel Tm by ~5 K in a diC16-PC bilayer , which is associated with the preservation of the fluid phase at lower temperatures . We further associate this reduction in Tm to the disruption of lipid interactions , especially within the glycerol moiety region . Similar effects have been attributed to disaccharides [48] , although the critical concentration is several orders of magnitude higher . From the biological perspective , the observation that mycolactone affects the elastic properties of the model diC16-PC lipid membrane is of relevance . Our calculations suggest a reduction of ~30% and 70% of the compressibility modulus with the addition of 5% and ~10% mycolactone , respectively . Furthermore , mycolactone reduces line tension of membranes , including the critical line tension that reports on the porosity of membranes . The magnitude of changes in membrane properties caused by mycolactone could , in principle , influence the activity of membrane-embedded proteins [49] . Previous studies have addressed how changes in membrane elastic properties can affect glucose transporters [50] , the stability of voltage sensor segments [51] , mechano-sensitive channels of large conductance [52] , and clathrin protein ordering [53] . However , the most likely connection between membrane association and Sec61 inhibition is that trafficking to the endoplasmic reticulum could facilitate mycolactone binding to the Sec61 translocon . In the context of our model biological membrane with ordered and disordered domains , mycolactone potentially behaves as a linactant . Addition of this toxin results in a marked reduction in the lateral line tension between lipid domains in a ternary lipid mixture . Moreover , mycolactone preferentially localizes at the interface between liquid-ordered ( Lo ) and liquid-disordered ( Ld ) domains . The mechanism of disruption seems to correlate with an individual destabilization , with no aggregates or mesoscale structures formed [54] at low concentrations . Interestingly , the toxin shows a slight preference to interact with the unsaturated tails of lipids from the Ld domain . However , the toxin does not accumulate in the bulk of the Ld domain , but rather at the interface , as seen in the electron density profile . The preferential localization of mycolactone at the interface between Ld and Lo domains allows mycolactone to decrease the line tension . This linactant activity of mycolactone could impact the intracellular signaling pathways that are coupled to T-cell receptor ( TCR ) activation . It is possible that mycolactone behaves as a linactant where it can enhance the mixing of disordered and ordered domains , thus promoting the transfer of Lck into a lipid-raft like ordered domain . Finally , our study provides the first glimpse of how mycolactone interacts with membrane and alters membrane stability and dynamics . It is unclear how the properties described above ( i . e . , a preferential localization in the lipid phase , particularly at the interface of ordered and disordered lipid domains , and altered biophysical properties of lipid bilayers ) relate to the cytotoxic nature of mycolactone [19] . Clearly , they will influence the processes by which mycolactone penetrates the host cell and how it is trafficked , both intra- and extra-cellularly . It is also likely that mycolactone will not be presented during infection as a monomeric , water-soluble molecule; rather it will be bound to and carried by host molecules with hydrophobic domains ( e . g . , WASP , N-WASP ) and/or lipid assemblies ( e . g . , bacterial vesicles , host cell membranes , and likely other assemblies such as high density lipoproteins ) . Also , localization at the ordered-disordered interface could assist in the exchange of mycolactone between bacterial outer membrane vesicles and host membranes . Further modeling and biophysical experimental studies are required to clarify these aspects of mycolactone cytotoxicity . We believe that molecular understanding that we gained on how mycolactone interacts with lipids could help the rational development of diagnostics and adjunctive therapies that target mycolactone as it appears during infection . In fact the development of antibodies that target mycolactone derivatives is well underway [34] , but their efficacy in toxin neutralization hinges on mixing the antibodies and mycolactone prior to cell exposure . This again suggests that in a cellular environment , mycolactone is hidden from water-soluble antibodies . Additionally , understanding mycolactone distribution and trafficking may help answer some of the remaining questions about how the toxin kills cells , evades immune responses , and plays a role in angiotensin pathways . Recently , it was shown that mycolactone exerts local analgesia by binding to angiotensin II type 2 receptors and leads to potassium-dependent hyperpolarization of neurons [24] . These findings have led to the exploration of using mycolactone and its derivatives therapeutically to suppress pain and inflammatory responses [55] . Therefore , understanding the nature by which mycolactone interacts with lipids may help in realizing the potential of mycolactone as a therapeutic agent . In summary , understanding the details of the mycolactone interaction with lipids , its influence on membrane dynamics and stability , and its thermochemical behavior in lipid and aqueous environments will be tremendously useful in understanding the variety of ways in which this toxin induces host disease , in the development of diagnostic tools for Buruli ulcer . Moreover , mycolactone has proven to be a helpful test bed for understanding host-pathogen interactions involving amphiphilic molecules . We used coarse-grained molecular dynamics simulations to explore the molecular behavior of mycolactone in lipid membranes . From simulations of mycolactone in aqueous , pure lipid , and mixed lipid bilayer systems , we find that mycolactone prefers the lipid to the aqueous environment , and that it could potentially perturb the thermochemistry of biological membranes in terms of transition temperature , compressibility , and line tension . Interestingly , mycolactone acts as a linactant , i . e . , it localizes at the boundary between different fluid lipid domains in membranes and promotes inter-mixing of domains . It should be kept in mind that in these simulations , we haven’t considered the diversity of components present in a typical biological membrane , but rather very simple models of lipid membranes . Regardless , it is striking to observe disruptive effects induced by this toxin on a lipid bilayer system . We speculate that such effects could easily translate to a broad number of biochemical signatures typical of the pathogenesis of the Buruli ulcer disease [8 , 22 , 24] . In this context , we suggest further studies , both experimental and theoretical , that can capture the effects of mycolactone in more complex systems in which the relevant biological components are present . Note added in proof: Subsequent to the acceptance of this paper we became aware of a recent independent publication that provides experimental confirmation of our prediction that mycolactone would decrease the line tension and thereby disrupt lipid domain formation [56] .
The structural coordinates of mycolactone were downloaded from the PubChem database ( CID 5282079 ) . Mycolactone is structurally inhomogeneous , with variants showing a common core macrocycle and differences in the southern acyl chain . The best studied is mycolactone A/B which is prevalent in Africa , Malaysia , and Japan . The A and B isomers of mycolactone differ by 1800 rotations about the C4’-C5’ double bond and C5’-C6’ single bond in the southern chain; they are in dynamic equilibrium with a 60/40 ratio of mycolactone A to B [57] . Here , we focus our studies on mycolactone B since that conforms to the structural form of mycolactone deposited in the PubChem database . We refer to mycolactone B simply as mycolactone in the manuscript ( Fig 1A ) . The octanol-water partition coefficient ( logPow ) for mycolactone was calculated for both AA and CG systems . Similar calculated logPow will ensure a consistent representation of the model in both representations . Given the appropriate free energy of solvation , the computation of the partition coefficient is straightforward . The difference between the solvation free energy in aqueous ( ΔGW ) and octanol ( ΔGO ) phase is the partitioning free energy ( ΔΔGOW ) ; ΔΔGOW=−2 . 3RTlogPOW where R and T correspond to the universal gas constant and the temperature of the system , respectively . ΔGW and ΔGO were calculated directly by uncoupling the non-bonded interactions of the solute with the respective solvent using the thermodynamic integration approach: ∆FBA=FB−FA=∫λAλBdλ〈∂Uuv ( λ ) ∂λ〉λ Here λ is a coupling parameter that regulates the strength of the interaction of FB ( fully uncoupled ) and FA ( fully coupled ) . Uuv ( λ ) denotes the potential energy function describing the total solute-solvent interaction . The average <…> is taken over the MD trajectory . Calculations were performed at 25 independent λ points . For each individual λ , simulations were run for 50 ns ( AA ) or 100 ns ( CG ) respectively . The obtained water-octanol partition coefficients were compared with values obtained using the Bennett’s acceptance ratio[64] . The results are highlighted in S3 Fig for both octanol and water solvation free energies using the CG and AA simulations . In general , we observe that both approaches provide similar logP values , although the latter slightly increases the preference for the organic phase at AA resolution . The membrane translocation potential of mean force ( PMF ) by mycolactone was calculated using the umbrella sampling approach on the CG system . The simulation was composed of 40 independent windows spaced by 1 Å . A restraining potential of 1000 kJ mol-1 nm-2 was applied to the center of mass of the entire mycolactone with respect of the center of mass of the lipid bilayer and along the normal ( z ) coordinate . For each window , 10 μs long simulations were performed . We should state , however , that MARTINI based simulations lead to an effective speed up of a factor of ~4 [36] , effectively giving a 40 μs time per window . PMFs were reconstructed using the weighted histogram [65] approach and convergence was assessed using the bootstrap method . Also , we assessed the convergence of the calculations using block averaging where the trajectories for each window were divided into independent blocks of 1 μs each . The different PMFs were calculated for each point and averaged . These results are presented in S1 Fig and in good agreement with the bootstrap method . The line tension , or edge energy , of the lipid membrane can be obtained from the critical line tension at which pores can be stabilized inside a membrane . According to a theoretical model [66]: E ( r ) =2πλ−πr2γ where E ( r ) = energy of a pore of radius r inside a membrane , λ is the line tension that opposes the pore formation , and γ is the surface tension that reduces the energetic barrier for pore formation . At low tension , pores are unstable , however , at a critical line tension γ* , the edge energy can be overcome . Thus , pores can be stabilized at r = γ*/λ . The line tension at the interface of two different lipid domains , σ , can be obtained from the bulk pressures measured during the simulation . For an interface along the X-dimension , σ can be obtained through the following expression [44]: σ=12〈LYLZ ( PYY−PXX ) 〉 where <…> denotes an ensemble average , LY and LZ are the box dimensions along the Y and Z edges respectively , and PYY and PXX are the pressure tensors perpendicular and parallel to the interface . The factor ½ accounts for the fact that there are 2 interfaces in the simulation system . The area compressibility modulus , KA , defined within one plane can be calculated from the fluctuations in the membrane area per lipid [67]: KA=kT<A0>N< ( A−A0 ) 2> where N denotes the number of lipids per monolayer , A0 is the averaged area per lipid and A the fluctuation of the area per lipid from the simulation trajectory . kT is 2 . 49 kJ mol-1 at 300 K . It has been observed that better agreement with experimental data is found for simulations of large membrane patches , as undulation modes can be captured .
|
Mycolactone is a macrolide exotoxin secreted by Mycobacterium ulcerans , which causes a skin disease called Buruli ulcer , a neglected emerging disease . It is the third most common mycobacterial disease after tuberculosis and leprosy . Studies have shown how mycolactone plays a pivotal role in Buruli ulcer pathogenesis , and identified it as an attractive therapeutic target . This multifunctional cytotoxin exerts multiple local and global responses , including ulcerative , analgesic , and anti-inflammatory effects . Prompted by its lipid-like structure , we used extensive multi-resolution simulations to probe mycolactone’s interactions with model membranes . Our results suggest that mycolactone is sequestered in membranes where it alters several dynamical , physical , and mechanical properties . It also behaves as a linactant , localizing at the interface between lipid domains and decreasing the inter-domain line tension . Our results shed light on how mycolactone permeates host cell membranes and is distributed between lipid and aqueous environments . These findings have significant implications for the toxin’s distribution in the host environment and mechanisms of pathogenicity . Understanding the toxin’s distribution and mechanism of trafficking will have ramifications for targeted diagnostics , therapeutic approaches , and our understanding of Buruli ulcer pathogenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"toxins",
"pathology",
"and",
"laboratory",
"medicine",
"membrane",
"potential",
"tropical",
"diseases",
"electrophysiology",
"toxic",
"agents",
"toxicology",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"cellular",
"structures",
"and",
"organelles",
"thermodynamics",
"infectious",
"diseases",
"lipids",
"buruli",
"ulcer",
"cell",
"membranes",
"free",
"energy",
"physics",
"biochemistry",
"biochemical",
"simulations",
"cell",
"biology",
"physiology",
"lipid",
"bilayer",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology"
] |
2018
|
Membrane perturbing properties of toxin mycolactone from Mycobacterium ulcerans
|
Structural inhomogeneities in synaptic efficacies have a strong impact on population response dynamics of cortical networks and are believed to play an important role in their functioning . However , little is known about how such inhomogeneities could evolve by means of synaptic plasticity . Here we present an adaptive model of a balanced neuronal network that combines two different types of plasticity , STDP and synaptic scaling . The plasticity rules yield both long-tailed distributions of synaptic weights and firing rates . Simultaneously , a highly connected subnetwork of driver neurons with strong synapses emerges . Coincident spiking activity of several driver cells can evoke population bursts and driver cells have similar dynamical properties as leader neurons found experimentally . Our model allows us to observe the delicate interplay between structural and dynamical properties of the emergent inhomogeneities . It is simple , robust to parameter changes and able to explain a multitude of different experimental findings in one basic network .
Distributions of synaptic weights are known to have a large influence on the dynamics and information-processing properties of neural circuits [1–5] . Recent electrophysiological studies have shown that the distributions of the amplitudes of excitatory postsynaptic potentials ( EPSPs ) in cortical [6 , 7] as well as hippocampal [5] networks are typically long-tailed and span several orders of magnitude . This characterizes a topological configuration in which the majority of synaptic connections are weak and a small number are much stronger [6–9] . At the same time , distributions of firing rates during spontaneous activity have been found to be long-tailed [10] and there is increasing evidence that long-tailed distributions play a fundamental role in brain functioning [11–13] . Studies of microcircuit connectivity have also demonstrated a number of significant inhomogeneities , as well as correlations amongst cell activity . Fine-scale functional subnetworks have been found in cortical networks [1 , 3 , 14] , and it has been shown that cells with strong outgoing synapses cluster together [3 , 7] . Both the clustering of the highly active cells [3] and the presence of strong synapses [1] are likely to play an important role for network dynamics and information processing in the neocortex . Apart from structural inhomogeneities in networks , the impact of individual neurons on the dynamics of neural networks may also differ substantially . A number of recent in vitro studies of 1D and 2D dissociated developing cortical and hippocampal cultures have shown that such networks typically express spontaneous neural activity characterized by network bursts , and ongoing repetitions of distinctive firing patterns within those bursts [15–19] . Furthermore , several studies have shown that the activity of certain neurons reliably precedes population spikes [16–19] . These early-to-fire neurons have been termed leader neurons [18] and have been found to form functionally connected networks , the activity of which collectively precedes most of the observed population bursts [19] . In the 1D case , population bursts have been found to be triggered by “burst initiation zones” [17] and in the 2D case recent studies [18 , 20] have shown that leader neurons not only precede but also are able to initiate population bursts . Nevertheless , the underlying network structure and specific topological properties of leader neurons and subnetworks of such cells remain to be discovered . Experimental studies in constrained [21] or chemically modulated [22] cultures give reason to believe that a complicated process of self-organization underlies their emergence . However , from the modeling point of view , little is understood about how strong inhomogeneities , such as the aforementioned , could evolve in a self-organized manner by means of activity-dependent synaptic plasticity . Synaptic plasticity is widely believed to be the basis for learning and memory and shapes the distribution of synaptic weights , which has been found to be typically long-tailed in a number of recent experimental studies [6 , 7] . The influence of long-tailed weight distributions on network dynamics has been studied in a number of recent works [2 , 4 , 5 , 12 , 23] , and it has been shown that such distributions can increase spike-based information transfer and facilitate information processing in neural networks [5 , 24] . Yet , also in the latter case little is known about how experimentally observed activity-modulated synaptic plasticity , such as spike-timing-dependent plasticity [25–27] ( STDP ) and homeostatic plasticity [28 , 29] could lead to a symmetry-breaking in the distributions of synaptic weights of an initially homogeneous network , or to the preservation of inhomogeneities prescribed by certain initial conditions [30 , 31] . Even less is known about how networks could self-organize to simultaneously express the aforementioned properties such as having both long-tailed distributions of weights and firing rates at the same time [2] , cells with different dynamical effects on network activity , and subnetworks with distinct dynamical and structural properties similar to the ones observed experimentally . Although recent theoretical models have proposed mechanisms that lead to the emergence of long-tailed synaptic weight distributions [32 , 33] , and in case of the SORN model [33] to other interesting aspects of self-organization , these either employ a specially tailored plasticity rule for this purpose [32] , or do not express long-tailed firing-rate distributions [33] . It is well known that networks of spiking neurons can exhibit strongly irregular dynamics if excitatory and inhibitory inputs to each neuron balance , such that the network is driven by fluctuations in its input , resulting in each neuron producing spikes at irregular times [24 , 34 , 35] . Such networks are called balanced state networks . They combine computational simplicity and dynamics that closely resembles activity recorded electrophysiologically in vivo from cortical cells of behaving animals [36 , 37] . This makes balanced state networks a very attractive and widely used theoretical model for cortical tissue [34 , 35 , 38 , 39] . Our goal for this study was to investigate processes of self-organization in such networks brought about by means of synaptic plasticity . We therefore consider a random network of spiking neurons in the balanced state , operating in the asynchronous irregular ( AI ) regime [34 , 35] that is believed to be a good fit to the activity of cortical networks in vivo [36 , 37] . We endow it with two activity-dependent synaptic plasticity rules , namely spike-timing-dependent plasticity [26 , 40 , 41] ( STDP ) and synaptic scaling [29] . In its prototypical form , STDP causes long-term potentiation ( LTP ) of a synapse if presynaptic spikes repeatedly occur some milliseconds before postsynaptic ones , whereas a reversed temporal order causes long-term depression ( LTD ) . Since its initial discovery at glutamatergic synapses [25 , 26 , 41] , many forms of STDP have been observed experimentally [42] , also such forms acting at GABAergic synaptic connections [43–47] and many models have been proposed [27 , 48–52] to describe the mechanisms and dynamics of STDP . In particular , STDP acting at inhibitory-excitatory connections has been shown to influence spiking dynamics of hippocampal pyramidal neurons [45] . Several recent modeling studies have shown that inhibitory STDP has a stabilizing effect on network dynamics [53–56] , and others also have started addressing questions of functional aspects of inhibitory plasticity [57 , 58] . Another well-studied form of synaptic plasticity is synaptic scaling [29 , 59] , a form of homeostatic synaptic plasticity [28] that describes the up- and down-regulation of a neuron’s synaptic input in order to keep a preferred mean target firing rate . It is well known that STDP alone can lead to instabilities in network dynamics due to effects of positive resonance which result in runaway excitation , and that endowing a random network solely with a multiplicative or power-law STDP rule acting at excitatory-excitatory synaptic connections while keeping the other synaptic efficacies fixed does not lead to stable effects of self-organization [30 , 31] . Combinations of STDP and synaptic scaling , however , are known to be able to keep network dynamics in a stable and biologically plausible regime [60] , and to support non-trivial computations underlying many optimization tasks [61] . Furthermore , it has been shown that combining Hebbian and homeostatic plasticity rules both has a stabilizing effect on network dynamics [62] , and it has been found to express structure-building properties in simple model networks [63] of rate-based neurons . Here , we investigate how the inclusion of synaptic scaling and inhibitory STDP could bring about self-organization in spiking networks . Our model network develops both long-tailed distributions of synaptic weights and firing rates , similar to those observed experimentally [6 , 10] . Moreover , a delicate interplay between dynamics and synaptic plasticity leads to the emergence of a special group of neurons that we call driver neurons . They form subnetworks that can take strong influence on network dynamics and share properties of certain neurons and subnetworks found experimentally [1 , 3 , 18 , 19] . The phenomena that we observe are generic and hold under alternations of the plasticity rules and their parameters .
After a transient phase , the weight distributions of the dynamic E-E and I-E synaptic connections have settled to their new stable shapes ( see Fig 1 ) . The mean synaptic strength of E-E connections is kept fixed at a value of 1 by the synaptic scaling rule , but the variance grows rapidly ( see Fig 1 , 1D ) . We observed that E-E connections undergo a symmetry breaking and that we obtain a long-tailed distribution of synaptic weights after convergence ( see Fig 1 , 1B ) . The EPSP distributions found in cortical and hippocampal networks of excitatory neurons are typically long-tailed [5–7] . Such types of weight distributions can lead to optimal enhancement of the responses of individual neurons to input and are thus beneficial for information transmission at strong synapses , as was shown recently [4 , 5 , 24] . We used a maximum likelihood estimator for the exponent to fit a power-law distribution with a cutoff into the middle part of the synaptic weights distribution , omitting the strongest 5% of the excitatory synaptic connections . Our fitting procedure was modified from [64] . We found a power-law exponent α = −1 . 92 and an upper cutoff xmin = 0 . 205 , Fig 1B . Although visually the fit does look good , with the amount of data we produced , it is possible to reject the power-law hypothesis [64] . However , it is clear that the truncated distribution has a long-tail behavior and that the probability decays approximately as a power-law . We obtained similar results when we exchanged the additive STDP rule with a partly or fully multiplicative one ( see Section Different STDP rules ) . The weights of the I-E connections evolve to a near Gaussian form . This is due to the fact that inhibitory STDP is subject to negative feedback and thus yields unimodal distributions of synaptic strengths even in the case of a purely additive plasticity rule [54] . In addition to expressing long-tailed weight distributions after convergence , the network rests in the AI regime and expresses approximately log-normal firing rate distributions throughout the transient state ( see Fig 1A ) . We observe many cells firing at very low rates close to 0 Hz and only a few cells firing at rates up to 30 Hz , a property in line with experimental data obtained during spontaneous cortical activity in vivo [10] and ubiquitous in brain networks [12] . While log-normal rates are known to be a general and robust property of random balanced state networks with homogeneous weights [23] , the combination of both long-tailed distributions of firing rates and synaptic weights is not a straightforward property [2] . Plasticity in the network leads to the development of few exceptionally strong excitatory synapses ( see Fig 1B ) . Interestingly , many of these synapses are found on excitatory neurons which have predominantly strong outgoing connections and which fire at higher than average rates . As we will show in the following , the elevated firing rates are in fact causal for the emergence of their strong outgoing weights ( see Section Emergence of driver neurons ) . We call these neurons driver cells ( or driver neurons ) and characterize them by distributions of outgoing excitatory synaptic weights with a high mean value ( see Fig 2A ) . To define the group of driver neurons we take the top 0 . 5 percent of the excitatory cells with the largest mean outgoing synaptic strength ( see Fig 2A ) . As our network consists of 4000 neurons , this amounts to 20 driver cells in the network . As the distribution of mean outgoing weights in the network is unimodal , there is no clear-cut threshold separating any group of cells from the rest of the population in this distribution . The particular choice of the threshold results in a very strong dynamical impact of the driver neurons as discussed in section Dynamical impact of driver neurons . Another possibility to define driver neurons would be to use a threshold located at three standard deviations above the mean . In this way we select neurons that have much stronger outgoing weights than expected under the hypothesis of normally distributed mean outgoing strengths . In contrast to the case of a normal distribution where this choice will result in an expected 0 . 23% of all cells , we classify 3% of the population as driver neurons using the aforementioned criterion . Using this choice , there is also a detectable dynamical difference between the driver neuron group and the rest of the network , though not as strong as for the 0 . 5% threshold ( see Section 9 in S1 Text ) . The clustering of strong outgoing synapses in the network is shown in Fig 3 ( see also Fig U , top left in S1 Text ) . Here , we plotted the quantiles of the distribution of the mean synaptic strength per neuron in the original network . For comparison we also plotted quantiles of surrogate data obtained by shuffling the synaptic weights among all excitatory synapses . This operation destroys all correlations in synaptic weights introduced by the plasticity mechanisms , while leaving network topology and the overall distribution of synaptic strengths unchanged . As we can see in Fig 3 , for the shuffled networks ( we show the mean and the standard deviation of 100 shufflings , standard deviation very small and almost invisible in the plot ) , most cells have a mean outgoing weight of around 1 , the mean excitatory weight in the network . The self-organization of driver cells is due to a delicate interplay of network dynamics and synaptic plasticity . For example , driver neurons are much less pronounced in a model network where we do not include inhibitory STDP ( see Fig 2B ) . We will describe the process of their emergence in more detail in the following . Looking at driver cells in the equilibrium network and at the same cells early in the network evolution ( that we call future driver cells ) , we found that they fire at rates much higher than the network average ( network average approximately 5 Hz , driver group average approximately 25 Hz , see Fig 2A ) . In the following , we will show that this is the main reason for the emergence of their strong outgoing synaptic connections: STDP dynamics of the excitatory synaptic weights in our network can be seen as a random walk on the closed interval [0 , wmax] . Here , the probability to increase the weight grows with increasing synaptic weight , and for weights above a certain threshold the average impact on synaptic weight of each presynaptic spike is positive . In a model without homeostatic plasticity we observed that once a synaptic weight reaches this threshold it converges to its maximum with high probability , with a velocity proportional to the firing rate of the neuron . As we included a postsynaptic homeostatic plasticity rule in our model at E-E synapses that constrains the total sum of weights onto each neuron , this led to a competition of all excitatory synapses converging onto a given excitatory postsynaptic cell over a limited pool of total synaptic efficacy . One characteristic of many STDP rules as well as the one that we include in our model is that synapses connecting a highly active presynaptic cell with a less active postsynaptic one ( in terms of their mean firing rates ) tend to undergo LTP [65 , 66] . Thus , outgoing synapses of driver cells that fire faster than the average cell have a higher probability to undergo LTP . Synapses from future driver cells are the ones to predominantly win that competition over available synaptic efficacy , diminishing the influence of other cells ( see Fig 4B and Section 4 in S1 Text ) . This ultimately allows driver cells to emerge and to have strong influence on their postsynaptic networks . For our model we support this by analytical considerations ( see Methods ) . We observe that the higher firing rates of ( future ) driver cells is due to reduced inhibitory currents those cells receive ( see Fig C , left in S1 Text ) . Currents in our model are influenced by two variables , synaptic weights and presynaptic firing rates . We find that the reduced inhibitory currents that drivers receive are a result of two separate effects of local network topology: First of all , driver cells have a lower than average number of converging inhibitory synapses . Secondly , inhibitory cells which are presynaptic to driver cells have on average a higher number of converging inhibitory synapses than randomly selected inhibitory cells of the network . The latter results in lower than average firing rates of the inhibitory cells presynaptic to driver cells ( see Fig D in S1 Text ) and the combination of these two effects leads to a permanently reduced inhibitory drive to driver cells . As the included inhibitory STDP rule is subject to a form of self-regulatory dynamics with negative feed-back that becomes stronger with increasing synaptic weight [54] , inhibitory plasticity cannot fully compensate this reduced inhibitory drive to driver cells by increasing inhibitory weights: On the one hand , inhibitory STDP in our model tries to compensate the high rate of the under-inhibited neurons by increasing the converging inhibitory weights to those cells . On the other hand , each inhibitory presynaptic spike delays the postsynaptic spike for an increasing period of time with increasing inhibitory synaptic weight , thus decreasing the positive contribution of the STDP rule . This results in a situation in which the converging inhibitory weights become stationary at a value below the maximal inhibitory weight , even in the case of high postsynaptic firing rates as seen in the case of driver neurons ( see Fig 4 ) . Altogether , we thus find that driver cells in our model are mainly determined by ( local ) network topology and that their emergence is due to an interplay of all three plasticity rules active in the network . In order to test the sensitivity of the observed effects to changes in the network size , we simulate networks of 10 , 000 and 20 , 000 cells and find that also in these we obtain qualitatively similar results ( see Section 10 in S1 Text ) . In the following , we will investigate dynamical and topological properties of the group of driver cells and compare them to randomly sampled groups of non-driver cells of the same size . One question we wanted to answer is whether the emergence of driver neurons influences the dynamics of the network . To answer this question , we forced both the group of driver cells and a group of randomly selected non-driver cells to fire two consecutive spikes . We observed the network response in both cases . We achieved this by providing two brief pulses lasting 0 . 5 ms of a very strong constant current input to the group of stimulated cells . Those two pulses were separated by a delay of 2 ms to allow all cells to leave their refractory periods after emitting the first spike . We furthermore only considered cases in which none of the stimulated cells was refractory prior to stimulation so that all cells of the stimulated groups fired exactly two spikes within 2+ϵ ms , where ϵ ≪ 1 ms is dependent on the membrane potential of the cell prior to the stimulation . We chose this stimulation protocol to imitate a bursting activity in the chosen subpopulation . To probe a baseline response of the network , we stimulated the same number of randomly selected non-driver cells with the same protocol . In the latter case , the network firing rate rises shortly due to the induced simultaneous firing of the stimulated neurons , but there are no lasting effects on network dynamics ( see Fig 5B ) . On the other hand , the stimulation of the driver cell group results in a prolonged elevation of the firing rate similar to a population spike ( see Fig 5A ) . To get a more precise picture , we observed the network dynamics subject to the condition that a number of driver neurons spontaneously fire in a synchronous way . We considered events in which a certain fraction of the driver cells all fire within a 1 ms time bin and average the excitatory population rates before and after this event , obtaining a synchrony triggered average ( STA ) curve ( see Fig 6A ) . For comparison , we sampled a random group of non-driver cells of the same size and considered coincident spikes from its members ( see Fig 6B ) . The STA curves for the random group are symmetric around the moment of synchronization . This indicates that the probability of finding a given number of neurons from this group firing coincidently within 1 ms is higher when the network rate is higher than usual , but this event has no effect on network dynamics . On the contrary , two synchronous spikes from the driver group are sufficient to result in a detectable and causal elevation of the population firing rate . This effect becomes more pronounced for larger groups of driver cells firing synchronously ( see Fig 6A ) . Choosing the top 3% of cells with the highest mean outgoing weights as drivers predictably diminishes the absolute impact of synchronous driver spiking on network dynamics , but qualitatively the results remain unchanged ( see Section 9 in S1 Text ) . We also performed the same averaging in a setting in which a selected number of driver and random non-driver neurons are forced to spike coincidently by providing a brief stimulation with a strong constant current input ( see Fig 6C ) . The response of the network to the additional synchronous spiking of driver neurons is similar to the case of spontaneous synchrony . A prolonged phase of elevated activity is observed and the effect grows with the number of activated neurons . On the other hand , even after a stimulation of many non-driver neurons , activity rapidly returns to the base-line level and no prolonged change in the network dynamics can be observed . A recent series of experimental studies performed in vitro [18–20] in dissociated hippocampal and cortical cell cultures reported the existence of certain special neurons termed leader neurons . They were characterized by their firing activity stably preceding population bursts in the culture . Leader neurons were found in a wide variety of dissociated cultures obtained from both hippocampal and cortical cells from embryonic , newborn ( < 24h ) and juvenile ( P16–17 ) rats [18] and under a wide variety of feeding protocols . It was found that synchronized firing of leader neurons increases the probability of the initiation of a population burst above chance level [18] . Furthermore , recent experimental work [20] shows that leader neurons do not just passively precede bursting activity in the cultures , but can actively trigger it . Whereas our model network stays in the asynchronous irregular activity regime that differs greatly from the synchronized bursting behavior of the hippocampal cultures , there are still periods of elevated activity , and driver neurons fire preceding them and can cause such events if many drivers are triggered to fire simultaneously , a property shared with leader neurons . There are further properties that the driver neurons in our model share with leader neurons , for example the tendency to form functional subnetworks [18 , 19] ( see Section Stability and topology of the driver neuron subnetwork ) . Moreover , it was shown [16–18] for leader neurons that the property of early spiking during population bursts is very likely to be caused by synaptic input ( reduced inhibition or increased excitation , or a combination of both ) to those cells and not intrinsic cell properties ( e . g . a reduced firing threshold ) , a finding we also made for the driver neurons in our model . It is an interesting open question to elucidate what will happen if the network is set to stay in the bursting regime throughout its development , that we consider for a follow up publication . Moreover , a recent experimental study [3] assesses the existence of subnetworks of highly active excitatory cells in the somatosensory cortex of juvenile mice , expressing both characteristics of leader neurons and driver neurons and thus giving further experimental support for a unification of the two concepts . In earlier studies of balanced state networks with a power-law or multiplicative STDP rule acting at excitatory-excitatory connections [30 , 31] , strong synapses arising due to a temporary symmetry breaking were not stable and disappeared after some time . In our model the strong synapses diverging from the driver neurons remain strong over long periods of time . Consequently , the property of belonging to the group of driver cells is stable over long periods of network time . This is due to the fact that in our model the property of becoming a driver cell is mainly determined by local network topology . It is an interesting question what will happen in much larger networks . To test for the sensitivity of the results to network size , we simulated networks of 10 , 000 and 20 , 000 neurons . We found that increasing network size in the considered ranges did not significantly change the shape of the converged weight distributions or the clustering of strong outgoing weights ( see Section 10 in S1 Text ) . We furthermore observed that driver cells form “rich club” subnetworks in which most synapses are strong . This can be explained by the fact that driver cells emerge in waves , recruiting cells from their postsynaptic networks . As the outgoing synaptic connections of future driver cells which fire at much higher than average rates undergo LTP , this leads to elevated excitatory currents in the cells postsynaptic to driver cells and to an increase in their firing rates . This , in turn , increases the chances of the postsynaptic cells of also becoming driver cells . This iterative process terminates at some point as the total available synaptic weight in the network is limited by a homeostatic rule and the fact that inhibitory STDP up-regulates converging inhibitory weights onto cells with elevated firing rates . To illustrate the effect , we simulated 1000 different networks and looked at the driver cells that emerged in those networks and their subnetworks . To measure the connectedness of the subgroups we studied the number of synaptic links within two groups of n = 20 neurons each , the group of driver neurons and a group of randomly sampled non-driver neurons . We found that the driver cell group has a significantly higher number of synaptic connections Cdriver compared to the number of synaptic connections Crand in the random group . The random group on average expressed Crand = 7 . 35 ± 3 . 30 ( mean ± standard deviation ) synaptic connections in their subnetwork . Not surprisingly , this is very close to the number of expected synaptic connections in a random network of 20 nodes with connection probability p = 0 . 02 which is C0 = 20⋅19⋅0 . 02 = 7 . 6 . In contrast , we found on average Cdriver = 12 . 14 ± 2 . 65 predominantly strong synaptic connections in the driver neurons subnetwork , an almost twofold increase compared to the random group . Similar relations were found in recent experimental studies of developing cortical networks . Stable subnetworks of more active cells were found to express a higher amount of connectedness [3] . Moreover , a tendency to higher mean EPSP amplitudes with lower variances within highly connected subnetworks was found [7] . The question remains whether there are other topological properties that distinguish driver neurons from the rest of the population , apart from the reduced inhibitory in-degrees . To answer this question we measured both excitatory and inhibitory in- and out-degrees throughout 1000 different networks . We found that the number of incoming and outgoing excitatory synaptic connections does not distinguish driver neurons from the rest of the network in our model , as we can already see in the example network ( see Fig 7A ) . To asses the observed difference in inhibitory in-degrees more clearly , for each network simulation we extracted the driver group after convergence of the synaptic weights and took one random group of non-driver neurons of the same size . As expected from the definition , the distribution of outgoing excitatory synaptic weights allows to distinguish the two groups of neurons dramatically ( see Fig 7C ) . The distributions of incoming inhibitory degrees are also easily separable across the two groups ( see Fig 7D ) . To compensate the smaller incoming inhibitory degree , inhibitory STDP up-regulates inhibitory weights converging onto driver cells such that they receive on average much stronger incoming inhibitory weights ( see Fig 7E ) . But as discussed earlier , they none the less receive reduced inhibitory currents when compared to the network average ( see Section Inhibitory STDP ) . To verify our hypothesis that driver cells in our model are mainly distinguished by properties of local network topology , we performed simulations of 1000 networks with varying initial conditions but a fixed network topology . To induce statistical fluctuations , we stimulated the networks with Poisson noise and then looked at the probability ( assessed via the relative frequency ) of each cell to belong to the driver group after the synaptic weights have converged . We found that the driver neuron population remained mainly unchanged independently of initial conditions and that roughly 30 different neurons were found in the driver cell group across all trials ( see Fig 7B ) . The source of the variation between the outcomes is in the random initialization and input fluctuations . Altogether , the property of belonging to the driver group is thus not solely but mainly dependent on network topology . Another interesting question is how large the variation in local network connectivity has to be in order to allow for the emergence of driver cells . To answer this question , we simulated a fully homogeneous network in which all cells have the same in-degree of both excitatory and inhibitory connections . In this case the weight distributions almost remained delta peaks , i . e . each synaptic weight stayed w ≈ 1 even subject to plasticity and no driver cells emerged ( see Fig I , left in S1 Text ) . Interestingly , already a slight amount of under-inhibition suffices to allow for the emergence of driver neurons . We demonstrated this by taking a random group of 50 cells in the fully homogeneous network and selectively pruned 10% of the inhibitory synapses converging onto each cell of the group . We found that already this small change in homogeneity suffices to allow the group to become driver cells ( see Fig I , right in S1 Text ) . What is the role of inhibitory plasticity in our model ? Similar to recent theoretical studies [53–56] , inhibitory STDP plays a stabilizing role in our network setup . Yet , inhibitory plasticity furthermore plays a crucial role in the emergence of driver neurons , as we will see in the following . In order to assess the effect of inhibitory STDP more closely , we simulated networks with static I-E synaptic connections of constant weight ( see Section 7 in S1 Text ) . The magnitude of the constant inhibitory weights was selected to be equal to the mean of the equilibrium distribution of inhibitory weights in the same network with inhibitory STDP . Without inhibitory STDP , the network still exhibits mostly asynchronous irregular activity , but with a higher amount of oscillations and with some cells expressing firing rates of up to 100 Hz ( see Fig F in S1 Text ) . Also , the excitatory population firing rate almost triples ( with a mean rate of around 15 Hz ) in this case , whereas the inhibitory population firing rate does not increase so drastically . The excitatory weight distribution is also similar to the case of a network including inhibitory STDP ( see Fig G in S1 Text ) , but outgoing strong weights cluster much less on the subset of neurons that constitutes the driver cell group in the former case ( see Fig 2B and Fig U in S1 Text ) , making their dynamical impact on network dynamics much smaller ( see Fig H in S1 Text ) . We furthermore found that this result does not depend on the actual value of the fixed inhibitory weight , but that it can be observed even for very weak or strong fixed inhibitory connections . How can this be explained ? The difference between plastic and static inhibitory connections lies in the selective nature of the Hebbian inhibitory STDP rule [44] . Namely that it increases the synaptic weights converging on cells with higher firing rates more strongly than the ones converging onto cells with lower firing rates ( see Fig 4A ) . This is the mechanism underlying its stabilizing property . Without inhibitory STDP , slightly under-inhibited neurons ( that could develop into driver neurons in the fully plastic network ) fire with higher rates and by this alone increase the firing rate of their postsynaptic partners . This happens on time-scales that are much shorter than the ones of synaptic plasticity . Thus , neurons that are post-synaptic to under-inhibited cells attain higher firing rates and start competing with their presynaptic partners over the available pools of postsynaptic weights limited by the synaptic scaling rule ( see Fig 4B ) . This process , although also yielding long-tailed distributions , prevents the strong clustering of strong outgoing synapses on single cells ( see Fig 2B and Section 13 in S1 Text ) . The cells with the highest mean outgoing weight also have an impact on network dynamics in this case , but this is much less pronounced than in the case of a network including inhibitory STDP ( see Fig H in S1 Text ) . Do the results depend on parameter tuning ? Do the the results generalize to learning rules other than additive STDP ? For this we will refer to the previously described model with additive excitatory and inhibitory STDP rules and a synaptic normalization at excitatory synapses as the base model and study variations of it .
In this paper we examined the self-organization of inhomogeneous synaptic strengths in balanced networks . Beyond the development of long-tailed weight and rate distributions , we observed a clustering of the strongest outgoing synapses on a few neurons that we call driver neurons . This clustering stays qualitatively the same for different modifications of the STDP rules , homeostatic regulations , and network topology . Our analytic results demonstrate how the network enhances small initial inhomogeneities by a combination of three plasticity rules: excitatory STDP , inhibitory STDP , and homeostatic plasticity . We furthermore showed that inhibitory STDP can serve not only the purpose of circuit stabilization , but also how it might be central for structure formation in networks .
The sub-threshold membrane potential of each LIF neuron obeys τ m d V d t = - ( V - E L ) + I e syn - I i syn , where τm = 20 ms is the membrane time constant , EL = −60 mV denotes a leak term and I e syn , I i syn denote excitatory and inhibitory synaptic currents , respectively . Whenever the membrane potential crosses a spiking threshold Vthres = −50 mV , an action potential is generated and the membrane potential is reset to the resting potential Vreset = −60 mV , where it remains clamped for a refractory period τref = 2 ms . The excitatory synaptic currents are given by I e syn = w e c e norm g e , where ge denotes the presynaptic spike train that is convolved with a synaptic kernel function , c e norm = 1 mV is a normalizing factor and we denotes the synaptic weight normalized to values [0 , wmax] with w e max = 20 , and an initial weight we = 1 . The negative synaptic current I i syn is defined analogously with c i norm = − 9 mV , w i max = 5 , and an initial synaptic weight wi = 1 . Synaptic connections are current-based with exponential kernel functions τ e d g e d t = − g e and τ i d g i d t = − g i . Here , τe and τi denote excitatory and inhibitory synaptic time constants , respectively . They are chosen as τe = 5 ms and τi = 10 ms , in accordance with fast-acting excitatory and inhibitory neurotransmitters . Synaptic parameters are chosen so that effective EPSP and IPSP amplitudes are comparable with experimental data [88] . EPSP amplitudes take values between 0 mV and 2 . 25 mV ( corresponding to a synaptic weight of w e max = 20 ) , with 0 . 16 mV corresponding to an excitatory synapse with weight w = 1 . IPSP amplitudes take values between 0 mV and −11 . 23 mV ( corresponding to a synaptic weight of w i max = 5 ) , with −2 . 25 mV corresponding to an inhibitory synapse with weight w = 1 . We consider a random , balanced state network of leaky integrate and fire neurons consisting of N = 5000 cells of which 4000 are excitatory ( E ) and 1000 inhibitory ( I ) . The network is fully recurrent with all four connection types E-E , E-I , I-E and I-I present . The probability of a synaptic connection between any two neurons is p = 0 . 02 , a value chosen as a compromise between the higher connection probabilities found for neighboring cortical neurons and the lower values for more distant cells [6] . Synaptic connections are current-based with an exponential decay and modeled to be in accordance with fast acting glutamatergic and GABAergic neurotransmitters . In order to ease simulation and analysis we restrict our model to have mono-synaptic connections between pairs of cells , aggregating possibly several synaptic contacts of one pair of cells into one postsynaptic potential ( PSP ) . The E-E ( STDP+homeostatic plasticity ) and I-E ( inhibitory STDP ) synaptic connections are dynamic , whereas the I-E and I-I ones are static ( see Fig A in S1 Text ) . STDP in our model is implemented in a standard on-line fashion with exponential kernels and all-to-all spike pairings so that the weight update for a synapse connecting a pre- and a postsynaptic cell is given by Δ w ( w ) = { A + ( w ) exp ( - ( t post - t pre ) / τ + ) if t post - t pre > 0 - A - ( w ) exp ( ( t post - t pre ) / τ - ) if t post - t pre ≤ 0 , where tpre and tpost denote pre- and postsynaptic spike times . For excitatory-excitatory synapses , we consider either additive ( A+ ( w ) and A− ( w ) constant ) , partly multiplicative ( A+ ( w ) constant ) or fully multiplicative STDP rules , see S1 Text for a description of the different rules . For inhibitory-excitatory synapses we only consider additive rules as in this case additive and multiplicative rules are equivalent [54] . For both excitatory and inhibitory connections , time constants were set to τ+ = τ− = 20 ms . For additive excitatory STDP , the amplitudes of LTP and LTD were chosen as A+ = 10−3 and A− = 1 . 05A+ , respectively , resulting in a negative integral of the STDP window . For the case of partly multiplicative and multiplicative STDP at excitatory synapses we chose A+ = A− = 10−3 . For inhibitory connections , we set A− = 10−3 and A+ = 4A− as motivated by experimental findings [44] and yielding a positive integral of the STDP window [54] . In the simulations , A+ and A− are multiplied with the respective maximal weight for additive rules to obtain their effective values . Like the excitatory STDP rule , inhibitory STDP [44] in our model is Hebbian , increasing the synaptic weight if the postsynaptic cell fires within τ+ ms after a presynaptic spike , and decreasing it when a presynaptic spike occurs within τ− ms after a postsynaptic spike . We verified that the results do not strongly depend on the learning rates of the STDP rules by varying them one order of magnitude into each direction . This influences the convergence speed of the synaptic weights to the equilibrium distribution , but not the shape of the distribution itself . Homeostatic plasticity is implemented in form of synaptic weight normalization acting at the postsynaptic site of excitatory-excitatory connections . The normalization rule is defined by w scaled ( i ) : = 1 ∑ j w j in ( i ) deg EE in ( i ) w in ( i ) , where for an excitatory cell i , win ( i ) denotes the vector of incoming excitatory weights with components w j in ( i ) and deg EE in ( i ) denotes the excitatory in-degree . In the simulations , the above normalization of weights is applied every 100 ms , replacing the weight vector at each cell with its normalized version . We note that for long simulation times synaptic normalization is equivalent to synaptic scaling , assuming equal mean rates of the presynaptic cells and slow time-scales of plasticity . Yet , we also performed simulations of the network with synaptic scaling mechanisms acting on a slower timescale according to the following rule , analyzed for rate-based models as in [62]: d w d t = - γ ( ν - ν 0 ) w 2 . Here , ν denotes the firing rate of the cell , ν0 a target rate and γ a learning rate . For the simulations we chose γ = 10−6 and ν0 = 0 Hz . A scaling step was performed each 50 ms during the simulation and firing rates were computed using a sliding window of length 100 s . We observed that these networks show qualitatively the same behavior as the ones with the synaptic normalization rule , see Section 12 . 3 in S1 Text . At the start of the simulation , the membrane potentials of the neurons were drawn from a uniform distribution between Vrest and Vthres . Subsequently , the network was driven by a constant depolarization of each cell , sufficient to depolarize each cell by 11 mV . This input drove the network to the asynchronous irregular ( AI ) regime of activity with a mean population firing rate close to 5 Hz . We chose a constant depolarization as input since we wanted to study self-organization in the network brought about by its own dynamics , rather than some structure present in the input . Additionally , this case was previously studied in [71] , where also some properties of the static network were assessed , such as the expression of asynchronous irregular spiking activity . The distribution of firing rates and interspike intervals ( ISIs ) in the network is long-tailed with few cells firing at rates up to 30 Hz and many at low rates below 0 . 1 Hz . As expected , the mean value of the distribution of coefficients of variation of the ISIs is close to 1 , indicating irregular spiking activity of the network . In an alternative setup , we tested a network in which each cell is stimulated by a Poisson spike train of the same mean intensity and noted that this setup results in qualitatively the same results . We started with a network in which initially all synaptic weights have a constant value of 1 for all four types of synaptic connections between excitatory and inhibitory cells and then activate the synaptic plasticity rules . We also simulated networks with initially Gaussian and uniform weight distributions and obtained qualitatively identical results . After a transient phase lasting around 5 hours of network activity , the weights stabilized to their new long-tailed distributions . We observed that during this transient phase the network rests in the asynchronous irregular regime of activity . No significant difference in the mean firing rate of the different populations and no apparent visual difference in raster plots before and after plasticity can be observed , see Fig B in S1 Text . We furthermore verified that the obtained results do not strongly depend on the learning rates chosen by systematically varying them around the chosen value , increasing and decreasing them by up to one order of magnitude . This change in parameters influenced the speed of convergence of the synaptic weights to their equilibrium distributions , but not the shape of the distributions itself , while at the same time taking no influence on network dynamics . We studied analytically tractable reduced models that enable us to calculate STDP weight updates in a mean-field fashion and that allow us to give an explanation of the observed processes of self-organization in the network .
|
It is widely believed that the structure of neuronal circuits plays a major role in brain functioning . Although the full synaptic connectivity for larger populations is not yet assessable even by current experimental techniques , available data show that neither synaptic strengths nor the number of synapses per neuron are homogeneously distributed . Several studies have found long-tailed distributions of synaptic weights with many weak and a few exceptionally strong synaptic connections , as well as strongly connected cells and subnetworks that may play a decisive role for data processing in neural circuits . Little is known about how inhomogeneities could arise in the developing brain and we hypothesize that there is a self-organizing principle behind their appearance . In this study we show how structural inhomogeneities can emerge by simple synaptic plasticity mechanisms from an initially homogeneous network . We perform numerical simulations and show analytically how a small imbalance in the initial structure is amplified by the synaptic plasticities and their interplay . Our network can simultaneously explain several experimental observations that were previously not linked .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Self-organization in Balanced State Networks by STDP and Homeostatic Plasticity
|
Leishmaniasis is an antropozoonosis caused by Leishmania parasites that affects around 12 million people in 98 different countries . The disease has different clinical forms , which depend mainly on the parasite genetics and on the immunologic status of the host . The promastigote form of the parasite is transmitted by an infected female phlebotomine sand fly , is internalized by phagocytic cells , mainly macrophages , and converts into amastigotes which replicate inside these cells . Macrophages are important cells of the immune system , capable of efficiently killing intracellular pathogens . However , Leishmania can evade these mechanisms due to expression of virulence factors . Different strains of the same Leishmania species may have different infectivity and metastatic phenotypes in vivo , and we have previously shown that analysis of amastigote proteome can give important information on parasite infectivity . Differential abundance of virulence factors probably accounts for the higher virulence of PH8 strain parasites shown in this work . In order to test this hypothesis , we have quantitatively compared the proteomes of PH8 and LV79 lesion-derived amastigotes using a label-free proteomic approach . In the present work , we have compared lesion development by L . ( L . ) amazonensis PH8 and LV79 strains in mice , showing that they have different virulence in vivo . Viability and numbers of lesion-derived amastigotes were accordingly significantly different . Proteome profiles can discriminate parasites from the two strains and several proteins were differentially expressed . This work shows that PH8 strain is more virulent in mice , and that lesion-derived parasites from this strain are more viable and more infective in vitro . Amastigote proteome comparison identified GP63 as highly expressed in PH8 strain , and Superoxide Dismutase , Tryparedoxin Peroxidase and Heat Shock Protein 70 as more abundant in LV79 strain . The expression profile of all proteins and of the differential ones precisely classified PH8 and LV79 samples , indicating that the two strains have proteins with different abundances and that proteome profiles correlate with their phenotypes .
Leishmaniasis is an antropozoonosis that affects around 12 million people in 98 different countries in Europe , Africa , Asia and America [1] . More than 1 , 5 million new cases are reported every year , 0 , 7 to 1 , 2 of them of the tegumentary forms and 0 , 2 to 0 , 4 million of the visceral form [1] . The clinical form of the disease depends mainly on the Leishmania species and on the immunologic status of the host [2] . In Brazil , Leishmania ( Viannia ) braziliensis and Leishmania ( Leishmania ) amazonensis are the species most frequently involved in tegumentary leishmaniasis [3] . The human L . ( L . ) amazonensis symptomatic infection frequently leads to the localized cutaneous leishmaniasis ( LCL ) , with moderate cellular hypersensitivity , and more rarely to the diffuse cutaneous leishmaniasis ( DCL ) , associated with anergy to parasite’s antigens [3] . The parasite has two main forms: promastigotes , transmitted by an infected female phlebotomine sand fly , and amastigotes , which live and replicate in phagolysosomes of phagocytic cells , mainly macrophages [4 , 5] . Macrophages are important cells of the immune system , capable of directly killing intracellular pathogens and triggering adaptive responses against them [6] . When activated , these cells produce cytokines and reactive oxygen species , nitric oxide , lysosomal enzymes and proteases with microbicidal effects [5] . Leishmania , however , can evade these mechanisms and replicate inside macrophages due to parasite´s virulence factors [7 , 8] . The importance of specific virulence factors may vary according to the Leishmania species . Protein A2 , LACK ( homolog of receptor for activated C kinase ) and cathepsin L-like cysteine protease B ( CPB ) , for instance , are considered important factors for L . ( L . ) donovani , L . ( L . ) major and L . ( L . ) amazonensis , respectively [2] . Inositol phosphosphingolipid phospholipase C-like ( ISCL ) is also considered an essential factor for L . ( L . ) major survival inside the acid phagolysosome [9] . Curiously , while L . ( L . ) major ISCL knock out parasites lost virulence in BALB/c mice , L . ( L . ) amazonensis ko parasites had similar virulence compared to wild type in this mouse strain [10] . Lipophosphoglycan ( LPG ) and major surface glycoprotein GP63 are by far the most studied Leishmania virulence factors . LPG is the most abundant molecule in promastigote´s surface [11] . It inhibits macrophage nitric oxide production , signal transduction and apoptosis , delays phagolysosome maturation and induces RNA double strand-dependent protein kinase ( PKR ) , which increases parasite growth [12–14] . Although essential for L . ( L . ) major and L . ( L . ) donovani infectivity , LPG is not necessary for L . ( L . ) mexicana infection in vitro and in vivo [15 , 16] . The zinc-metalloprotease GP63 is an important antigen in promastigotes , also expressed ( at lower levels ) in amastigotes [17] . GP63 facilitates Leishmania infection and survival since it degrades extracellular matrix , decreases kinase and upregulates phosphatase activity in infected macrophages , and enhances the resistance to antimicrobial peptides . Besides , GP63 cleaves C3 to C3b and C3bi , increasing parasite resistance to complement-mediated lysis , and directly cleaves the pro-inflammatory factors AP-1 and NF-κB ( reviewed in [11 , 17] ) . Interestingly , it was recently shown that cysteine peptidase B , an important virulence factor for L . ( L . ) mexicana and L . ( L . ) amazonensis [18] , regulates the levels of LPG and GP63 in L . ( L . ) mexicana [19] . While some factors are restricted to the parasite surface , others can be secreted . GP63 , elongation factor 1 alpha ( EF-1α ) , frutose-1 , 6-bisphosphate aldolase , secreted acid phosphatase ( SAcP ) , heat shock proteins ( HSPs ) 10 and 70 and tryparedoxin peroxidase , among others , are produced and secreted by amastigotes [8 , 20] . Not only GP63 , as previously mentioned , but also EF-1α , aldolase and SAcP , interact with macrophage kinases and phosphatases , reducing cell activation and microbicidal capacity [8] . Cysteine peptidases may either accumulate inside amastigotes or be secreted in exosomes , depending on the Leishmania species . These important virulence factors have roles both inside the parasite and in the host [11] . It is well known that Leishmania species differ in terms of virulence , as illustrated by the fact that several mouse lineages are resistant to L . ( L . ) major and susceptible to L . ( L . ) amazonensis [2 , 21] . It is also known that strains of the same Leishmania species may show different infectivity and metastatic phenotypes in vivo [22–24] . Although proteome comparison has been extensively employed for the identification of proteins involved in resistance to drugs [25–29] , few studies have used this strategy to identify virulence factors . One of them compared different clones of L . ( V . ) guyanensis and identified two proteins associated with metastatic capacity [22] . Another study analyzed two strains of L . ( L . ) infantum with different infectivity in vivo and found that proteins such as KMP-11 , heat shock proteins , tryparedoxin peroxidase ( CPx ) and peroxidoxin were differentially expressed [23] . A recent work compared L . ( V . ) braziliensis isolates from mucosal and cutaneous lesions of the same patient and observed overexpression of prostaglandin f2-alpha synthase and HSP70 in cutaneous isolates [24] . We have previously shown that LV79 strain of L . ( L . ) amazonensis develop small lesions in C57BL/6 mice . In fact , LV79 lesions in this mouse strain increase until six weeks after inoculation and decrease thereafter , although parasites can still be found in lesions until thirteen weeks post infection [30] . On the other hand , PH8 strain was shown to generate lesions of increasing size in the same mouse strain [31] . In the present work , we show that promastigotes from LV79 and PH8 strains induce different lesion development in BALB/c and C57BL/6 mouse strains , and that amastigotes from PH8 are more infective . Differential abundance of virulence factors probably accounts for the higher virulence of PH8 amastigotes . In order to test this hypothesis , we have quantitatively compared the proteomes of PH8 and LV79 lesion-derived amastigotes using a label-free proteomic approach . The comparison of the proteomes of lesion-derived amastigotes from the two strains identified proteins such as CPx , SOD and HSP70 as significantly more abundant in LV79 amastigotes , and GP63 as more abundant in PH8 parasites . The expression profile of all proteins and of the differentially expressed ones precisely classified PH8 and LV79 samples , indicating that protein abundance profiles correlate with the phenotypes of the two strains .
All animals were used according to the Brazilian College of Animal Experimentation ( CONEP ) guidelines , and the protocols were approved by the Institutional Animal Care and Use Committee ( CEUA ) of the University of São Paulo ( protocol number 001/2009 ) . Euthanasia was performed in CO2 camera . Promastigotes of Leishmania ( L . ) amazonensis LV79 ( MPRO/BR/72/M 1841 ) and PH8 ( IFLA/BR/67/PH8 ) strains were cultured at 24°C in M199 medium supplemented with 10% fetal calf serum ( FCS ) . Parasites were sub-cultured every 7 days to inoculums of 2 × 106/mL . For differentiation of amastigotes into promastigotes , lesion-derived parasites were counted using Neubauer chamber and transferred to M199 medium with 10% FCS at densities of 103 , 104 and 105 parasites/mL . Cells were incubated at 24°C for 4 days and promastigote densities were determined . Four to 8-week-old BALB/c and C57BL/6 mice maintained in our facilities were infected in the left hind footpads with 2 × 106 promastigotes of L . ( L . ) amazonensis LV79 or PH8 in the beginning of stationary-phase ( day 5 , see S1 Fig ) in a final volume of 20μL . Footpad thickness was measured weekly using a caliper ( Mitutoyo Corporation , Japan ) . For histological analysis , we employed five BALB/c animals for each parasite strain . Animals were euthanized , infected paws were removed and control footpads were removed from uninfected mice with similar ages . Fragments of these tissues were fixed in 10% buffered formalin for 18 h , washed and dehydrated in graded concentrations of ethanol , diaphanized and embedded in paraffin . The 4 μm paraffin sections were stained with hematoxylin and eosin . For immunohistochemistry , sections were deparaffinized , blocked with 5% BSA in PBS for 30 minutes , and incubated in 0 , 1% sodium azide , 3% H2O2 in methanol for 30 minutes for blocking endogenous peroxidase . After incubation with rabbit anti-Leishmania serum ( gently provided by Prof . Mauro Cortez ) 1:1000 in PBS 2% ( w/v ) BSA for 18h at 4°C , slides were washed in PBS and incubated with secondary anti-rabbit peroxidase-conjugated antibody ( Imuny , Brazil ) 1:2000 in PBS 2% ( w/v ) BSA for two hours . Slides were then washed in PBS , incubated with DAB ( DAKO , Denmark ) for 2 minutes , washed in water and counterstained with hematoxylin . Samples were dehydrated and diaphanized , mounted with Permout ( Sigma ) and analyzed in Nikon Eclipse E200 LED microscope with Moticam 580 ( Motic ) camera . Amastigotes were purified as previously described [32] . Briefly , lesions were minced and homogenized in 5mL PBS using a tissue grinder ( Thomas Scientific ) . After centrifugation at 50 x g for 10 min at 4°C , the supernatant was recovered and centrifuged at 1450 x g for 17 min at 4°C . Supernatant was then discarded and the pellet was washed three times with PBS followed by centrifugations at 1450 x g for 17 min at 4°C . After 3h of incubation in RPMI with 4% serum under rotation at room temperature to liberate endocytic membranes , amastigotes were further centrifuged , resuspended in 2mL of erythrocyte lysis buffer ( 155mM NH4Cl , 10mM KHCO3 , 1mM EDTA , pH7 , 4 ) and incubated for 2 min in ice . Parasites were washed twice in PBS , resupended at 109 cells/300μL in PBS + Proteoblock ( a protease inhibitor cocktail from Fermentas ) and lysed by 8 cycles of freeze thaw in liquid nitrogen-42°C . Soluble proteins were obtained after centrifugation at 12 . 000 x g for 3 min and quantified by Bradford ( Biorad ) . MTT assay was performed using MTT ( 3-[4 , 5-dimethylthiazol-2-yl]-2 , 5- diphenyltetrazolium bromide ( Sigma ) as previously described [13] . Briefly , 2x107 lesion-derived amastigotes were resuspended in 100 μL PBS with 5mM glucose , transferred to 96 well plates and incubated with 20 μL of MTT ( 5mg/ml in PBS ) at 34°C for 50 minutes . 100 μl of SDS 10% were added and absorbance at 595 nm ( reference at 655nm ) was measured in BioTek ELx800 equipment ( Biotek , Winooski , VT , USA ) . Trypsin-like activity in amastigote extracts was assayed as we recently described [33] . 100 μg of soluble amastigote proteins from each sample were digested with trypsin . The resulting peptide mixture was analyzed on a LTQ Velos Orbitrap mass spectrometer ( Thermo Fisher Scientific ) coupled with LC-MS/MS by an EASY-nLC system ( Thermo Fisher Scientific ) through a nanoelectrospray ion source . Sample concentration and desalting were performed online using a pre-column ( 2 cm; 100 μm ID; 5 μm C18-A1; Thermo ) . Separation was accomplished on Acclaim PepMap 100 C18 column ( 10cm; 75um ID; 3um C18-A2; Thermo ) using a linear gradient of A and B buffers ( buffer A: A = 0 . 1% formic acid; Buffer B = 99% ACN , 0 . 1% formic acid ) from 1% to 50% buffer B over 60 for a total of 77 min at a flow rate of 0 . 3 μL/min to elute peptides into the mass spectrometer . Columns were washed and re-equilibrated between LC—MS/MS experiments . Mass spectra were acquired in the positive-ion mode over the range m/z 400–1500 at a resolution of 30 , 000 ( full width at half-maximum at m/z400 ) and AGC target >1 × e6 . The 20 most intense peptide ions with charge states ≥2 were sequentially isolated to a target value of 5 , 000 and isolation width of 2 and fragmented in the linear ion trap using low-energy CID ( normalized collision energy of 35% ) with activation time of 10 ms . Dynamic exclusion was enabled with an exclusion size list of 500 , exclusion duration of 30 s , and a repeat count of 1 . Three biological replicates ( amastigotes from three independent mice infections ) were performed with two technical runs for LV79 and PH8 . For protein identification and quantification , raw files were imported into MaxQuant version 1 . 5 . 2 . 8 [34] . The database search engine Andromeda [34 , 35] was used to search MS/MS spectra against a database composed of Uniprot Mus musculus ( release May 5th , 2016; 50 , 189 entries ) and Leishmania sp ( release May 5th 2016 , 50 , 820 entries ) databases . Database search employed the following parameters: ( i ) mass tolerance of 4 . 5 ppm and 0 . 5 Da for MS and MS/MS , respectively; ( ii ) trypsin cleavage at both ends and two missed cleavage allowed; ( iii ) carbamidomethylation of cysteine ( 57 . 021 Da ) was set as a fixed modification , and oxidation of methionine ( 15 . 994 Da ) and protein N-terminal acetylation ( 42 . 010 Da ) were selected as variable modifications . All identifications were filtered to achieve a protein and peptide FDR of 1% . One peptide was set as the minimum number for protein identification , and all proteins identified with one peptide had this peptide as unique peptide that could unambiguously identify that protein . For protein quantification , a minimum of two ratio counts were required . All identifications were filtered to achieve a protein and peptide FDR of less than 1% as recommended in the proteomic community for large scale mass spectrometry-based experiments acquired in the data-dependent mode used in this study . Protein quantification was based on the MaxQuant label-free algorithm using both unique and razor peptides for protein quantification , and at least 2 ratio counts were required for considering a protein quantification valid . Protein abundance was calculated based on label-free protein quantification ( LFQ ) values , which are normalized intensities calculated by the MaxQuant software [36] . LFQ-based quantification was shown to provide very accurate and robust quantification and has been validated in many diverse biological contexts [37] . Fold changes were calculated by dividing the average of the LFQ intensities from LV by the average of LFQ intensities from PH replicates . Statistical analyses of the proteome data were performed using Perseus v . 1 . 5 . 4 . 1 in the MaxQuant environment . First , proteins identified in the reverse database , potential contaminants and proteins identified only by site were excluded . The LFQ intensities were log2 transformed and the averages of the two technical replicates values for each independent experiment were calculated . T-test analysis was applied on the PV and PH groups with a p value set to p<0 . 05 . Hierarchical clustering of significantly altered proteins was performed using the Z-score calculation on the log2 intensity values , and the results were represented as a heat map . Principal component analysis was constructed in the web-based chemometrics platform MetaboAnalyst 2 . 0 [38] . Western blots were performed as previously described [32] using 25μg of soluble amastigote proteins and 12% acrylamide gels . After incubation with ECL Prime Western Blotting Detection Reagent ( GE healthcare ) for five minutes , membranes were developed using ChemiDoc XRS+ ( BioRad ) and analyzed using Image Lab ( BioRad ) software . The results were normalized to actin band intensities .
Both LV79 and PH8 L . ( L . ) amazonensis strains cause lesions in BALB/c and C57BL/6 mice , but lesions were smaller and decreased with time in C57BL/6 mice ( Fig 1A ) . On the other hand , BALB/c lesions were significantly larger than C57BL/6 for both parasite strains , as we have already described [30] . PH8 lesions were significantly larger than LV79 in both mouse strains ( Fig 1B , 1C and 1D ) , and parasite loads tend to be higher in infections with this L . ( L . ) amazonensis strain ( Fig 1E ) . We also compared histological sections of PH8 and LV79 lesions in BALB/c mice . After twelve weeks of infection , BALB/c mice showed disrupted footpad structure and high abundance of infected macrophages for both parasite strains , and more abundant necrosis in LV79 lesions ( Fig 2A and 2B ) . Immunohistochemistry indicated higher abundance of parasites ( labeled in brown ) in PH8 lesions ( Fig 2C versus Fig 2D ) , corroborating the higher parasite recovery ( Fig 1E ) and lesion size ( Fig 1D ) observed for this L . ( L . ) amazonensis strain . Infections shown in the previous experiments were initiated with promastigote cultures in stationary phase . To verify whether infections using amastigotes of PH8 and LV79 also generated lesions with significant different sizes , we isolated lesion-derived parasites and inoculated them in naïve BALB/c footpads . Before inoculation , we estimated parasite viability by MTT assay and analyzed trypsin-like activity , used as a measure of metacaspase activity , which is directly associated to parasite death [33] . We also compared parasite differentiation into promastigotes . In Fig 3A we show that lesion-derived amastigotes from PH8 strain have higher viability than LV79 , and , accordingly , lower trypsin-like activity ( Fig 3B ) . As expected , PH8 amastigotes generate cultures with higher numbers of promastigotes ( Fig 3C ) . Lesions generated after inoculation of PH8 amastigotes were bigger than the ones generated by LV79 amastigotes , as shown in Fig 3D , 3E and 3F . To analyze if the larger sizes of PH8 lesions could be attributed to a higher number of viable parasites , we adjusted LV79 parasite numbers considering their viability , so that we would inoculate the same number of viable amastigotes for LV79 and PH8 . As shown in Fig 3D , 3E and 3F , infections with normalized LV79 parasites still led to smaller lesions than PH8 , indicating that the higher virulence of PH8 cannot be solely attributed to the increased viability of lesion amastigotes . In fact , only in infections using 5 or 10 times more LV79 amastigotes we observed a lesion development pattern similar to PH8´s ( S2 Fig ) . Differential abundance of virulence factors probably accounts for the higher virulence of PH8 amastigotes . In order to test this hypothesis , we have quantitatively compared the proteomes of PH8 and LV79 lesion-derived amastigotes using a label-free proteomic approach . Amastigote loads for LV79 strain in C57BL/6 mice lesions 13 weeks after infection are around 7 x 104 parasites/footpad , much lower than the 1 . 5 x 108 parasites/footpad of BALB/c , as we have recently shown [30] . This low parasite recovery precluded the use of C57BL/6-derived amastigotes for proteome analysis . Three independent experiments ( named 1 , 2 and 3 ) were performed with BALB/c mice infected with stationary promastigotes of the two strains , and each amastigote sample was analyzed in technical duplicates . The total number of proteins identified in the Leishmania database , considering all experiments and replicates , was 301 . Fig 4A indicates that 276 of the 301 proteins were detected in the proteomes of both strains , while 15 and 10 proteins were detected only in LV79 and PH8 amastigotes , respectively ( S1 Table ) . Among the proteins identified in both samples , 12 were significantly more abundant in PH8 amastigotes and 25 in LV79 ( Table 1 ) . Among these 37 proteins , 16 had fold changes of at least 2 ( ratios LV79/PH8 higher than 2 or lower than 0 , 5 ) : 11 more abundant in LV79 and 5 more abundant in PH8 , which are now depicted in bold in Table 1 . Although most fold changes were not very high , they are robust since they have statistical significance after t-test of three independent experiments . These results indicate that among the 301 proteins identified , 20% ( 62 proteins ) were either exclusively detected or increased in one of the strains . It is important to mention that among the 301 proteins , 218 ( 72% ) were common across all experiments ( PH8 and LV79 ) and replicates . We also observed that the R2 correlation value of the quantified protein signals between individual replicates was excellent , with a range of 0 . 929–0 . 975 , indicating high reproducibility among replicates . The pattern of expression of the 37 differential ( but not exclusively detected ) proteins precisely clustered PH8 and LV79 samples in two separate branches , as shown in Fig 4B . When we employed expression data of all identified proteins , including the two technical replicates of each sample , PH8 and LV79 samples still clustered ( Fig 4D ) . Samples were also efficiently grouped based on principal component analysis ( Fig 4C ) , indicating that the two strains have remarkable differences in terms of protein abundance . Proteins with different abundance comparing PH8 and LV79 are involved in several cellular processes , among them metabolism/ ATP synthesis , signaling , proliferation/replication , translation , and oxidative stress ( Table 1 ) . These proteins included some known Leishmania virulence factors such as cysteine protease , tryparedoxin and tryparedoxin peroxidase ( CPx ) , superoxide dismutase ( SOD ) , GP63 , heat shock protein 70 ( HSP70 ) and elongation factor . Proteins showing subtle differences are more difficult to validate in “semi-quantitative” Western blot assays , and for this reason we have chosen to validate proteins with ratios higher than 2: tryparedoxin peroxidase , with fold 4 , 62 in LV79/PH8 , and GP63 , with fold 0 , 34 in LV79/ PH8 ( 2 , 94 times more abundant in PH8 ) . Both were analyzed using antibodies developed against Leishmania ( anti-GP63 , anti-CPx ) . The images and corresponding bar graphs shown in Fig 5 validate proteome analysis ( Table 1 ) : GP63 is indeed more abundant in PH8 proteomes , and CPx is more abundant in LV79 proteomes .
We have shown that BALB/c and C57BL/6 mice infected with promastigotes of LV79 and PH8 strains develop lesions with striking different sizes according to the parasite and mice strains . In comparison to BALB/c mice , C57BL/6 lesions were smaller and decreased with time for both parasite strains , different from the huge increasing lesions previously reported for PH8 [31] . This discrepancy may be attributed to C57BL/6 strain maintained in different animal facilities or to the parasite strain from different labs . Anyway , lesions caused by PH8 strains were significantly smaller than the ones induced by LV79 in both BALB/c and C57BL/6 mice . Amastigotes from the two strains were compared in terms of protein abundance , as shown by proteome analysis . The 301 Leishmania proteins identified in this study represent a small fraction of the 6000 proteins predicted to be expressed in amastigotes , and several reasons may explain this fact . First , the study of lesion-derived amastigotes´ proteome presents some technical challenges due to the interference of host proteins , which are carried along amastigote purification and protein extraction steps even using a well stablished protocol such as ours . The presence of host proteins certainly diminishes our capacity of identifying a higher number of parasite proteins . In fact , after protein identification using a database composed of Uniprot Mus musculus and Leishmania sp , a total of 213 and 301 proteins were identified in the mouse and Leishmania databases , respectively , and 815 of the peptides detected belong to mouse proteins and 875 to Leishmania proteins . Moreover , we have analyzed the iBAQ values , which may be used as a measure of protein abundance [40] , and are calculated by dividing the total intensity of a protein by the number of tryptic peptides between 6 and 30 amino acids in length . Comparing the total iBAQ value for Leishmania proteins to the total iBAQ value of mouse proteins , we found that Leishmania proteins accounted for the double of the iBAQ value of mouse ones . Besides , we did not perform sub-cellular fractionation or peptide fractionation prior to the LC-MSMS analyses . Instead , we only considered soluble proteins from a non-detergent-based protein extraction , since our main interest was on soluble amastigote virulence factors that could modulate macrophage infection and parasite survival . This strategy probably leads to a lower number of proteins compared to total extract preparations using detergents [41] or sub-cellular fractionation . At last , biological or chemical post-translational modifications as well single nucleotide polymorphism were not included as variable modifications in the MSMS search , which may represent a fraction of MSMS that was not identified . Proteins considered as virulence factors in Leishmania such as CPx , SOD , GP63 and HSP70 were identified as differentially expressed between the two parasite strains . SOD , CPx and HSP70 are known to reduce oxidative damage in Leishmania . SODs are important in antioxidant defense in many organisms , metabolizing superoxide ( O2- ) into oxygen ( O2 ) and hydrogen peroxide ( H2O2 ) . They are organized in three families based on the metal ion that supports activity: Ni , Cu complexed with Zn , and Mn or Fe [42] . Eukaryotes including mammals have Cu/ Mn/ ZnSODs , whereas FeSODs have been found in prokaryotes , protozoans , plants , and algae [43] . Different FeSOD species ( FeSOD-A and FeSOD-B ) have been characterized in L . ( L . ) chagasi , L . ( L . ) tropica , and L . ( L . ) donovani [44–46] , and in this work we have identified a Fe SOD in L . ( L . ) amazonensis proteome similar to L . ( L . ) mexicana enzyme . CPx has been shown to increase oxidative resistance in L . ( L . ) donovani [47] , L . ( L . ) infantum [48] and L . ( L . ) amazonensis [49] . This enzyme also augments infection [47] and virulence [23] of L . ( L . ) donovani . High levels of the enzyme were reported in antimony resistant L . ( L . ) donovani [48] , L . ( L . ) braziliensis and L . ( L . ) chagasi [29] , in L . ( L . ) amazonensis resistant to arsenite [49] and in metastatic L . ( V . ) guyanensis [50] . HSP70 also protects Leishmania from toxic environmental conditions reducing heat-induced denaturation and cell death [51] . Indeed , HSP70 has been shown to be increased in L . ( L . ) infantum and L . ( L . ) donovani under heat shock or oxidative and nitrosative stresses [23 , 51] , and the overexpression of this protein conferred increased resistance to H2O2 in L . ( L . ) donovani [51] and in L . ( L . ) amazonensis [9] . Like CPx , HSP 70 is overexpressed in antimonial resistant L . ( L . ) infantum and L . ( V . ) braziliensis parasites [29] . Besides , more virulent isolates of L . ( V . ) braziliensis showed increased HSP70 expression [24] . SOD , CPx and HSP70 were all more abundant in LV79 amastigotes . Interestingly , parasites from this strain generated smaller lesions and showed lower viability after isolation from lesions . It is possible that other virulence factors compensate for the lower expression of these three proteins and account for PH8 higher virulence and survival in the host , or that post translation modifications of one or some of these proteins generate more active protein species in PH8 . In fact , we have previously described different species of CPx and HSP70 in L . ( L . ) amazonensis amastigotes [32] , and HSP70 activity is known to be influenced by phosphorylation at specific residues [52] . Among the virulence factors mentioned above , only GP63 had higher abundance in the most virulent PH8 strain . Considering that this molecule favors binding of promastigotes to macrophages and intramacrophage survival and replication [53] , as well as parasite survival in BALB/c mice [54] , it is conceivable that a higher abundance of GP63 may contribute to PH8 virulence . The results presented here show that amastigotes from L . amazonensis strains PH8 and LV79 , which have different virulence in mice , also have proteins with different abundances . To our knowledge , this is the first gel free proteome of lesion-derived amastigotes . Despite the difficulties of working with lesion-derived parasites and the detection of a relatively low proportion of the predicted products , the comparison of PH8 and LV79 strains enabled the reproducible identification of several proteins that distinguish the two strains and that may be involved in virulence in L . amazonensis . In fact , samples from the same strain are efficiently grouped using expression data from all proteins and from the differentially expressed ones . These results indicate that PH8 and LV79 can be distinguished by comparison of protein abundances and that proteome analysis may be used to characterize Leishmania phenotype and eventually predict the virulence of other L . ( L . ) amazonensis strains or isolates .
|
Leishmaniasis is an antropozoonosis caused by Leishmania parasites that affects around 12 million people in 98 different countries . Cutaneous leishmaniasis caused by Leishmania amazonensis can have different clinical forms and severities depending on the parasite strain . We have here shown that two Leishmania amazonensis strains , named PH8 and LV79 , which have different virulence in mice , also have different protein signatures . In fact , samples from these strains can be distinguished based on the abundance of all proteins detected and of the differential ones . Differential proteins identified in this work may be employed in the future to predict virulence of parasite strains or isolates .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"cellular",
"stress",
"responses",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"cell",
"processes",
"immunology",
"microbiology",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoan",
"life",
"cycles",
"developmental",
"biology",
"protozoans",
"leishmania",
"promastigotes",
"heat",
"shock",
"response",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"life",
"cycles",
"amastigotes",
"biochemistry",
"eukaryota",
"cell",
"biology",
"virulence",
"factors",
"proteomes",
"biology",
"and",
"life",
"sciences",
"protozoology",
"cellular",
"types",
"macrophages",
"organisms"
] |
2017
|
Quantitative proteomic analysis of amastigotes from Leishmania (L.) amazonensis LV79 and PH8 strains reveals molecular traits associated with the virulence phenotype
|
Successful host defense against pathogens requires innate immune recognition of the correct pathogen associated molecular patterns ( PAMPs ) by pathogen recognition receptors ( PRRs ) to trigger the appropriate gene program tailored to the pathogen . While many PRR pathways contribute to the innate immune response to specific pathogens , the relative importance of each pathway for the complete transcriptional program elicited has not been examined in detail . Herein , we used RNA-sequencing with wildtype and mutant macrophages to delineate the innate immune pathways contributing to the early transcriptional response to Staphylococcus aureus , a ubiquitous microorganism that can activate a wide variety of PRRs . Unexpectedly , two PRR pathways—the Toll-like receptor ( TLR ) and Stimulator of Interferon Gene ( STING ) pathways—were identified as dominant regulators of approximately 95% of the genes that were potently induced within the first four hours of macrophage infection with live S . aureus . TLR signaling predominantly activated a pro-inflammatory program while STING signaling activated an antiviral/type I interferon response with live but not killed S . aureus . This STING response was largely dependent on the cytosolic DNA sensor cyclic guanosine-adenosine synthase ( cGAS ) . Using a cutaneous infection model , we found that the TLR and STING pathways played opposite roles in host defense to S . aureus . TLR signaling was required for host defense , with its absence reducing interleukin ( IL ) -1β production and neutrophil recruitment , resulting in increased bacterial growth . In contrast , absence of STING signaling had the opposite effect , enhancing the ability to restrict the infection . These results provide novel insights into the complex interplay of innate immune signaling pathways triggered by S . aureus and uncover opposing roles of TLR and STING in cutaneous host defense to S . aureus .
Cells of the innate immune system , including macrophages and neutrophils , are tasked with initiating the rapid and robust response to invading microbial pathogens . Armed with a variety of pattern recognition receptors ( PRRs ) , these sentinel cells sense pathogen associated molecular patterns ( PAMPs ) displayed by microbes to trigger the appropriate antimicrobial response . Specific gene programs activated by microbial pathogens include genes that promote inflammation , genes whose products are directly microbicidal , and genes that induce and regulate adaptive immune responses . However , pathogens may also activate a transcriptional program that interferes with host defense pathways . Understanding this complex interplay of host immune responses to pathogens and evasion of host defense by the pathogen is critical for the discovery of new therapeutics . While much work has been performed identifying how individual PAMPs trigger transcriptional cascades , less work has been performed examining how a complex , living pathogen can trigger transcriptional responses in immune cells and whether these initial transcriptional responses result in protective immunity or favor pathogen persistence . Staphylococcus aureus is a Gram-positive microorganism that is the leading cause of skin and soft tissue infections in humans . Dissemination of S . aureus can result in life-threatening infections , and invasive infections with S . aureus result in more deaths annually than infections with any other infectious agent in the United States[1] . Through its acquisition of resistance to multiple antibiotics , methicillin resistant S . aureus ( MRSA ) has reemerged as a major public health concern . A major goal is to define the regulatory mechanisms responsible for host recognition of S . aureus while identifying pathways triggered by S . aureus to evade host defense for the purpose of developing strategies to improve immune responses and better combat infection . Through studies of mutant mice , S . aureus has been shown to activate a myriad of PAMP/PRR pathways . Toll-like receptors ( TLRs ) are one such family of cell surface and endosomal PRRs that recognize a variety of microbial PAMPs . TLR1/2 or TLR2/6 recognizes cell wall components of S . aureus including Pam3CysK4 and lipotechoic acids [2 , 3] . RNA and DNA from S . aureus have been shown to activate TLR8 and TLR9 , respectively [4 , 5] . Activation of TLRs results in signaling through the adaptor proteins Myeloid differentiation factor 88 ( MyD88 ) and/or Toll-IL-1 receptor domain containing adaptor inducing interferon-β ( TRIF ) to initiate transcriptional responses . Additionally , S . aureus can either escape endosomes to enter the cytosol [6 , 7] or secrete various virulence factors into the cytosol from within the endosome , including cell wall components ( muramyl dipeptide ) , toxins , secreted substances , or DNA; these factors can trigger cytosolic receptors , such as nucleotide oligomerization domain ( NOD ) 2 , the NOD-like receptor ( NLR ) NLRP3 , and Absent in melanoma 2 ( AIM2 ) , respectively . Many of the above pathways converge to activate IL-1β transcription ( via TLRs ) and processing ( via the inflammasome ) , with IL-1β serving as a critical host defense factor required to control S . aureus [8–10] . While the above pathways are involved in host defense , S . aureus can also induce type I interferon ( IFN ) production [11] , which has been associated with evasion of host defense . This type I IFN has been shown to directly inhibit IL-1β transcription and IL-1β processing through the inflammasome in response to multiple infections [12–14] , decreasing host defense in certain infections , but diminishing systemic inflammation in response to others [15] . A variety of pathways have been reported to contribute to the type I IFN response , including the aforementioned TLR2 [16] , TLR8 [5] , TLR9 [4] , and NOD2 pathways [17] , as well as a recently identified pathway involving activation of Stimulator of IFN Genes ( STING ) [18] . While STING activation was shown to be dependent upon DacA-mediated production of the small molecule , di-cyclic AMP in response to extracellular S . aureus biofilms , a cytosolic DNA sensor , cyclic GMP-AMP synthase ( cGAS ) , can also activate STING in response to cytosolic DNA from many other bacterial pathogens [19–22] . TLR pathways can be triggered by ligands present on either living or killed pathogens taken up from the extracellular milieu , whereas the cytosolic PRR families only sense living pathogens that can gain access to the cytosol ( through direct invasion or secretion of factors ) . Although DNA microarray analyses of mRNA from infected macrophages have allowed an examination of individual host defense pathways in the response to pathogens such as Mycobacterium tuberculosis and Listeria monocytogenes [23–25] , a comprehensive analysis of the relative contributions of multiple PRR pathways has not been performed . Furthermore , how these pathways enhance or inhibit early host defense has not been fully elucidated . Herein , we defined the transcriptional programs induced by live and heat-killed ( HK ) S . aureus using RNA Sequencing ( RNA-seq ) of macrophages from wild type ( WT ) and mutant mice . We examined the extent to which each of several pathways contributes to the S . aureus response . Live and HK S . aureus induced similar transcriptional programs , but utilized different sensing mechanisms for the induction of many genes . While TLR signaling pathways were responsible for a high percentage of the response to HK S . aureus , live S . aureus utilized both TLR and STING signaling to induce the transcriptional program . The STING response required live bacteria , was activated by multiple Staphylococcal species , and was activated in mouse and human cells predominantly through cGAS recognition of S . aureus DNA . Interestingly , a small number of TLR- and STING-independent genes activated by live S . aureus were linked to a hypoxia sensing pathway . Using a cutaneous S . aureus infection model , we found that , while activation of TLRs through MyD88 resulted in protective immunity by inducing IL-1β production , neutrophil infiltration , and neutrophil activity in the skin , activation of STING antagonized protective immunity by limiting neutrophil recruitment and IL-1β production . These findings highlight the complex interplay between TLR-dependent host defense and cGAS-STING-dependent immune evasion mechanisms in the early transcriptional responses elicited by S . aureus .
To fully understand the gene programs induced by infection with live S . aureus , we first used RNA-seq to characterize the genome-wide response to the bacterium in WT mouse C57BL/6 bone marrow-derived macrophages ( BMDMs ) . During initial experiments , we found that the rapid proliferation of live S . aureus can quickly overtake BMDM cultures . We therefore examined the gene expression program during the first four hours of culture , prior to bacterial overgrowth . We did not add antibiotics to the culture medium as this would skew responses to those induced by killed bacteria . We also found that cultured BMDMs were exquisitely sensitive to hemolysin-producing S . aureus strains , which induced considerable cell death ( 30–40% ) within two hours . To focus on immune pathways and minimize the effects of apoptosis and cell stress pathways on the transcriptional signature , we used the DU5938 ( α , β , γ hemolysin-deficient ) strain of S . aureus [26] , which induced only minimal cell death ( ~10% ) during the four-hour culture period . We first examined the gene expression profile of mouse BMDMs cultured with live S . aureus ( MOI 10 ) . We focused on highly expressed genes ( at least 1 RPKM at one time point ) that were strongly and significantly induced ( at least 5-fold at one or more time points , at p-adjusted<0 . 05 by DESeq ) for all subsequent analyses . Inclusion of genes induced more weakly would have made the results of loss-of-function experiments more difficult to interpret . 369 genes met our induction criteria , and we focused the remainder of our analysis on these genes . We next separated the 369 genes into four clusters based on the time point of maximum expression ( Fig 1A and 1B ) . In general , genes that reached their maximum expression at later time points were more strongly induced than genes that peaked early in the time course ( Fig 1B ) . Only two genes , Fos and Egr1 , reached peak expression within 30 minutes of treatment with live S . aureus ( Fig 1A ) . These genes are well-known targets of serum response factor ( SRF ) , and their promoters displayed enrichment of binding sites for SRF , Activating Transcription Factor ( ATF ) , and cyclic AMP Response Element Binding ( CREB ) ( Fig 1B and 1C ) . Motifs for additional transcription factors were enriched in this analysis and in those shown below , but we will focus on those that have been studied most extensively . Genes that reached peak expression at 60 minutes also were induced by small magnitudes ( 10-fold average ) . However , unlike the transiently induced genes that reached maximal expression at 30 minutes , these genes generally maintained substantial expression throughout the remainder of the four-hour time course ( Fig 1B ) . These genes also displayed enrichment for ATF and CREB binding sites in their promoters , but lacked significant SRF site enrichment ( Fig 1C ) . The largest groups of genes induced by live S . aureus reached their peak expression at 120 minutes or 240 minutes . The 70 genes induced maximally at 120 minutes displayed a greater range of induction ( 5- to 313-fold ) and , on average , were induced more strongly ( 27-fold ) than those induced maximally at earlier time points . The promoters of these genes possessed significant enrichment for NF-κB family members binding sites , consistent with classic inflammatory genes ( Fig 1B and 1C ) [27] . The average fold induction for the 289 genes that peaked at 240 minutes was considerably higher ( 57-fold average , range 5- to 1617-fold ) . The promoters of these genes also exhibited enrichment for NF-κB family member binding sites , but additionally exhibited enrichment for IFN Regulatory Factor ( IRF ) and Signal Transducer and Activator of Transcription ( STAT ) 1 and STAT2 binding sites; these factors are often associated with a type I IFN response ( Fig 1B ) . Overall , these findings highlight a highly ordered transcriptional response , with different sets of transcription factors coordinating the transient and sustained responses to live S . aureus , similar to previous findings with macrophages treated with a TLR4 agonist [27 , 28] . We next compared the macrophage gene expression responses to live and HK S . aureus . RNA-seq was performed with BMDMs treated with HK S . aureus ( equivalent to MOI = 10 ) and the data sets were analyzed as above . We found 364 genes that met the induction criteria ( >5-fold induction; maximum expression >1 RPKM; p-adjusted<0 . 05 ) following treatment with HK S . aureus , similar to the number induced by live S . aureus ( 369 ) ( Fig 1D and 1E ) . The kinetics of the response to HK S . aureus was similar to , but slightly slower than , the response to live S . aureus ( Fig 1D and 1E ) . Enrichment of transcription factor binding motifs was similar to that observed with live S . aureus , with CREB motifs associated with the promoters of early genes , followed by NF-κB motifs , and finally NF-κB , STAT and IRF motifs enriched in the promoters of genes that peaked at the 240-minute time point ( Fig 1F ) . Fos and Egr1 , the two genes significantly induced at 30 minutes by live S . aureus , were weakly induced ( between 3-4-fold ) at the 30-minute time point , but they did not meet the 5-fold or p-adjusted <0 . 05 criteria and therefore were excluded from the analysis ( S1 Table ) . In total , our gene sets include 432 genes that were induced by either live or HK S . aureus , or both ( Fig 1G ) . Although 68 genes were considered to be unique to the live S . aureus set and 63 genes were considered to be unique to the HK S . aureus set , the vast majority of these genes were actually induced by both the live and HK organism; they simply missed the thresholds required for inclusion in one of the gene sets ( S1A and S1B Fig ) , with most missing the 5-fold induction threshold with the second stimulus . Of the 63 genes appearing uniquely induced by HK S . aureus , 57 were , in fact , induced 3-5-fold by live S . aureus , with only 6 genes being induced 2–3 fold ( S1B Fig ) . Similarly , most of the 68 genes that appeared uniquely induced by live S . aureus were also induced 3–5 fold by HK S . aureus; however , 20 genes did not reach 3-fold induction , and 9 of these genes were not induced 2-fold ( S1B Fig ) . These experiments revealed a highly similar overall progression of the transcriptional responses induced by live and HK S . aureus , with the exception of a very small cluster of genes induced by live but not HK S . aureus . The findings therefore support an initial hypothesis that similar PRR pathways are responsible for gene induction by the live and HK organisms . Our next goal was to provide a mechanistic framework for gene induction by S . aureus . Because S . aureus possesses agonists to various TLRs , we first examined the effect of abolishing all TLR-dependent signaling through an analysis of BMDMs from Myd88-/-Trif-/- mice; all TLR-dependent signaling is thought to be abolished in these mice because all TLRs are thought to use MyD88 and/or TRIF as essential downstream adaptors . RNA-seq was performed in parallel with mRNA from WT and Myd88-/-Trif-/- BMDMs treated with live S . aureus for 0 , 30 , 60 , 120 , and 240 minutes . Our initial analysis revealed that 308 of the 369 genes induced by live S . aureus in WT mice were expressed in the mutant cells at a level that was less than 50% of the WT level; 276 of those genes were expressed at a level that was less than 33% of the WT level ( <33% expression; Fig 2A ) . Motifs for NF-κB family members were most strongly enriched in the promoters of the genes that exhibited the greatest MyD88/TRIF-dependence ( Fig 2B ) . In contrast , motifs for IRF and STAT family members ( and NF-κB to a lesser extent ) were most strongly enriched in the promoters of the genes that exhibited little or no dependence on MyD88/TRIF ( Fig 2B ) . These results confirm a major role for TLR/MyD88/TRIF signaling and NF-κB in the early inflammatory response activated by live S . aureus , with an apparent association between TLR/MyD88/TRIF-independent genes and IRF/STAT factors that may be reflective of a type I IFN response . We next performed RNA-seq with mRNA from WT and Myd88-/-Trif-/- BMDMs macrophages treated with HK S . aureus for 0 , 30 , 60 , 120 , and 240 minutes . Interestingly , in response to HK S . aureus , the induction of a robust TLR-independent program was absent; even the type I IFN program reflected in the enrichment of STAT/IRF motifs was found to be dependent on TLR signaling in response to HK S . aureus ( S2A and S2B Fig ) . In fact , only 6 of the 364 genes induced by HK S . aureus were induced to 50% of WT values , confirming that TLRs are the dominant PRRs that contribute to nearly the entire early transcriptional program in response to HK S . aureus , including the type I IFN signature ( S2A–S2C Fig ) . These data also suggest that while TLR signaling is fully capable of inducing a type I IFN response in macrophages treated with dead S . aureus , macrophages utilize a different mechanism to induce a type I IFN dependent program in response to live S . aureus . Many immune pathways have been reported to contribute to the induction of the type I IFN response by live S . aureus , but several of these pathways would be inactive in Myd88-/-Trif-/- macrophages [4 , 5 , 11 , 16] . Only a few MyD88/TRIF-independent PRR pathways are triggered by living microorganisms and capable of inducing type I IFN , including the NOD2 , STING , and retinoic acid-inducible gene I ( RIG-I ) /mitochondrial antiviral signaling ( MAVS ) pathways . While a NOD2-IRF5 pathway was shown to induce type I IFN in epithelial and dendritic cells in response to a different strain of S . aureus ( 502A ) , the S . aureus strain more likely to cause disease , USA300 , was a poor activator of this pathway [17] . Consistent with these earlier findings , in preliminary qRT-PCR studies , we were unable to detect a role of NOD2 in the response to live S . aureus at select TLR signaling-dependent ( Tnf , Il6 , Cxcl1 ) and -independent ( Rsad2 , Ifit1 , Ifit3 , Mx1 , Mx2 ) genes . Similarly , Mavs-/- macrophages did not reveal diminished expression of the above genes tested in response to live S . aureus , consistent with previous reports demonstrating no role of MAVS in S . aureus-induced type I IFN production [29] . We next tested whether STING is responsible for the type I IFN response in macrophages treated with live S . aureus . In preliminary experiments , BMDM from StingGt/Gt mutant mice [30] demonstrated an impairment in the induction of several type I IFN-induced genes in response to live S . aureus when compared to WT BMDMs . We therefore used RNA-seq to compare the transcriptional programs in these two strains . We found that 65 genes induced by live S . aureus were expressed at a level that was <50% of the WT expression level , and 49 of these genes were expressed at a level that was <33% of the WT expression level , in StingGt/Gt macrophages ( Fig 2C ) . Promoter enrichment analysis revealed strong IRF and STAT signatures in the promoters of genes regulated strongly by STING ( Fig 2D ) . The genes that reached >50% of WT expression level in the absence of STING signaling possessed strong NF-κB and EGR1/EGR2 signatures . Genes that only reached 33–50% of WT expression level in mice devoid of STING signaling displayed modest promoter enrichment for IRF/STAT , NF-κB , and EGR binding sites , again suggesting that both TLR- and STING- dependent signals can regulate the induction of these partially inhibited genes ( Fig 2D ) . We next examined whether STING participates in the early transcriptional response to HK S . aureus . We found that the vast majority of genes that were induced by HK S . aureus in WT macrophages were induced similarly in STINGGt/Gt macrophages ( S1A and S1B Fig ) . Only 14 genes showed <50% of WT expression , and only 1 of those genes displayed <33% of WT expression , in StingGt/Gt macrophages . All of these genes were also inhibited in Myd88-/-Trif-/- macrophages to a similar or greater degree , and they did not display a strong IRF or STAT signature; they instead displayed POU2F1 and RUNX2 signatures and a weaker NFκB1 signature , highlighting a very minor , if any , role for STING in the response to HK S . aureus . Upon further analysis , while the majority of the genes that displayed impaired expression in StingGt/Gt macrophages did not display similar impairment in Myd88-/-/Trif-/- macrophages , several genes displayed partial inhibition in the absence of either pathway . We therefore wished to analyze the datasets simultaneously to obtain an overall framework of the role of TLR and STING dependent signaling to the response to live S . aureus . We identified 4 groups of genes based solely on their need for TLR and/or STING signaling for >50% expression . The majority of live S . aureus-induced genes ( 284/369 ) was diminished to <50% of WT expression in the absence of TLR signaling ( Group I; Fig 3A and 3B ) . The transcription factor binding motifs in the promoters of these genes , not surprisingly , exhibited strong enrichment for NF-κB motifs ( Fig 3C; S1 Table ) . When Gene Ontology ( GO ) Biologic Process analysis [31] was used to evaluate this gene family , the main processes included inflammatory response , response to molecules of bacterial origin or lipopolysaccharide , and regulation of cytokine production ( Fig 3D and S3A Fig ) . These results highlight the strong role of TLR signaling in the induction of the early inflammatory response induced by live S . aureus . The next transcriptional program included a group of 41 genes whose induction was diminished to <50% of WT in the absence of STING signaling ( Group II; Fig 3A and 3B; S1 Table ) . The promoters of these genes were enriched in IRF and STAT motifs , which are associated with a type I IFN response ( Fig 3C ) . GO Biologic Process analysis confirmed that this group exhibited an enrichment of genes associated with a type I IFN gene program; the enriched processes included defense response to virus , type I IFN signaling pathway , response to type I IFN , cellular response to type I IFN , and cellular response to cytokine stimulus ( Fig 3D and S3A Fig ) . The third group includes 24 genes induced by live S . aureus that was diminished to <50% of WT in macrophages devoid of either TLR- and STING-signaling ( Group III; Fig 3A and 3B; S1 Table ) . These genes display strong enrichment for STAT1/2 , IRF1 , IRF2 , STAT1 and NFκB1 promoter motifs ( Fig 3C ) . When GO Biologic Process analysis was applied to this group , enrichment was observed for regulation of T-helper cell differentiation , regulation of CD4-positive alpha-beta T cell differentiation and activation , positive regulation of nitric oxide biosynthetic processes , and tyrosine phosphorylation of STAT3 protein ( Fig 3D and S3A Fig ) . Not surprisingly , a subset of these genes , including Il12b , Il6 , Il27 , and Nos2 , are among the most strongly induced genes by PAMPs in macrophages , and both the TLR and type I IFN programs can result in their induction . These results suggest that Group III genes require strong and/or multiple signals for their maximal induction [27] . The final group of 20 genes was induced >50% of WT in the absence of either TLR- or STING dependent signals ( Group IV; Fig 3A and 3B; S1 Table ) . Interestingly , promoter motif analysis revealed weak enrichment for hypoxia inducible factor ( HIF ) -1α motifs ( Fig 3C ) . When GO Biologic Process analysis was applied , 4-hydroxyproline metabolism , response to oxygen levels , response to decreased oxygen levels , and response to hypoxia were the major biological processes found to be enriched ( Fig 3D and S3A Fig ) , suggesting that treatment with live S . aureus induced a hypoxia response or activated HIF-1 dependent signaling independently of TLR and STING signaling . To gain further insight into whether a hypoxia response was in fact induced by live S . aureus , we tested whether this group of genes had been previously found in hypoxic macrophages . Indeed , previous work suggested that a metabolic/bioenergetic phenomenon occurs upon infection of cells or tissues with live pathogens , including S . aureus , resulting in the PRR-independent induction of a hypoxic response [32] . We compiled a list from 4 manuscripts [33–36] that tested the direct effects of hypoxia on human or mouse myeloid populations in culture . While undoubtedly TLR signaling can trigger a hypoxia response [37] , we focused only on those genes that were induced strongly in the absence of TLR- or STING-dependent signaling , as it appeared that S . aureus was inducing an additional hypoxia response that did not require TLR signaling . When the 20 genes that did not display an appreciable TLR/STING requirement in our analysis were compared against a list of 352 genes derived from 4 datasets of hypoxia-induced genes in monocytes or macrophages , 8 of our 20 genes were found in the hypoxia-induced gene set ( Fig 3E; p = 1 . 32e-08 ) . Next , we examined whether HK S . aureus induced the genes associated with the hypoxia signature . We created a similar framework using Myd88-/-Trif-/- and StingGt/Gt macrophages treated with HK S . aureus . We found only 3 groups of genes based on TLR and/or STING signaling ( S1A and S1B Fig ) . These include a group of 340 genes diminished to <50% of WT in the absence of TLR signaling , 18 genes that were diminished to <50% of WT when either TLR and STING pathways were inactivated , and 6 genes that induced at >50% of WT expression levels in the absence of either TLR or STING signaling ( Groups I , II , and III , respectively , S1A Fig ) . Notably , all 18 genes that were diminished to <50% of WT in the absence of STING were diminished to a similar or greater extent in the absence of TLR signaling . Of note , the promoters of Group I genes were enriched for NF-κB , IRF , and STAT motifs and included many of the genes that were diminished to <50% of WT in the absence of STING signaling or in the absence of either TLR and STING signaling in response to live S . aureus ( S1C Fig , S1 and S2 Tables ) . The Group II genes in response to HK S . aureus exhibited promoter enrichment for IRF and STAT signatures , while the Group III genes did not have significant enrichment for any transcription factor binding motifs , as it is often difficult to obtain motif enrichment with only 6 genes ( S1 Fig ) . Interestingly , similar to the effects of live versus HK group A Streptococcus on cultured macrophages [21] , HK S . aureus was a more potent inducer of many of the genes that exhibited STING-dependence in response to live S . aureus , including Ifnb , Cxcl9 , Cxcl10 , Il12b , Il27 , Ifit1 , Oasl1 , and ligp1; this finding suggests that HK S . aureus activates the type I IFN response through TLRs more strongly than live S . aureus does through STING ( S1 and S2 Tables ) . To evaluate whether the hypoxia program required the presence of live S . aureus , we further examined the 68 genes that were strongly induced by live but not by HK S . aureus ( see Fig 1G ) . As mentioned above , several of these genes were induced by HK S . aureus , but did not reach the thresholds needed for inclusion in our gene sets . Surprisingly , 54 of these 68 genes were diminished to <50% of WT in the absence of TLR signaling ( Fig 4A , Group I ) , with only 2 genes each exhibiting the same degree of impairment in the absence of either STING signaling alone or either TLR or STING signaling ( Fig 4A , Groups II and III ) . The remaining 10 genes were induced when either TLR or STING pathways were not functional ( Fig 4A , Group IV ) . These genes included 10 of the 20 genes found to be induced in the absence of TLR- or STING-signaling in response to live S . aureus ( Group IV from Fig 3 ) . When GO Biologic Process analysis was applied to the 10 genes that are induced by live but not HK S . aureus ( Fig 4A Group IV genes ) , response to oxygen levels , response to decreased oxygen levels , response to hypoxia , and 4-hydroxyproline metabolism were found to be the most strongly enriched GO signatures ( Fig 4B and S3B Fig ) and included 6 of the 8 genes that overlapped with published datasets of hypoxia stimulated macrophages ( Fig 4C ) . These results therefore provide strong confirmation that the hypoxia response is induced specifically by live S . aureus , and that this response accounts for most of the TLR/STING-independent genes . Notably , the 10 remaining genes from the group of 20 that were induced by live S . aureus in a TLR/STING-independent manner ( Fig 3 , Group IV ) exhibited enrichment for second messenger-mediated signaling and cardiac contractility when examined by GO Biologic Process analysis , but lacked enrichment for hypoxia ( Fig 4B ) . Because live S . aureus induces a strong , STING-dependent , type I IFN response , we next wished to confirm that IFN-β secretion was dependent upon STING . Indeed , treatment of StingGt/Gt macrophages with live S . aureus resulted in a 75% decrease in IFN-β production when compared to WT macrophages ( Fig 5A ) . We next examined how the STING adaptor is activated by live S . aureus . STING can be activated either by di-cyclic nucleotides produced by bacteria or indirectly by the cytosolic DNA sensor cyclic GMP-AMP synthase ( cGas ) , which produces the small molecule cGAMP to activate STING [19 , 38 , 39] . S . aureus can produce a similar di-cyclic nucleotide , di-cyclic AMP , through the dacA protein , which encodes a di-adenylate cyclase; this was recently shown to induce type I IFN in response to S . aureus biofilms [18 , 40] . S . aureus also possesses an Esx secretion system [41] , similar to those in L . monocytogenes and M . tuberculosis that are responsible for activation of cGAS [19 , 20] . To clarify the mechanism by which S . aureus activates STING , we compared induction of Ifnb1 mRNA and IFN-β1 protein by live S . aureus in StingGt/Gt and cGas-/- macrophages . Because all of the experiments to this point utilized the DU5938 strain of S . aureus that lacks hemolysins , we also wished to examine whether the cGAS-STING pathway activates Ifnb1 expression in response to strains of S . aureus that are more clinically relevant . We chose a USA300 LAC strain of methicillin-resistant S . aureus ( MRSA ) , one of the major causes of community and hospital acquired S . aureus infections in the United States , as well as S . epidermidis , a commensal staphylococcal species . The DU5938 strain of S . aureus induced the dose-dependent activation of Ifnb1 mRNA through cGAS-STING ( Fig 5B ) , and induced IFN-β protein production through cGAS-STING as well , with a greater dependence on STING ( Fig 5C ) . Infection of macrophages with USA300 strain resulted in greater induction of Ifnb1 mRNA and IFN-β1 protein from mouse macrophages when compared to the DU5938 hemolysin-deleted strain , and this induction was similarly dependent on cGAS and STING ( Fig 5B and 5C ) . The S . epidermidis strain led to weaker but significant induction of Ifnb1 mRNA and IFN-β protein , with strong STING and cGAS dependence ( Fig 5C ) . Additionally , highly purified DNA from an overnight culture of the USA300 strain of S . aureus induced Ifnb1 mRNA expression in a dose dependent manner in WT , but not StingGt/Gt or cGas-/- macrophages ( Fig 5D ) . These findings indicate that the cytosolic DNA sensor cGAS is the predominant sensor resulting in STING-dependent activation of type I IFN in murine macrophages in the early response to S . aureus , although residual activation through STING appears to occur either through another DNA sensor or through direct activation by di-cyclic nucleotides produced by S . aureus , as observed with Streptococcal species [22] . We next wished to determine whether the cGAS-STING pathway contributes to induction of the type I IFN response in human cells . To accomplish this , we used human THP-1 cells that had CRISPR/Cas9-mediated deletions of the cGAS or STING genes [22 , 42] . Similar to mouse cells , induction of Ifnb1 mRNA was significantly diminished in both cGas-/- and StingGt/Gt cells , with a slightly stronger inhibition in the STING mutant cells ( Fig 5E ) , confirming that human monocytic cells utilize cGAS-STING to induce type I IFN . Since signaling through the TLR and STING pathways contributes to the majority of transcriptional response to live S . aureus in cultured macrophages , we next wished to determine the roles of these pathways in cutaneous infection . MyD88 signaling was previously shown to be critical for cutaneous host defense to S . aureus [1 , 2] , but the role of STING in the cutaneous response to S . aureus is unclear . To test the role of these PRR pathways , we utilized a model of cutaneous infection using a bioluminescent S . aureus strain ( Xen36 ) to track the infection over time , as bioluminescence directly correlates with viability and bacterial numbers [43] and a similar bioluminescent S . aureus was used to show the importance of MyD88 in cutaneous infection with S . aureus [3] . We first confirmed that Xen36 induced type I IFN responses through cGAS-STING in vitro , and found that Xen36 similarly activates early Ifnb1 mRNA through cGAS and STING ( Fig 5F ) . We next compared cutaneous infection of Myd88-/- and StingGt/Gt mice to WT mice . We chose to use StingGt/Gt mice instead of cGas-/- mice because cGAS-/- mice exhibited residual Ifnb1 induction . In our model , following injection of 1x106 bacteria in the skin of WT mice , all mice developed an erythematous , indurated plaque with scale on their back by the second day post-infection . About 50% of wild type mice developed an ulceration by the 4th day post-infection , likely depending on the depth of injection of the Xen36 inoculum ( dermal or sub-dermal ) . The bioluminescence reached its maximal intensity on Day 3 following infection regardless of whether the mice developed an ulceration , and the signal then slowly lost intensity until approximately Day 10–12 , when the signal was lost . First , we confirmed that TLRs participate in cutaneous host defense to S . aureus . Unlike the WT mice , all Myd88-/- mice injected with the same inoculum of S . aureus , developed progressively enlarging ulcerations that started four days after infection . MyD88-/- mice displayed a greater bioluminescence signal than WT mice starting at 3 days post infection , and this signal remained elevated until mice were euthanized at 12 days post infection ( Fig 6A and 6B ) . After confirming that Myd88-/- mice displayed impaired clearance of S . aureus , we performed similar experiments in StingGt/Gt mice . We found that StingGt/Gt mice displayed a more rapid clearing of bioluminescent S . aureus in comparison to WT mice ( Fig 6A and 6B ) . While lesions in about half of StingGt/Gt mice also developed an ulceration ( similar to WT mice ) , the StingGt/Gt mice that ulcerated typically ulcerated earlier than WT mice , with the majority developing ulceration by two days post infection . This finding suggests that STING signaling in vivo impairs cutaneous host defense against S . aureus . Whereas many WT mice still exhibited a bioluminescent S . aureus signal in their wound at 7 and 10 days post-infection , the StingGt/Gt mice displayed significantly improved clearance at Day 3 and Day 5 , and complete clearance of bioluminescent signal at Day 7 . Standard bacteriological plating of infected skin from skin harvested on Day 3 following S . aureus infection confirmed that StingGt/Gt mice possessed decreased bacterial burdens when compared to WT mice ( Fig 6C ) . To determine whether STING signaling contributes to early IFN-β production in vivo , we measured IFN-β protein in the skin 8 hours after infection . While WT mice were capable of inducing IFN-β in the skin , StingGt/Gt mice had a significantly reduced ability to produce IFN-β ( Fig 6D ) . Myd88-/- mice did not display a significant difference in early IFN-β production in the skin in comparison to WT mice . These data suggest that STING activation contributes to IFN-β production and diminishes the ability to clear S . aureus . To further understand the mechanism by which STING activation influences cutaneous host defense to S . aureus , we first examined tissue histologically at 18 hours ( prior to any ulceration ) following infection . Although the infiltration of polymorphonuclear cells was present diffusely in the dermis and panniculus in the skin of WT mice , the presence of a well-defined abscess was only rarely seen ( 1/8 ) ( Fig 7A ) . However , in StingGt/Gt mice , the majority of the samples already possessed a well-defined abscess ( 5/8 ) with greater numbers of polymorphonuclear cells in the well-organized collection . Myd88-/- mice displayed small numbers of neutrophils in the panniculus , but no mice ( 0/5 ) displayed a well-defined abscess . Immunohistochemical staining for Ly6G confirmed the presence of neutrophils in the WT skin , with lower numbers in Myd88-/- skin , as well as the well-formed collection of neutrophils in the StingGt/Gt skin ( Fig 7A ) . To quantify neutrophil infiltration into tissue , we measured myeloperoxidase , a marker of neutrophil activity following S . aureus infection . Myeloperoxidase was increased in skin of StingGt/Gt mice compared to WT mice at 8 and 18 hours following infection ( Fig 7B ) . On the other hand , there was a large impairment in the ability of Myd88-/- mice to recruit neutrophils to the skin , as they demonstrated an approximately 70% reduction in myeloperoxidase levels 8 hours after infection ( Fig 7B ) . These data demonstrate that S . aureus activation of STING impairs while MyD88 activation is required for neutrophil recruitment in the skin . IL-1β is known to be critical for neutrophil recruitment in the skin in response to S . aureus , and type I IFN can inhibit inflammasome-mediated induction of IL-1β [1 , 12] . Furthermore , prior studies of group A Streptococcus revealed a role for type I IFN in regulating IL-1β expression from the skin [15] . We therefore investigated whether S . aureus activation of STING impairs local IL-1β production in the skin . Immunohistochemical analysis revealed that IL-1β was produced in the skin of WT mice infected with S . aureus at 18 hours; the majority of the IL-1β was produced by recruited neutrophils ( Fig 7A ) . There was stronger immunohistochemical staining in neutrophils of StingGt/Gt mice and very low levels of IL-1β staining in neutrophils of Myd88-/- mice ( Fig 7A ) . We next measured IL-1β in skin homogenates to quantify whether IL-1β levels are higher in StingGt/Gt mice . Indeed , we found that IL-1β was increased at 8 and 18 hours following S . aureus infection in WT mice , and StingGt/Gt mice demonstrated a significant increase in IL-1β expression ( Fig 7C ) . Myd88-/- mice demonstrate a 60% reduction in IL-1β at 8 hours post-infection in comparison to WT mice , confirming the essential role of MyD88 for IL-1β induction ( Fig 7C ) . Taken together , our findings demonstrate that although TLR-dependent activation of MyD88 signaling induces protective immunity to limit S . aureus infection , S . aureus utilizes STING signaling to subvert cutaneous host defense .
Studies of the innate immune response to infection by S . aureus and other bacterial pathogens have typically identified several PRR pathways capable of sensing the pathogen and contributing to host defense . Although multiple PRR pathways can play a role in host defense to a given pathogen , the relative contribution of each for the gene expression response elicited by infection generally has not been characterized . An understanding of relative contributions requires a quantitative analysis of the response in WT cells , as well as the response that remains after each PRR pathway is eliminated . This general approach has limitations , in that multiple pathways may each make small contributions to the activation of some genes; other genes may be activated by two or more pathways in a redundant fashion . However , the approach allows the identification of dominant PRR regulators of many genes and represents an essential first step toward a full understanding of the host-pathogen interactions that regulate the immune response , as well as evasion of the response . In this study , we carried out an initial , quantitative analysis of the gene expression response in mouse BMDMs to infection by live S . aureus , and a comparison of the BMDM response to the live and HK pathogen . We focused on the most strongly induced genes because of the difficulty interpreting loss-of-function results when examining weakly induced genes . For example , if a gene is induced by only 2 . 5-fold by the WT pathogen , it is difficult to interpret the importance of a PRR pathway when elimination of that pathway reduces induction to 1 . 7-fold or 2 . 1-fold . This issue remains a challenge even when analyzing strongly induced genes . However , by focusing on genes that are induced by at least 5-fold , it is possible to have greater confidence in the contribution of a PRR pathway when a strong quantitative effect is observed in a loss-of-function experiment . For these loss-of-function experiments , we arbitrarily chose to define a gene as “dependent” on a particular signaling pathway if its expression level in the mutant cells was reduced to <50% of the expression level observed in WT cells . We chose to emphasize these relatively strong effects to increase confidence in the results and also to identify the dominant regulators of each gene . The pathways studied clearly make smaller contributions to the induction of additional genes , and some of those contributions reach statistical significance . Furthermore , we would not be surprised to find modest contributions of pathways that were not examined in this study . Nevertheless , our finding that the vast majority of the strongly induced genes exhibit expression levels that are diminished to <50% of WT in either Myd88-/-Trif-/- or StingGt/Gt BMDMs highlights the dominant roles of these two pathways in the early transcriptional response to S . aureus . One notable feature of our study is that we used a strain of S . aureus deficient in the pore-forming toxins ( α/β/γ hemolysin- and Panton-Valentine leucocidin toxin ) . The parental strain ( 8325–4 ) from which DU5938 was derived also lacks phenol soluble modulin α 3 ( PSM-α3 ) , another virulence factor that can cause cytotoxicity in macrophages , and the DU5938 strain is , therefore , likely to also lack PSM-α3 [7] . This strain was chosen because macrophages in a cell culture system do not have the benefit of complement , antimicrobial peptides , or natural antibodies that can effectively neutralize S . aureus in vivo , and the use of toxin-sufficient strains rapidly killed macrophages , with significant macrophage death starting as early as 2 hours . In response to the WT strain , this time point corresponded to an acute burst of transcriptional activity , but by 4 hours , many ( 8 of 10 ) genes tested had decreased to within 25–50% of basal activity , likely because considerable cell stress or death was occurring . Thus , we wished to evaluate the mechanisms responsible for early gene induction by the live and HK pathogen in the absence of the effects of apoptosis/necroptosis/necrosis that these pore-forming toxins can cause , which include the non-specific release of danger associated molecular patterns ( DAMPs ) from the dying cells . We acknowledge that the pore-forming toxins can contribute to gene expression by activating inflammasome pathways . However , the inflammasome generally requires more than 4 hours to be activated , as activation of caspase-1 , transcription and translation of pro-IL-1β , and proteolytic cleavage to mature IL-1β , are all required before the inflammasome can activate further transcription through IL-1R and MyD88 . Although many innate immune sensors are undoubtedly activated by S . aureus , our results revealed that two pathways , the TLR and DNA-cGAS-STING pathways , represent dominant regulators of early transcription of the majority of genes that are strongly induced by live S . aureus in cultured mouse BMDMs; a hypoxia response appears to account for the induction of many of the small number of early genes induced in a TLR/STING-independent manner . Although the gene expression responses to live and HK S . aureus are very similar at first glance , many of the genes that exhibit STING-dependence in response to the live pathogen are activated by TLR pathways in response to the HK pathogen . These results are consistent with the fact that HK S . aureus remains in the phagolysosome , while live S . aureus is capable of either escaping the phagolysosome and entering the cytosol or secreting substances into the cytosol that can activate cytosolic PRRs , including cGAS , STING , or NOD receptors [6 , 7] Furthermore , the hypoxia response elicited by the live pathogen was not observed with the HK pathogen . An analysis of mutant mouse strains demonstrated that the STING pathway serves as a negative regulator of host defense to S . aureus infection , with the STING pathway promoting a type I IFN response and negatively regulating IL-1β expression and neutrophil recruitment . Our identification of a TLR/STING-independent hypoxia program in macrophages exposed to live bacteria was also of special interest . Hypoxia has been found to serve as a metabolic switch to “train” macrophages to become more antimicrobial; it has been detected in response to S . aureus infections in vitro and in vivo , and may be related to a decrease in oxygen tension in infected tissue/cells [32 , 44] . While TLR signaling alone can induce a hypoxia program , we found an additional hypoxia signature in the absence of TLR signaling . Recently , oxygen consumption and glycolysis by S . aureus was shown to contribute to a hypoxia response by activating HIF-1α [45] . This HIF-1α activation , in turn , augmented IL-1β production and promoted keratinocyte and macrophage activation , while blocking glycolysis worsened cutaneous infections . Acute and chronic hypoxia can alter host defense responses to S . aureus and triggering the hypoxia machinery has been suggested as a potential therapeutic target [46] . The cytosolic DNA pathway can be activated by multiple viruses , bacteria , and mycobacterial species upon access of the pathogen or its DNA to the cytosol [20 , 47] . Streptococcus pneumoniae and group B streptococcus , both of which are Gram-positive bacteria similar to S . aureus , were recently shown to activate type I IFN responses in macrophages through this same pathway[22 , 48] . Other PRRs , including TLR2 , TLR7 , TLR9 , and NOD2 , were shown to be involved in type I IFN induction in response to S . aureus in other cell types using different treatment conditions and lengths of time in culture [4 , 5 , 16 , 17] . Importantly , we focused on early time points to less ambiguously determine the pathways involved in macrophage activation and found that the activation of STING requires the presence of living bacteria , and can be induced by both less virulent S . epidermidis and S . aureus species along with more virulent USA300 strains of MRSA . S . aureus used the cytosolic DNA sensor cGAS to activate STING , but it can also activate STING in a cGAS-independent manner , demonstrating similar redundant activation of the pathway to group B streptococcus [22] . The fact that the USA300 strain resulted in higher IFN-β induction through cGAS-STING is likely due to the increased cell death associated with this strain , as release of host DNA from dying cells likely contributes to the increased IFN-β production , since basal and stimulated type I IFN is regulated by cell death [49] . This is corroborated by the fact that S . epidermidis , which did not cause considerable cell death , also induced lower levels of IFN-β than either DU5938 or USA300 . The role of type I IFN in response to bacterial pathogens , especially extracellular bacteria , is complex [50 , 51] . An inability to induce type I IFN was shown to be detrimental in S . aureus skin infection and adding IFN-β resulted in improved clearance of S . aureus [29]; however , in a pneumonia model , type I IFN was shown to be detrimental [4 , 11 , 17] . Type I IFN can decrease neutrophil chemotaxis and limit neutrophil infiltration , and it leads to increased susceptibility to S . aureus and Streptococcus pneumonia following influenza infection [52 , 53] . A strain of S . aureus that elicits high levels of type I IFN induction also displays increased virulence in a pneumonia model [11 , 47 , 53] . Molecularly , type I IFNs can directly inhibit the transcription of IL-1β as well as the processing of pro IL-1β protein into mature IL-1β through the inflammasome [12–14] . This inhibition of IL-1β processing by type I IFN has been shown to limit host defense to M . tuberculosis , Candida albicans , and Group A Streptococcus , and is likely the mechanism by which STING-dependent type I IFN is limiting IL-1β production in response to S . aureus . Understanding how S . aureus triggers immune pathways and the relative contributions of known and/or novel PRR pathways to S . aureus-initiated transcriptional cascades may lead to novel strategies to combat infection . We are the first to identify the activation of cGAS by S . aureus DNA through the STING adaptor as a major contributor to the initial macrophage type I IFN response to live S . aureus . For our studies , we implemented a model of cutaneous infection by S . aureus to determine the role of the STING pathway in local host defense . We wished to avoid the systemic inflammatory response that a large inoculum of bacteria would cause , since systemic inflammation and bacteremia can compromise local host defense and make interpretation of results more difficult . Using our model , we found that S . aureus activates STING in the skin , which results in the elaboration of IFN-β . Activation of this STING-dependent IFN-β production results in impaired local clearance of S . aureus . Since IL-1β is critical for neutrophil recruitment to S . aureus infection sites [3 , 54] , the loss of early type I IFN in STINGGt/Gt mice likely accounts for the increase in IL-1β and enhanced neutrophil recruitment at the site of infection . The two major early pathways activated by S . aureus within hours of infection in macrophages led to diametrically opposed outcomes following in vivo infection with S . aureus . While both pathways synergized to induce the transcription of some genes in macrophages , the two pathways had opposite effects on host defense in vivo . This likely signifies that the effects of STING signaling on downstream transcriptional cascades are uncoupled from the effects of IFN-β produced in vivo . During cutaneous infection , the TLR-dependent pathway is critical for early local host defense while activation of the cGAS-STING pathway results in immune evasion by S . aureus . Recently , Castiglia et al . showed that during an overwhelming cutaneous S . pyogenes infection that resulted in bacterial dissemination , activation of type I IFN systemically suppressed IL-1β driven inflammation , limiting organ damage [15] . These results were consistent with previous findings that IFN-β produced during influenza infection can diminish IL-1β and host immunity to S . aureus [53 , 55] . Furthermore , activation of the cytosolic DNA pathway through STING in response to the intracellular pathogens , Listeria monocytogenes and M . tuberculosis , was shown to similarly limit host defense to infection [39 , 56] . Taken together , these studies suggest that TLR- and inflammasome-dependent IL-1β production is required for host defense to bacterial infection , but may become detrimental in the case of overwhelming infections . Thus , type I IFN , including that produced through the cGAS-STING pathway , can inhibit IL-1β processing , providing a rheostat on IL-1β production . During local infections , this type I IFN can impair host defense , whereas in severe infections , this type I IFN can diminish systemic inflammation caused by overproduction of IL-1β . Hence , targeting cGAS and/or STING may remove detrimental type I IFN without affecting protective innate immune host defense pathways . Another interpretation is that part of the gene program that requires STING and type I IFN includes genes that are critical in triggering T cell activation and adaptive immunity ( Fig 3D ) . Perhaps the host is activating the STING pathway , a pathway that may lead to diminished initial host defense , in order to trigger activation of adaptive immunity to confer longer protection against a rechallenge .
The mouse studies described in this manuscript were performed under the written approval of the UCLA Animal Research Committee ( ARC ) in accordance to all federal , state , and local guidelines . All studies were carried out under strict accordance to the guidelines in The Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the accreditation and guidelines of the Association for the Assessment and Accreditation of Laboratory Animal Care ( AALAC ) . The protocol/permit/project license number assigned by the IACUC/ethics committee that approved under UCLA ARC Protocol Number 2015-021-01 . S . aureus infections were performed under isofluorane anesthesia and all efforts were made to minimize animal pain and discomfort . All mice used have a C57BL/6J genetic background , Myd88-/- mice , StingGt/Gt mice , and cGas-/- mice were purchased from Jackson Laboratory ( Bar Harbor , ME ) . Myd88/Trif-/- mice were a kind gift of Greg Barton[57] . Human wild type THP-1 cells or those deleted of cGAS and STING were a kind gift of V . Hornung and were cultured in RPMI-1640 ( Gibco Laboratories; Gaithersburg , MD ) with 10% FBS ( Omega Scientific; Tarzana , CA ) and 0 . 05 mM 2-mercaptoethanol ( Sigma-Aldrich; St . Louis ) [42] . Cells were grown to ~80% confluence and treated with 10 ng/ml phorbol myristate acetate ( Sigma ) for 2 hours prior to stimulation . BMDMs were isolated and differentiated from WT or mutant/deficient mice as previously described in [28] . Briefly , bone marrow isolated from femurs and tibias from male mice at 6–10 weeks of age were cultured in BMDM media consisting of DMEM ( Gibco ) with 20% FBS ( Omega ) , CMG conditioned media containing M-CSF ( cell line was a kind gift from G . Cheng ) [58] , and 1X Pen/Strep ( Gibco ) for 6 days . All media was washed twice with PBS ( Corning ) , then antibiotic-free BMDM media was added prior to addition of bacteria . CMG was prepared from L929 cells . Strain DU5938 , a mutant of the WT 8325–4 lab strain triple defective mutant in α , β , and γ hemolysins , was a kind gift of T . Foster [26] . In other experiments , the USA300 LAC strain ( JE2; ATCC/BEI Resources; Manassas , VA ) was used . For S . epidermidis , the Winslow and Winslow strain ( ATCC ) was used . The Xen36 strain ( bioluminescent strain of 8325–4 lab strain ) was purchased from Perkin Elmer ( Walthum , MA ) . Staphylococcal bacteria were inoculated from single colonies and grown shaking at 200 rpm overnight in LB ( Fisher Scientific ) at 37°C . The following morning , a 1:40 dilution of the overnight culture was shaken in LB for 3 hours , washed twice with PBS ( Corning Life Sciences; Manassas , VA ) , and resuspended . Bacterial concentrations were estimated with a spectrophotometer ( Beckman DU 640B Spectrophotometer , Beckman Coulter , Inc . , Fullerton , CA ) by determining the absorbance at 600 nm ( A600 ) . Colony-forming units ( CFUs ) were verified by plating dilutions of the inoculum onto LB agar overnight . HK S . aureus was obtained by placing live S . aureus culture in a 65°C water bath for 1 hour[26] . To obtain highly purified DNA from S . aureus , DNA was extracted from an overnight culture using the DNAZol Reagent ( ThermoFisher; Canoga Park , CA ) per manufacturer recommendations . Following extraction , DNA was treated with 1μg/ml of RNase A . Then the DNA was purified by phenol chloroform extraction . The A260/A280 ratio of the purified DNA that was used in the experiments herein was 1 . 81 . Mouse macrophages were stimulated on day 6 following differentiation . Human and mouse macrophages were infected with bacterial strains at 10 M . O . I . in antibiotic free MDM and BMDM media . Macrophages were infected with S . aureus for 0 , 30 , 60 , 120 , and 240 minutes at 37°C or treated with HK S . aureus for the same time points . Media was obtained at the 240-minute time point for cytokine determination by ELISA . RNA was isolated and reverse transcribed as described[58] . Briefly , RNA was extracted using TRI-reagent ( Molecular Research Center; Cincinnati , OH ) , treated with RNase-free DNaseI , and purified using an RNeasy kit ( Qiagen Inc; Valencia , CA ) . Quantified RNA ( 2 μg ) was reverse-transcribed using Omniscript RT Kit ( Qiagen ) and random hexamer primers . cDNA fragments were analyzed by qPCR using SensiMix Plus ( Bioline; Taunton , MA ) and the iCycler System ( Bio-Rad Laboratories; Irvine , CA ) or a 7900HT ( Applied Biosystems ) . PCR amplification conditions were 95°C ( 3 min ) and 45 cycles of 95°C ( 15 sec ) , 60°C ( 30 sec ) , and 72°C ( 30 sec ) . Primer pairs for mouse and human IFN-β , respectively , are as described [59 , 60] . Results for IFN-β were compared to β-actin or h36b4 ( Thermo-Fisher ) , which were used as housekeeping internal controls for mouse or human , respectively . RNA was isolated as described[27] . Strand-specific libraries were generated from 400 ng total RNA using the TruSeq RNA Sample Preparation Kit v2 ( Illumina ) , with modifications using the “dUTP” method [61] . cDNA libraries were single-end sequenced with a length of 50bp on an Illumina HiSeq 2000 . All bioinformatics analyses were conducted using Galaxy[62] . Reads were aligned to the mouse genome ( NCBI37/mm9 ) with TopHat v1 . 3 . 3 and allowed a maximum of one alignment with up to two mismatches per read . mRNA RPKM values were calculated using Seqmonk ( http://www . bioinformatics . babraham . ac . uk/projects/seqmonk/ ) . RPKMs were calculated by dividing mapped exonic reads by the length of the mature mRNA product . All RPKMs represent an average from at least two biological replicates . A gene was included in the analysis if it met all of the following criteria: The maximum RPKM reached 1 at any time point , the gene was induced at least 5-fold , and the induced expression was significantly different from the basal ( P<0 . 05 ) as determined by the DESeq package in R Bioconductor[63] . P-values were adjusted using the Benjamini-Hochberg procedure for multiple hypothesis testing[64] . The splice variant with the largest RPKM was included in the analysis . To normalize the data , the basal RPKM in WT samples was set at 0% and the maximum WT RPKM at 100% for each gene . In the mutant strains , percent expression was calculated using this scale substituting only the maximum WT RPKM with the maximum mutant RPKM . For promoter enrichment , PScan software ( http://159 . 149 . 160 . 88/pscan ) using Jaspar 2016 motif analysis examining -450 to +50 base pairs of the promoter of groups of interest . For Gene Ontogeny analysis , gene families were inputted into ENRICHR software[31 , 65] . The GO Biologic Process function was used for gene ontogeny classification . The top 5 significant GO terms were used . Both the lighter color and longer length of the bars denote stronger significance of the identified pathway . All procedures were approved by UCLA Animal Research Committee . The mice were shaved on the back and inoculated subcutaneously with 100 μl of mid logarithmic growth phase S . aureus strain Xen36 ( ∼1 × 106 CFUs/100 μl = 1:10 dilution of A600 of 0 . 5/ml ) in sterile pharmacy grade saline ( 0 . 9% ) by a 27-gauge needle and a tuberculin syringe ( Abbott Laboratories; Chicago , IL ) . In one experiment , a subset of WT and StingGt/Gt mice ( n = 4–5 mice ) was used to confirm bioluminescence results by determining CFU from skin on Day 3 . In vivo bioluminescence was performed with the Xenogen IVIS imaging system ( Xenogen Corporation; Alameda , CA ) at the Crump Institute for Molecular Imaging at UCLA as previously described[66] . Mice were anesthetized via isofluorane injection . Data are presented on color scale overlaid on a gray-scale photograph of mice and quantified as total flux and average radiance ( photons/s ) within a circular region of interest ( 1 × 103 pixels ) with Living Image software ( Xenogen ) ( lower limit of detection: 1 × 104 photons/s ) . For histological analysis , lesional 8 mm punch biopsy ( Acuderm; Ft . Lauderdale , FL ) specimens were bisected and fixed in formalin ( 10% ) and embedded in paraffin . Hematoxylin and eosin ( H&E ) and immunoperoxidase labeling was performed on paraffin sections ( 4 μm ) by the Tissue Procurement & Histology Core Laboratory and by the Histopathology Laboratory at UCLA , according to guidelines for clinical samples . Detection of Gr-1 ( Ly-6G ) -positive cells and IL-1β expression on paraffin embedded specimens of lesional skin punch biopsy specimens were performed with a biotinylated rat anti-mouse Ly6G mAb ( clone 1A8 ) ( 1 μg/ml ) ( BD Pharmingen , San Diego , CA ) or rat anti-mouse IL-1β mAb ( clone H153 ) ( 1 mg/ml ) ( Santa Cruz Biotechnology; Dallas , TX ) by the immunoperoxidase method . Myeloperoxidase activity was used to assess neutrophil accumulation in the skin tissue using a previously reported method[67] . Briefly , 8mm punch biopsy samples were weighed and homogenized on ice in 0 . 01 mol/L KH2PO4 at a ratio of 1 volume tissue to 15 volumes of buffer . After centrifugation at 10 , 000g for 20 min at 4° , the pellets were resuspended by sonication in cetyltrimethylammonium bromide buffer ( 13 . 7 mM cetyltrimethylammonium bromide ( Sigma ) , 50 mM KH2PO4 ( Sigma ) , 50 mmol/L acetic acid; pH 6 . 0 ( Sigma ) at a ratio of 1 to 5 weight to volume . The supernatant was kept for ELISA analysis ( see previous text ) . The suspension was centrifuged again at 10 , 000g for 15 min , and the pellet was discarded . The supernatant was then incubated in a 60°C water bath for 2 h . Myeloperoxidase activity of the supernatant was measured by the H2O2-dependent oxidation of tetramethylbenzidine . Absorbance was determined at 650 nm and compared with a linear standard curve of recombinant mouse MPO ( R&D Systems , Minneapolis , MN ) . For detection of cytokines from skin , supernatants from the first step of the MPO assay were used . For the detection of cytokines from cell culture , supernatants from cultured macrophages were used . For the detection of IFN-β , the Mouse IFN-β Verikine ELISA kit ( PBL Assay Science , Piscataway , NJ ) was used according to manufacturer instructions . For detection of IL-1β , a Cytometric Bead Assay ( BD Biosciences , Piscataway , NJ ) was performed per manufacturer instructions .
|
Individual pathogen associated molecular patterns ( PAMPs ) induce gene expression in immune cells through distinct signaling pathways to protect cells from infection . However , pathogens typically possess many PAMPs , and the precise contribution of each PAMP to the gene expression program elicited by a live pathogen has not been clearly defined . Herein , we used gene expression profiling to examine the full early response of macrophages to Staphylococcus aureus , a major human opportunistic pathogen . Surprisingly , we found that two pathogen-sensing pathways , Toll-like receptor ( TLR ) and Stimulator of Interferon Signaling Gene ( STING ) pathways , contribute to the activation of ~95% of the genes induced by S . aureus infection . The remaining genes may be induced by hypoxia pathways . When the bacterium is dead , 98% of the gene induction occurs through TLR signaling , and neither STING nor hypoxia contributes greatly to the response . STING activation requires sensing of S . aureus DNA by the cytosolic DNA sensor , cGAS . During S . aureus skin infection , the TLR and STING pathways compete with each other to induce or suppress host defense , respectively , by counter regulating interleukin 1β production and neutrophil recruitment . A similar approach may allow delineation of the relative contributions of immune pathways in the response to various live pathogens .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"staphylococcus",
"aureus",
"immune",
"receptor",
"signaling",
"membrane",
"receptor",
"signaling",
"hypoxia",
"bacteria",
"neutrophils",
"bacterial",
"pathogens",
"immune",
"system",
"proteins",
"white",
"blood",
"cells",
"animal",
"cells",
"staphylococcus",
"medical",
"microbiology",
"proteins",
"gene",
"expression",
"microbial",
"pathogens",
"toll-like",
"receptors",
"biochemistry",
"signal",
"transduction",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immune",
"receptors",
"macrophages",
"cell",
"signaling",
"organisms"
] |
2017
|
Opposing roles of Toll-like receptor and cytosolic DNA-STING signaling pathways for Staphylococcus aureus cutaneous host defense
|
Cancer is a disease of cellular regulation , often initiated by genetic mutation within cells , and leading to a heterogeneous cell population within tissues . In the competition for nutrients and growth space within the tumors the phenotype of each cell determines its success . Selection in this process is imposed by both the microenvironment ( neighboring cells , extracellular matrix , and diffusing substances ) , and the whole of the organism through for example the blood supply . In this view , the development of tumor cells is in close interaction with their increasingly changing environment: the more cells can change , the more their environment will change . Furthermore , instabilities are also introduced on the organism level: blood supply can be blocked by increased tissue pressure or the tortuosity of the tumor-neovascular vessels . This coupling between cell , microenvironment , and organism results in behavior that is hard to predict . Here we introduce a cell-based computational model to study the effect of blood flow obstruction on the micro-evolution of cells within a cancerous tissue . We demonstrate that stages of tumor development emerge naturally , without the need for sequential mutation of specific genes . Secondly , we show that instabilities in blood supply can impact the overall development of tumors and lead to the extinction of the dominant aggressive phenotype , showing a clear distinction between the fitness at the cell level and survival of the population . This provides new insights into potential side effects of recent tumor vasculature normalization approaches .
Cancer is a disease of multicellular regulation , in which malfunctioning cells can break free of homeostatic regulations imposed by the host environment [1] . One of the main characteristics of cancer is the increased proliferation and mutation of cancerous cells due to malfunctioning control of growth and proliferation [1] . As these behavioral changes typically originate from mutations in the cells’ genetic material , excessively proliferating cells accumulate further alterations , leading to a possible amplification of malignancies . Traditional studies of altered cell traits primarily focus on genetic mutations , but neglect the multicellular nature and genetic variety of tumors . Tumor heterogeneity has been demonstrated experimentally and is an active field of research [2–4] . Neutral mutations may accumulate and contribute to intratumor heterogeneity [5] . An intermediate level of heterogeneity is correlated with low survival probability [6] . Heterogeneity may even promote the collapse of tumor development by inducing a clone population that supports and enhances the growth of other clones in mice [7] . Marusyk and colleagues [7] claim that these supporter clones may be outcompeted by the more aggressive subpopulation , leading to the disappearance of the supporters and to the consequent collapse of the tumor . Heterogeneity questions the validity of previous whole–tumor analyses , as “the most abundant cell type might not necessarily predict the properties of mixed populations” [8] , and emphasizes the need for more detailed approach . Initial phases of tumor development are increasingly thought to give rise to a Darwinian process [1 , 9] , where individual cells compete for growth space and nutrients . Selection is imposed by the microenvironment , a highly complex entity spanning several ranges in size from the endocrine regulation of the whole body down to the extracellular matrix ( ECM ) and neighboring cells . During cancer development this environment is changed due to changes in the cellular component , and due to the tissue and organism level reactions to the tumor . Intrinsic coupling between neighboring cells forms the basis of the plasticity-reciprocity model [10]: as cells alter their behavior ( plasticity ) , their contribution to the local environment changes through for example ECM remodeling , nutrient uptake , or adhesion molecule expression . In turn , the changed environment imposes an altered selection on the cells ( reciprocity ) , creating a feedback between cells and their local microenvironment . This cascading change in behavior is reminiscent of the behavioral changes associated with stages of cancer development [11] that was later described by the accumulation of mutations through which cells become increasingly malignant [12] . Although a strict sequence of mutations was not found , a general pattern was observed in the majority of cases [12] , linking the macroscopic stages to the cell-level changes . Computational modeling is an excellent tool for exploring , studying , and understanding such complex systems , because it provides complete control over assumptions and reveal the consequent behavior emerging from them . This allows the dissection of complex interactions and exploration of experimentally challenging cases . A variety of models have been applied to study cancer . Using a population level description , Basanta and co-workers studied how the combination of tumor treatments , p53 cancer vaccine and chemotherapy , can be optimized to yield the best results [13] . In a similar study , Sreemati Datta and coworkers [14] model tumor development with the inclusion of evolving mutation rates and show that the balance between inducing driver mutations and mutation rates plays a key role in tumor growth: at high mutation rates , genetic instability may counter tumor progression . Combined with close experimental verification , Marusyk and colleagues [7] used a similar model to suggest that interactions among clones may lead to an overall collapse of tumor development . This approach is able to incorporate evolutionary games to help cope with the development of treatment resistance . Such space-free models assume that all cells within the tumor are able to interact with all other cells directly , and are unable to deal with intra-tumor spatial heterogeneities . Studies incorporating this spatial aspect have mostly worked with cellular automata ( CA ) models . For example , Gerlee and colleagues were able to explain the ‘go or grow’ hypothesis through the emergence of haptotaxis in their CA model [15] . In a recent study , Waclaw and colleagues [16] used a CA model to show that cell motility together with cell turnover may prevent intratumor heterogeneity . Of particular interest is the study of Anderson and colleagues [17] , where the evolution of a growing cell population and the effect of a heterogeneous environment is explored . They represent the tumor environment as a distribution of ECM molecules that together with oxygen serve as nutrient after degradation . Cell evolution is modeled in the phenotype by selecting a set of cellular parameters ( matrix degradation rate , proliferation rate , etc . ) from a number of predefined phenotypes upon cell division . The new phenotype was either selected randomly or according to a predefined sequence progressing towards more aggressive behavior . The authors found that cells evolved into a similar , aggressive phenotype when applying the random mutation scheme . Heterogeneity in the population was found to give rise to irregular tumor surface , whereas environmental heterogeneity ( heterogeneity in the ECM distribution ) reduced population heterogeneity and favored the most aggressive cell types . Low concentration of oxygen was also found to reduce population heterogeneity and promote invasive finger formation; applying two bursts of oxygen in these simulations led to the segregation of the mixed population . Similar models have been used to show emergent progression of phenotypes related to hypoxia , glycolysis , and acid-resistance , by including neuronal networks in cells [18] , or angiogenesis by introducing blood vessels in hypoxic regions [19] . Enderling and co-workers [20] introduced the cancer stem cell ( CSC ) hypothesis in a similar model by incorporating cells with unlimited proliferation potential ( CSCs ) and cells with limited proliferation potential . They found that cell death induced by tumor therapy could lead to a more aggressive , proliferative tumor , as the CSCs were no longer competing for space after the treatment . Using a combination model of phenotype evolution and the CSC hypothesis , Sottoriva and colleagues showed that the presence of CSCs in the model tumors led to more invasive tumor morphology [21] . The CA model further allowed exploring efficient drug delivery through the neovasculature [22 , 23] , the role of spatial arrangement of the vasculature in radiation therapy [24] and even the effect of the different 2D representations of the 3D vasculature [25] . Using a more detailed model allows to explore the evolution of many other aspects of cell behavior , such as cell flexibility , cell adhesion , or cell shape . One such model used in tumor modeling is the cellular Potts model [26 , 27] . It has been used to describe , for example , the effect of nutrient limitation on tumor growth morphology [28] , or to compare the emergence of distinct developmental stages in terms of morphology and growth in vascular versus avascular tumors [29] . Using a heterogeneous evolutionary model of CSCs , a recent study reported the emergence of spatial stratification of tumors based on the evolution of adhesion molecule expression [30] . A further line of models explore the effect of oxygenation of tumors [31] in the light of chemotherapy [32] . An alternative way of introducing more detail is a spatial continuum representation of the cells . One example of this is the phase field method which has been used in combination with discrete modeling elements to investigate tumor growth morphology and its interaction with the vasculature [33 , 34] . Most of the above models applies a sharp distinction between the tumor and the healthy tissue . While the tumor is described in a cell-based detail , the environment is usually presented as a continuum . These models implicitly assume that stromal cells surrounding the tumor do not participate in the development of the tumor , and that these cells are fundamentally different from the cancerous tissue . The transition between the stromal and cancerous cell states is neglected . A recent exception from this is the work of Powathil and colleagues [35] where the effect of irradiation on a small set of “healthy” bystander cells is investigated . As a result of irradiation and induced signals , stromal cells may apoptose or may be converted to the tumor cell type in the model . However , in this and almost all of the above models the phenotypes available for evolving cells are restricted to a small , discrete subset of all possibilities . A continuous range of cell states could reflect a more biological picture including more subtle , for example epigenomic , changes . Furthermore , the nutrients are typically modeled as a single component , however , one signature of cancerous cells is the reduced oxygen consumption and increased glucose uptake . For this a more detailed nutrient description is required . Another less explored point of interest is the change in the temporal behavior of the larger environment , the nutrient supply . Initial tumors experience a stable blood supply in healthy tissues , that presents them with a relatively constant environment . In contrast , angiogenic tumors at a later stage develop neovasculature that is tortuous and leaky , presenting a fluctuating , unstable environment for the cells [36 , 37] . How does this temporal variation affect the population behavior ? Would the same aggressive phenotype dominate the population as in the stable environment , or would it fit less than the less aggressive and stable phenotype ? Here we present a cellular Potts model of a closely packed , mutating cell population representing an epithelial tissue within an organism ( Fig 1 ) . Mutation in the model allows cellular behaviors to vary continuously in a wide range of phenotype space , therefore evolution is governed by selection emerging naturally from the limitations of space , glucose , and oxygen . Cellular metabolism is modeled as a mixture of aerobic glycolysis and respiration . Small scale changes in the microenvironment are represented by the changes of the immediate cellular neighborhood , both by mutations and cell rearrangements . We model large scale environmental fluctuations through the fluctuating activity of spatially fixed nutrient sources , which represent the cross-section of blood vessels . Constructed in this way our model includes both the plasticity-reciprocity model of Friedl and Alexander [10] , as well as the large scale fluctuations imposed by the host . In the following sections we analyze the behavior of the model with and without mutating cells . We show that our model reproduces the Warburg shift , and exhibits stages of development similar to those observed in previous studies ( such as [17–19] ) . In our model cells undergo clonal expansion , hypoxia , followed by starvation , with the development of segregated populations around blood vessels . The spatial differentiation of cell populations is somewhat similar to the spatial diversity in real tumors as described by Alfarouk et al . [38] . Whereas Alfarouk and colleagues describe two main habitat zones concentrically surrounding the blood vessel , we observe only one of the zones with high proliferation rates and a robust cellular outflow from near the nutrient source . Finally , our results indicate that the dominant aggressive phenotype is more sensitive to fluctuations in the environment than the ones maintaining a stable phenotype without mutation .
To investigate the above questions , we model a monolayer of cells using a modified cellular Potts model ( CPM ) based on the CompuCell3D implementation [39] which can be obtained from http://www . compucell3D . org . Customized code for the simulations and example parameter and initial condition files can be found in S1 File . In the following we give an overview of the model; for more detail see the Methods section . Cells in the CPM are represented as confluent domains on a lattice on which an integer σ ( x → ) at every position x → indicates which cell is occupying the location x →; cell-free areas are designated by σ ( x → ) = 0 . Cell movement results from a series of elementary steps in which an attempt is made to copy σ ( x → ) at a randomly selected location x → to one of its randomly selected neighboring location x ′ → . This attempt is accepted with a probability based on a Hamiltonian goal function H that defines cell dynamics ( Eqs 1 and 2 ) . H is usually defined such that cells maintain a controlled size , perform amoeboid-like cell movement , and may exhibit adhesion or contact-repulsion . A time step in the model is defined as the Monte Carlo Step ( MCS ) consisting of N elementary steps where N is the total number of lattice sites in the model . In our model we apply the usual calibration by relating 1 MCS to 1 minute real time , and 1 lattice site to 2 μm . Diffusion of soluble substances are simulated on a lattice identical to the cellular-lattice . This calibration relates the simulated tissue area to 400μm × 400μm , and diffusion coefficients of simulated nutrients to realistic values ( glucose and lactate: Dg = 10−9 m2/s; oxygen: D O 2 = D l = 10 - 11 m 2 / s ) [40] . We implemented a metabolism whereby cells consume glucose and oxygen from their environment and use a mixture of lactic acid fermentation and cellular respiration . Cells metabolize oxygen and glucose to generate an abstract cellular energy that is used for their maintenance and growth . The amount of energy required for cells is controlled by the expression levels of glucose transporters in the cell membrane which is determined by an intracellular growth signal parameter ( N0 ( i , t ) for cell i at time t ) . The mode of metabolism is determined by an internal hypoxia inducible factor ( h ( i , t ) for cell i at time t ) , that is controlled through the oxygen levels inside and around the cell and the amount of intracellular reactive oxygen species ( ROS ) ( Eq 15 ) . This factor determines the ratio of respiratory and fermentative modes of cellular metabolism . In our model the speed of energy production , and hence cell growth , is independent of the mode of metabolism in order to avoid a selection bias towards the faster metabolic mode . If a cell reaches a pre-defined doubling size it divides , hence cell cycle time is determined by cell metabolism . Cell death may occur either due to starvation or age . If a cell generates less energy from metabolism than is required for maintenance , it converts the necessary amount of its cell mass into energy ( catabolism ) . Once the cell mass is exhausted , the cell is considered dead and is taken out of the simulation . Cells are also killed in the simulation with a 0 . 1% probability after each MCS to maintain cell turnover . Glucose and oxygen are supplied by a separate set of designated immobilized cells that play the role of blood capillaries ( Fig 1a ) . To allow the temporal control of nutrient supply , capillaries can be in an active or a blocked state . Nutrient levels are kept at a fixed concentration in the blood stream , that is inside the capillaries , when the capillaries are active . When a capillary is blocked , nutrient levels are kept at zero within the capillaries . The activity of the vessels are changed with a probability ( nutrient switching probability ) after each MCS . A high switching probability leads to a stable supply while a low switching probability results in an inconsistent supply , mimicking blocked or tortuous vessels without affecting the average activity time of the vessels . Lactate produced by cells as a waste from lactic acid fermentation is cleared out of the system by active capillary cells . For more detail on the model see the Methods section . We start by first verifying that the model is capable of simulating a sustainable homeostatic tissue . Therefore we studied the behavior of a healthy tissue in the absence of mutation and stable nutrient supplies . Fig 2a shows the model setup . We consider a two-dimensional square lattice , corresponding with a slice of tissue of 400 μm × 400 μm ( 40 × 40 cells , 200 × 200 lattice sites ) , containing four blood vessels arranged in a square formation ( Fig 1a ) . To achieve constant nutrient supply , the probability of a vessel to be blocked or unblocked in every MCS is 0 . 5 . In this case each vessel switches between active ( depicted in white ) and blocked ( depicted in gray ) states rapidly . As the nutrients diffuse away from the source , this rapid switching results in a continuous supply of nutrients ( Fig 2c ) . The color of tissue cells in Fig 2a indicates the intracellular pressure , defined as the difference between target volume ( biomass ) and actual volume of the cell ( π ( i , t ) = VT ( i , t ) − V ( i , t ) ) . This measure differs from the extracellular pressure as it includes any contribution of the contractile actin cortex surrounding biological cells . Note that the pressure within the population is distributed without any specific pattern . Fig 2b shows the dynamics of the tissue for 105 MCS corresponding to approximately 70 days . The number of cells fluctuates around a constant value throughout the simulation ( Fig 2b ) , and is sufficient to cover the whole system: the ratio of the cell-covered region over the total simulation area is close to one ( Fig 2b ) . As the average intracellular pressure does not increase over the course of the simulation ( Fig 2b ) , cell growth and proliferation are kept in balance with the basal metabolism and the constant cell turnover . Excessive growth is prevented by the lack of growth space through a negative feedback between the intracellular pressure and cell growth ( Eq 18 ) . Nutrient levels remain constant during the simulations and cells remain respiratory as indicated by the absence of lactate ( Fig 2d ) . Next we investigated how a cancerous tissue would behave in our model . Cancerous tissues are characterized by large number of mutations , chromosomal rearrangements and changes in gene expression level , all of which may result in phenotypic changes of the cells [1 , 41–43] . To mimic such changes , we allowed the set of 10 assigned phenotypic properties of cancerous cells to change upon division ( Fig 1b ) , including the division volume or adhesion parameters ( see Table 1 and Methods ) . After cell division , the daughter cells inherit the phenotypes of their parents with some small mutations . Every parameter is allowed to change with a fixed probability ( mutation rate μp for parameter p ) and independently of one another . The change in parameter p is drawn from a normally distributed random variable with a standard deviation of σp , that is: p ′ = p + N ( 0 , σ p ) . The parameters are allowed to change freely within a pre-defined range ( Table 1 ) with reflective boundary conditions . This allows an unbiased parameter to uniformly explore the available range ( Fig 1b inset ) . To determine how a single mutating cell would perturb the homeostasis of the above tissue , we inserted a cell with a mutation potential either near ( Fig 3a , red cell ) or far from ( Fig 3b ) the nutrient source . The single mutating cell persisted in 28 and 26 simulations out of 100 for the two different initiation positions . When the cell persisted , it expanded within the first 5000 MCSs and eventually colonized the population completely ( Fig 3a and 3b ) . In order to study the internal dynamics of the tumor , we will only focus on the case where the mutating cell persists and has colonized the population; therefore we will initiate our simulations with populations where all cells are allowed to mutate . Fig 3c–3f shows the behavior of the model with mutating cell populations , with 10% mutation rate . In comparison with the non-mutating populations , the intracellular pressure is higher ( Fig 3c ) . This shows that cells overcome the initial growth control mechanism . The number of cells initially increases , reaches a peak , and then declines to approximately half of the peak value , well below the initial numbers , where the population size stabilizes ( Fig 3d ) . These changes in cell numbers are followed by the intracellular pressure as well: in the expansion phase the pressure increases , but before the peak in cell number it sharply declines . After the population size settles to a lower value , the pressure settles to an approximately constant value . The full coverage of the tissue drops to approximately half coverage , showing that the cancerous cells cannot maintain a complete monolayer . Nutrients are depleted further from the blood vessels , resulting in a shortage of oxygen in the distant regions ( Fig 3e ) , shortly followed by the depletion of glucose ( Fig 3f ) . The depletion of oxygen triggers the cells to switch to fermentation , resulting in an increase in lactate . This switch accelerates the depletion of glucose , causing the decline in population size . After the population size is reduced , oxygen levels return to the same level as in the non-mutating populations , while cells still rely on fermentation as can be seen from the maintained lactate levels ( Fig 3f ) . Previous studies have reported distinct stages of development including hypoxia , glycolysis , or acid-resistance [17–19] . However , in these studies the evolution occurred in isolation from the stromal tissues and vasculature either using a limited set of phenotypic behaviors potentially constraining the degree of freedom of the evolutionary trajectories or with immobile cells . Therefore , we first asked if stages of tumor progression also occurred in our less constrained model . Fig 4 shows the behavior of our model with the nutrient concentrations and cell numbers averaged from 10 independent simulations with stable vasculature ( switching probability = 0 . 5 ) and 10% mutation rate . Based on this , we identified distinct stages in our model: expansion ( 1 ) , hypoxia ( 2 ) , starvation ( 3 ) . Insets show cell configurations characteristic of the three stages , color scale on the insets indicates the oxygen concentrations . These stages emerge as a result of an interplay between the cells and their environment , as in the proposed plasticity-reciprocity hypothesis of Friedl and Alexander [10] . Stages in our model relate to: 1 . Conditioning the environment; 2 . A reaction to the environmental change in the behavior of cells ( new phenotypes emerging , old phenotypes disappearing ) ; 3 . New environment created by the new population . Remarkably , this shows that despite the much larger number and freedom of mutating parameters in our model , we still find the same phenomena of emergent stages . In the first stage of our model the population expands by cell growth and division . A high intracellular growth signal N0 is selected for in the population , favoring fast growing cells ( Fig 4c ) . This parameter evolves much faster than any other of the 10 mutating parameters ( see also Table 1 ) while most of the other parameters do not exhibit such a strong and clear drift by the end of the simulations ( Fig 4d ) . Indeed , the overall behavior of the model did not change qualitatively in simulations where only N0 or N0 and the chemotactic parameters are allowed to evolve ( S1 Fig ) . Since nutrients in the environment are not limiting due to the assumed prior homeostasis , these cells simply outgrow the slower ones , creating patches of high growth ( Fig 4e ) . This leads to the expansion of fast growers , and as a result , cells from newer generations appear in clumps of growth hot-spots ( Fig 4f ) . At this stage expansion can occur at any position in the population since nutrients are available at any location . In the second stage of our model the population turns hypoxic . As the number of cells grows , the intracellular pressure increases rapidly in the tightly packed tissue until about t = 5000 MCS ( see Fig 3d ) . At this time oxygen is depleted at areas further away from the source ( see Fig 4a middle inset at t = 6000 MCS ) . Fast growing cells ( high N0 , Fig 4g ) are unable to fuel their increased metabolic need through oxidative respiration and turn hypoxic ( Fig 4h ) . These cells further increase their glucose uptake ( Fig 4i ) due to the HIF1-α→ GLUT signaling pathway in our model ( Eq 15 ) , and start the production of lactate . Finally , glucose is gradually depleted at regions far from the sources as a result of the elevated glucose consumption rate of fast growing cells . In the depleted areas cells die out , and with them the cell population is gradually decreased ( Fig 3b and 3d ) . The only cells remaining are around the vessels , that eventually hijack the source ( Fig 4j and 4k ) . These cells continue to compete as in stage 1 , since the change in the environment near the vessels is minimal , and keep increasing their internal growth signal from generation to generation ( Fig 4j ) . Increasing intracellular pressure near vessels ( Fig 4k ) exerted by neighboring cells and counteracts growth . Changes in nutrient levels indicate that our model selects for cells exhibiting the well-known Warburg effect , whereby cells metabolize glucose through glycolysis even in the presence of oxygen ( aerobic glycolysis ) [44] . Cells in our model initially shift to aerobic glycolysis to support their metabolic need escalated through competition , shown by the increasing levels on intracellular hypoxia and ROS ( Fig 4l ) . This results in an increase in extracellular oxygen ( Fig 4a ) . Despite the availability of oxygen , cells are unable to revert to a more efficient full respiratory metabolism due to production of ROS which stabilizes HIF1−α and limits the amount of metabolic flux through respiration ( Eq 15 ) , thus keeping it in a state of hypoxia in our model ( Fig 4l ) . Nevertheless , cells do consume oxygen but it is significantly lower than glucose uptake ( Fig 4m ) . Taken together , these results show that our model exhibits different stages of development similar to previously published studies . Remarkably , this progression emerges in spite of an almost completely unrestricted evolution of a large number of phenotypic parameters . Tumors in this model are initialized at random positions , but due to the explicit representation of localized nutrient sources , we show that they occupy the vicinity of blood vessels at later stages . This is enhanced by the more realistic representation of cells in the CPM where cell shape and compressibility allow cell rearrangements within the packed tissue as opposed to the more rigid CA models exploring progression [17–19] . Secondly , we show that our model selects for cells exhibiting the Warburg effect despite the lack of growth advantage of fermenting cells . To test if the stages of tumor progression depend on the phenotypic mutation rates , we simulated the model for a series of mutation rates . Whereas the non-mutating population keeps a constant size , all mutating populations exhibit an initial increase in cell numbers ( Fig 5a ) . In highly mutating populations ( 5% and 10% ) this increase is followed by a drop in cell numbers . This drop is observed later in populations with 5% mutation rate , and population decrease is just starting at the end of the simulations in populations with 1% mutation rate . Note that the repetitions reproduce the behavior fairly well , suggesting the robustness of the system . Therefore we suggest that similar stages occur at lower mutation rates , and the time needed for reaching each stage depends on the mutation rate . This is supported by the changes in nutrient levels in the simulations , which react faster to change than the total number of cells ( Fig 5b–5f ) . In healthy , non-mutating populations , nutrient levels and population size is stabilized ( Fig 2b ) . In mutating populations , cells evolve a higher metabolic demand in parallel with increased proliferation . This results in an increase in population size and decrease in oxygen levels , followed by a decrease in glucose levels . As oxygen becomes sparse , cells turn hypoxic and switch from oxidative phosphorylation to aerobic glycolysis , resulting in an increase in lactate levels . We have found the same behavior in populations with different mutational probabilities ranging from 0 . 1% up to 10% ( Fig 5b–5f ) , or higher ( S2a–S2h Fig ) . Again , a lower mutational probability only delayed the changes in the nutrient levels . We did not find a qualitative difference in the progression in our simulations at different mutation rates , showing that the emergent order of stages is robust in this system . To explore the structure of the population at different mutation rates in the system , we analyzed the distribution of cells in the space of normalized mutating parameters . After subtraction of the initial parameter values and normalizing with mutational step-size , the ten-dimensional parameter space of the population was reduced to the three most prominently changing axes within each population using principal component analysis ( see Methods ) . Fig 5g shows one example population at the final time point of the simulations for mutation rates 1% , 10% , and 20% with each dot representing a cell . At low mutation rate ( 1% ) the population splits up into well-defined clones which are more spread and less well-defined at higher mutation rates . The first three principal axes contain most of the information about the shape of the population , as can be seen by the normalized weights ( eigenvalues ) of these components ( Fig 5h ) . The composition of principal axes at the final time point of the simulations shown on Fig 5g differ ( Fig 5i ) with the growth signal N0 , cell rigidity λV , and glucose chemotaxis χg playing an important role at μ = 1% . At 10% mutation rate the lactate chemotaxis parameter χl plays a distinct role in segregating the population phenotypes , while at μ = 20% the segregation is less obvious ( Fig 5g ) and is driven mainly by parameters λV , doubling volume VD , and adhesions ( ρCAM , ρMAM ) . While the composition of the main axes varies across different simulation repeats with the same mutation rate , the populations are nevertheless well characterized by the first three components in all cases ( S2i and S2j Fig ) . To better understand population structure in the simulations , we categorized the cells at each time point in the 10-D phenotype space using hierarchical clustering ( Methods ) . We measured the displacement of each cluster as the Euclidean distance between the point of origin and its center of mass and its spread as the mean distance of points of the cluster from the cluster’s center of mass . As expected , we observed that populations with higher mutation rates reach further from the origin by the end of simulations; these clusters are more spread and less dense than clusters in populations with lower mutation rates ( Fig 5j and 5k ) . Considering the whole time course of the simulation , the cluster analysis reveals that as the population explores the phenotype space , clusters from the highly mutating populations first tend to spread out and dilute more than the clusters in populations from lower mutation rates ( Fig 5l and 5m ) . These results show that the population starts to dilute much faster in phenotype space at high mutation rates , but without affecting the progression of stages apparent from the nutrient levels . Next , we tested how feedback from a larger spatial organizational level , through the nutrient supply in our case , would affect populations of different mutation rates . In a healthy tissue , nutrient supply is relatively constant . The main source of fluctuations are the slow daily change according to the circadian rhythm , and the relatively fast blood pulse . In cancerous tissues the vasculature is remodeled through tumor vasculogenesis , resulting in tortuous and leaky vessels [36 , 37] . As these vessels are less reliable , here we assume that they dysfunction from time to time , for example by becoming temporarily blocked . To model vessel tortuosity , we introduced a blocking probability for the vessels . An open blood vessel will be blocked with a probability P at every time step , and a blocked vessel will be opened with the same probability . A high blocking probability ( P = 0 . 5 ) corresponds to a healthy situation , where the resulting fast switching of the vessel is smoothened out by nutrient diffusion . A lower blocking probability introduces longer periods of nutrient deprivation but also longer periods of nutrient supply . On average these systems receive the same amount of nutrients , but in different dosage . Healthy , non-mutating populations in our simulations survived over a wide range of nutrient fluctuations . The average number of cells from 10 simulation repeats showed that these populations keep roughly the same size at different blocking probabilities , as shown on Fig 6a . At the extreme blocking probability P = 0 . 001 , only 2 out of 10 populations died out . In comparison , mutating cell populations are unable to tolerate blocking probabilities lower than P = 0 . 01 , irrespective of their mutation rate ( Fig 6a ) . Note that at high P , populations with low mutation rates have an increased population size at the end of the simulations ( t = 105 MCS ) , compared to the healthy population . This increase results from the initial stage of progression , as these populations only reach the first stage ( expansion ) by the end of the simulations . However , this advantage disappears as P decreases . The observed reduction in population size due to decreasing P could work in two ways: either by killing cells through starvation , or by speeding up the progression of stages . In the previous section we showed that a higher mutation rate speeds up the progression of the population and thus results in a reduced population size ( Fig 5 ) . This might eventually lead to extinction . If the populations under fluctuating nutrient supply go through the same stages of progression as the ones in the stable environment , the environmental indicators used in Fig 5 ( levels of glucose , oxygen , lactate ) are unsuitable , as these might change in simulations with different P . Instead , we focus on the intracellular evolution of traits , that are not altered directly in these experiments . The first trait to be selected for is the intracellular growth signal of the cells ( N0 ( i , t ) ) that exhibits a run-away dynamics ( Fig 4c ) . If the blocking probability accelerates the progression through the stages , it should increase the selection pressure on the intracellular growth signal as well . Contrary to this expectation , we found that the trend in the average value of the intracellular growth signal remains approximately the same in simulations across different P values , and even slightly decreases at lower P ( Fig 6b ) . Similarly , other cellular measures ( such as hypoxia , ROS , or pressure ) , or cellular parameters ( such as the chemotaxis parameters ) show the same behaviors irrespective of P ( S3 Fig ) . Therefore , longer nutrient fluctuations do not accelerate the evolution of the population , and the reduction in population size is not a result of the acceleration of the same evolutionary dynamics . Instead we conclude that as cells deplete the nutrients in the environment due to their increased consumption , the chance for survival in systems with longer fluctuations is reduced . Next we altered P during the time of simulation runs . We tested how the population reacts if the blood vessels become increasingly tortuous , starting from a healthy state ( fast switching ) progressing to a tortuous vasculature ( slow switching ) . Blocking probability in these simulations is decreased gradually in the simulations , following a geometric progression P ( t + 1 ) = r P ( t ) with an initial value of P ( t = 0 ) = 0 . 5 and ratio of r = 0 . 999876 . Once the progression reaches P ( t f ) = 0 . 001 at tf ≈ 50 , 100 MCS the blocking probability is not decreased further ( P ( t > t f ) = P ( t f ) ) . Similar to the stable system , the mutating populations ( mutation rate μ = 0 . 1 ) are initially driven into the high consumer state by cell-cell competition ( Fig 6c solid red line showing growth signal parameter N0 ) . When the fluctuation probability reaches the order of P ( t ) = 0 . 01 ( t ≈ 31 , 500 MCS ) , the populations start to die out ( Fig 6c red dashed line showing number of surviving populations ) , similar to the case of static low blocking probabilities . Note however , that non-mutating populations ( Fig 6c blue ) are able to survive increasing fluctuations in the nutrient supply . This shows that a changing nutrient supply does not necessarily influence the direct competition among cells . Inconsistency of nutrients may emerge from the dysfunctional tissues of the emergent tumor occluding vasculature . To represent this feedback , we examined how the population behaves when the consistency of nutrient supply is related to the amount of cellular coverage in the tissue . We created a feedback between the density of the tissue ( measured as tissue surface coverage ρ , with 0 ≤ ρ ≤ 1 ) and the fluctuating source . For a fully populated tissue we kept the nutrient switching probability high , P = 0 . 5 , providing a smooth nutrient supply , and decreased it linearly with the cell density to model the variability in nutrient supply . Thus: P ( ρ ) = 0 . 499 ρ + 0 . 001 . In these simulations the mutating population persists at approximately the same level as in the healthy case ( Fig 6d ) . The emergent tissue coverage yields an approximate fluctuation probability of P ( ρ ) ≈ 0 . 27 , sufficiently high to support mutating populations . While the relationship between nutrient supply stability and living cell density is experimentally unclear , our results suggest that a linear relationship between stability and density is insufficient to drive the aggressive cells to extinction , similar to what is expected of an expanding pathological tumor . The effect of low nutrient levels with two consecutive surges of high oxygen levels has been shown to affect selection in a model of evolutionary tumor growth [17] . Here we used localized nutrient sources and stochastic supply , rather than two deterministic and uniform pulses . We show that in our system , inconsistent nutrient supply does not increase the speed of cellular evolution , however , it does reduce the viability of the unstable cell populations . We found similar results when inconsistency is considered progressively increasing or proportional to the tissue coverage . Our results suggest that , while consistent nutrient supply promotes cell-level selection of fast growing cells , inconsistent nutrient supplies that put a larger demand on the tissue exert selection at the tissue scale and provide higher chances of survival for populations with cellular quiescence . In a simulated healthy tissue devoid of mutants , cell growth is independent of spatial localization , growth is limited intrinsically by a constant growth signal balancing spatial confinement . After this intrinsic limitation is lifted through uncontrolled mutations and competition , growth becomes clustered in growth hot-spots ( Fig 4e ) . These spots are populated by overly proliferative cells , resulting in clonal expansion: Fig 7a shows configurations at two time points in the same simulation color-coded for descendants . At a later stage , when cells deplete nutrients , proximity to the sources dictates the growth rate ( Fig 4j ) . In the resulting environment cells closer to the source grow faster and divide , while cells further away starve and die . This differential growth gives rise to a directed cell movement from the vicinity of the sources to the depleted areas , apparent from the short cell trajectories shown in Fig 7b . To demonstrate that the segregation patterns were indeed linked to the nutrient sources and are not the result of finite system size , we performed simulations with randomly scattered blood vessels instead of the regular distribution , similar to previous studies [24 , 31] . The example trajectory plot on Fig 7c shows that the outflow of cells is correlated with the nutrient positions . The population becomes segregated , as the probability of a cell invading a neighboring region is diminished . As a result of this differential growth cells near the source take over the vicinity of the vessel and spread their phenotype in this region ( Fig 7d ) . While cells within each segregated part of the population have highly similar phenotype , the different parts evolve independently of one another . This is shown by the distribution of phenotypic traits over time in each population ( Fig 7e ) . This gives rise to independently evolving quasi-species-like populations within the tissue , and might be analogous to the observed heterogeneity in tumors . In simulations with low but constant vessel blocking probabilities cells around a blocked source are able to leave their region and migrate to another active source , as exemplified by the trajectories in Fig 7e . The ability to detect and to move to a neighboring active source is crucial for the cells for survival , and therefore is expected to be selected for . In our model , cells can achieve this by evolving chemotaxis towards glucose or oxygen , or chemotaxis away from lactate . Indeed , chemotaxis parameters are the second most affected parameters throughout the evolution of the population ( Fig 4c and 4d ) . Interestingly , chemorepulsion by lactate is strongly selected for although lactate in our model does not have any direct metabolic effect on the cells . This acquired property helps to orient the cells towards the nutrient sources better than oxygen , which becomes ubiquitous with more shallow gradients , or glucose , which does not diffuse far from the nutrient sources due to elevated uptake . The sum effect of these motility parameters can be expressed with a combined chemotaxis parameter defined as: χ ′ ( i , t ) = χ g ( i , t ) + χ O 2 ( i , t ) - χ l ( i , t ) . Fig 7g and 7h show the evolution of this combined chemotaxis parameter in populations from different vessel blocking probabilities and mutation rate of 10% . Indeed in all of the conditions the combined chemotaxis is selected for . In simulations with lower blocking probabilities the combined chemotaxis has increasingly higher fluctuations due to the random selection introduced by the blocking and opening of the vessels , often leading to extinction as well .
We show that a directional evolution emerges from random movement in phenotype space as the result of cell competition , driving cells from a healthy ( homeostatic ) state to a more aggressively expansive phenotype . This is consistent with previous findings of Anderson and colleagues [17] , who showed that a similar drift towards aggressive phenotypes emerges if cells are allowed to mutate randomly into one of a 100 predefined discreet phenotypes . In contrast to these abrupt changes in behavior , our model only allows small changes in mutation . This choice lends more persistence to the clones in the population , since more mutation events are required to diverge from an existing clone . Faster growing cells are selected in our model , which then go on to colonize the population by means of clonal expansion . Progression of phenotypes has been observed previously in other models of tumor evolution , where the authors also considered the toxic effect of acidification due to glycolysis [18 , 19] . In [18] , cells first develop the ability to survive in hypoxic environments by lowering their apoptotic response threshold to low oxygen levels . This is followed by a metabolic switch to glycolysis and finally the emergence of the acid-resistant phenotype . In contrast , acid-resistance was proposed to emerge first followed by evolution of glycolysis in another model where nutrient sources were represented as point sources [19] . The distinct stages exhibited in our model ( clonal expansion , oxygen depletion , and starvation ) are accompanied by a sudden increase in glycolytic activity which is then moderated onto a stable medium level ( Fig 4l ) . In our model the population turns hypoxic ( Fig 4l ) before developing a distinct chemo-repulsive response ( Fig 4c ) . We found the progression robust against changing the mutation rate , however , at higher mutation rates the population was able to explore the phenotype space faster and in more dispersed clusters ( Fig 5 ) . Although lactate in our model does not affect cells , cells gradually develop a chemorepulsion away from lactate ( a negative χl on Fig 4c and 4d ) . Lactate accumulates at regions where high consumers deplete glucose , therefore cells use lactate chemorepulsion as a compass to navigate away from the high-consumer niche towards a more supportive environment . This novel feature in our model highlights how phenotypes that show no apparent advantage at the cell level could be selected for based on the altered micro-environmental conditions . Local expansion in tumor growth models has recently been described by Waclaw and colleagues [16] . Their study focused on heterogeneity in passenger mutations while the probabilistically occurring driver mutations set the proliferation advantage . Cell motility there acts to blur intratumor heterogeneity by allowing proliferating cells to invade ‘cell-free’ areas where further proliferation is not inhibited by other cancer cells . While this model is able to reproduce the clonal expansion features of tumors , it neglects the potential inhibitory effect of spatial constraint produced by the surrounding stromal tissue . Proliferation diversity in tumors was shown to be essential to explain experimentally observed tumor morphology by an earlier study using the cancer stem cell hypothesis [20] which interprets the tumor as a conglomerate of self-metastases . The emergent clonal expansion patterns in our model ( Fig 7a ) are reminiscent of these self-metastases and therefore suggest that slight differences of proliferation rates within the population are sufficient to generate these patterns , as opposed to the sharp distinction between cancer stem cells and non-stem cells . Previous models of tumor evolution typically neglect the confining effect of the surrounding healthy tissue although spatial confinement can play an important role when studying the effects of treatment recovery in a tumor with a cancer stem cell population [21] . When a large portion of cells is killed by therapy , the internal cells have access to growth space and are able to regrow the tumor . In other words: space limitation keeps the growth inhibited . To mimic this situation here we restrict the growing tumor to a confined space , making them a model of an in vivo system . We note that simulations in our system do not develop a necrotic core as in other models , as dead cells simply shrink and disappear from the simulations without inducing any signaling response from neighboring cells or without increasing spatial confinement . Therefore our system focuses on live tumor cell evolution where necrotic cells may be considered as extruded from the simulated monolayer region into the underlying necrotic core , removed by the immune response and/or drained by the lymphatics . The source of nutrients in previous models is typically considered uniform , or is provided in the environment in a dispersed fashion . The source of nutrients are the blood vessels in our model , similar to the more recent studies of Shirinifard and colleagues [29] , Powathil and coworkers [31 , 32 , 35] , or Patel and colleagues using a hybrid cellular automaton model [45] . Two recent studies showed that arrangement of vessels and their 2D representation affects tissue oxygenation and may alter the outcome of a simulated radiotherapy [24 , 25] . Nutrient diffusion and utilization create a spatial gradient in the supplies which is translated into a differential growth pattern in our system . Differential growth creates a collective flow of cells outwards from the sources ( Fig 7b , 7c and 7f ) . This outward flow is counter-acted by the emergent chemotactic tendency of cells to move closer to the source ( Fig 4c and 4d ) . Chemotaxis in a changing environment provides additional advantage in locating active nutrient sources ( Fig 7f ) . Blood vessels in our model are immobilized entities , vascular remodeling is not included . Inconsistency in nutrient supply is implemented as stochastic switching of blood vessel activity , but without feedback from the pressure or hypoxia in the tissue as in the phase field model of Yan and colleagues [34] . Simulations where the blocking probability P is a function of time ( Fig 6c ) or tissue coverage ( Fig 6d ) showed that the progression of stages is not altered in our system . A higher number of connected loops have been shown to emerge in 3D models of tumor angiogenesis than in 2D , predicting that complete blockage of circulation is very rare [22 , 23] . Here we studied local vessel blockage which would still have a local effect on our system , even if at a reduced frequency . Previously Anderson and colleagues showed that surges of oxygen induce diversification into a population of simulated cells evolving under low oxygen levels [17] . A repeated , second surge was shown to induce phenotypic segregation in these simulations [17] . Here we studied how inconsistency of the nutrient supply affects the population using stochastic ( rather than deterministic ) switching . In contrast to Anderson’s study , cells do not receive a constant low oxygen supply in our model and are therefore prone to extinction . We found that populations with increasing mutation rate are increasingly sensitive to nutrient fluctuations ( Fig 6a ) , although the evolutionary trend at the cell level remained largely unaffected ( Fig 6b , S3 Fig ) . In our model the aggressive phenotype analogous to cancer is a natural consequence of selection on the cell-level and random mutation . When selection pressure is applied on the tissue level ( in the form of fluctuating nutrient supplies ) the overall fitness ( or survival ) of the cancerous population proves to be lower than that of the healthy population , potentially due to depletion of ambient nutrient resources as a form of competition . Cancerous cells create an insecure environment that is more sensitive to stress coming from outside the cell population . This theoretical exploratory study opens questions in how to best approach normalization of the tumor vasculature . In the model , healthy cells are able to withstand the irregularities in nutrient supply better than cancerous cells . Based on this observation , it is tempting to speculate that irregularities of the tortuous tumor vasculature might serve a similar role in real tumor development . It is important to bear in mind that these results are based on a rudimentary model of tumor development . In addition to the simplifications discussed above , cell-cycle regulation is overly simplified in our model as opposed to the study of Powathil and colleagues [31] where it is in the focus of the study; cells in our model lack an explicit way to store surplus energy as opposed to the cumulative health-factor of Swat and co-workers [30] representing the cells’ tolerance against starvation . Due to its 2D nature , our model is unable to account for the out of plane transport of nutrients or cells . However , using our quasi-2D model allows the study of several processes taking place in epithelial tissues , where most of the tumors arise . Importantly , the implemented cellular metabolism is overly simplified to enable the exploration of the system . After the foundation of this model framework is set out , it could be expanded with more detailed intracellular metabolic networks , for example , using spatial dynamic flux-balance analysis [46] . These multiscale models will lead to a tighter integration of computational and experimental work similar to the “symbiotic” approach described in a recent angiogenic sprouting study of Cruys and colleagues [47] . Further future work includes the more thorough exploration of different mutation rates for different traits . One of the ultimate goals of computational systems biology is constructing a virtual tissue in order to predict efficient treatment of various diseases . This can be achieved by adding homeostatic mechanisms to the tissues such as contact inhibition of growth , or introducing an internal energy storage as in the study of Swat and colleagues [30] . Inclusion of a dynamic angiogenesis model ( e . g . [22 , 29 , 48–50] ) might involve exploring the roles of different blood vessel placements , or implementing angiogenesis models already available in the same platform . Finally , the system enables the measurement of fitness at different spatial levels and in different ( micro- and macro- ) environments which could serve as a basis for exploring evolutionary trade-offs in using computational simulations . In this study we introduce an unbiased evolutionary approach to studying the evolution of interacting tissue cells . The model includes localized source of nutrients ( oxygen , glucose ) and sink for intermediate metabolites ( represented by lactate in our model ) , and a simplified cellular metabolism including glycolysis and respiration . Cells in the model are spatially confined and no explicit distinction is made between stromal or tumor cells . The model exhibits distinct stages of development with an emergent evolutionary drift towards rapid growth , high glucose consumption , and hypoxia . This is accompanied by a Warburg-effect , whereby cells become unable to return to respiratory metabolism even at high oxygen levels . The simulated tumor exhibits clonal expansion and eventually gives rise to similar phenotypes around each nutrient source which then evolve independently later on . Finally , we found that the emergent rapid growing population is highly sensitive to intermittent nutrient supply , such as caused by leaky tortuous blood vessels .
Cells are initialized in a monolayer with an initial volume and target volume of 25 lattice sites on a lattice of 200 × 200 and four endothelial cells ( Fig 1a ) . In the initial regime of the simulations , nutrients are allowed to diffuse into the system to obtain a natural distribution resulting in high glucose and oxygen and low lactate levels ( Fig 1b ) . During this equilibrating time cells are allowed to metabolize , but cannot grow , shrink , move , or mutate . This state represents a stable , homeostatic tissue with sufficient nutrient availability . When the temporal changes in diffusing nutrients is less than 5% in a time interval of 100 MCS , the initial regime is closed , cells are released and the simulation is started . Nutrient fields at the beginning of the initial regime ( Fig 1b ) are initialized using pre-generated concentration distributions to expedite equilibration . These initial concentration fields are generated by simulating a cell population in the initial regime starting with zero concentrations: ∀ x → ∈ Λ : g ( x → , t = 0 ) = 0 , O 2 ( x → , t = 0 ) = 0 , l ( x → , t = 0 ) = 0 . The concentration fields are saved when the total concentration levels remain within 5% over a 100 MCS iteration period: ∑ x → ( s ( x → , t = t save ) - s ( x → , t = t save - 100 MCS ) ) < 0 . 05 ∑ x → s ( x → , t = t save - 100 MCS ) for s ∈ {g , O2} . Note that no lactate is present as all cells are respiratory in the initial regime . All mutating parameters in all cells are initialized with the same value , after which all cells undergo a mutation attempt to provide an initial heterogeneity to the population . Adhesion parameters are set to J ( c , c ) = J ( c , EC ) = Jmax , and J ( c , m ) = Jmax/2 for the three region ( “cell” ) types: stromal cells ( c ) , endothelial cells ( EC ) , and cell-free areas ( m ) . To allow the full range of interactions for cells , we fix the adhesion molecule density values of the extracellular region and the endothelial cells as: k CAM , CAM × ρ CAM ( i ∣ τ ( i ) : EC ) = J ( c , EC ) = J max k MAM , MAS × ρ MAS ( i ∣ τ ( i ) : m ) = J ( c , m ) = J max / 2 ( 20 ) With this choice the effective Jeff values are allowed to change between 0 and Jmax for Jeff ( c , c ) , and Jeff ( c , EC ) . For interaction with the medium , Jeff ( c , m ) is allowed to change between 0 and Jmax/2 . Diffusion parameters for glucose , oxygen , and lactate were set to Dg = 10−9 m2/s and D O 2 = D l = 10 - 11 m 2 / s following the approximation of Jiang and coworkers [40] , and decay is neglected for all of these chemical species ( ωs = 0 , s ∈ {g , O2 , l} ) . Decay rate of ROS is set to a constant ωζ = 0 . 1 per MCS . The amount of ATP produced from 1 glucose molecule through respiration was chosen as αr = 38 . This approximation is a theoretical upper limit for the process , in reality this number is expected to be lower . However , due to the compensation with the number of glucose transporters ( see Eq 13 ) the exact value of this parameter is not expected to change the behavior of the system . To analyze the population behavior over the simulations we analyzed the evolving parameters in the following way . For every parameter p ( i , t ) of cell i at time t we calculated a normalized parameter as pn ( i , t ) = [p ( i , t ) − p ( i , t = 0 ) ]/σp where σp is the characteristic step size of the parameter . This way the normalized parameters reflect their distance from their origin in terms of mutational step-size . The distribution of cell phenotypes in this space was analyzed at time t by finding the principal components of the 10-dimensional set of pn ( i , t ) points for all i and p using singular value decomposition from scientific python ( SciPy ) . The axes corresponding to the first three largest eigenvalues were selected as the main components of the cloud . Clustering in the normalized phenotype space was performed using hierarchical clustering of SciPy with the Ward method and Euclidean distances . To distinguish clusters we established a cutoff cophenetic distance of 100 manually by evaluating a set of selected dendograms and distribution of clusters plotted on the first three main axes . Displacement of the clusters is calculated as the ( 10D ) Euclidean distance of the center of mass of the cluster from its initial point of origin . Spread of clusters was calculated as the mean distance of points in the cluster from its center of mass . Density of clusters was calculated by dividing the spread by the number of points in the cluster .
|
Multicellular organisms control their cells to facilitate higher level function of the whole organism . In tumors , this control is lost and cells are allowed to enhance their fitness by , for example , increased proliferation . Tumor cells continue to change their behavior through accumulating mutations , leading to a complex and highly heterogeneous structure . Several computational studies have investigated the emergent structures of such mutating group of cells and led to the recognition that cellular heterogeneity within tumors is essential to explain the observed morphologies . Most of these studies have considered a limited number of possible cell phenotypes , an isolated tumor cell population , unlimited growth space around the tumor , often with an inexhaustible source of nutrients . Here we introduce a modeling approach that takes into account the limited growth space around the tumor , localized nutrient sources , cellular metabolism , and mutation in a continuous phenotype space . The model reproduces the Warburg effect due to the limited nutrient supply leading to an irreversible switch in cellular metabolism , and , consistently with previous models , exhibits stages of development together with a natural selection for a rapidly growing phenotype . This phenotype locally emerges in stable environments , but when nutrient supply becomes erratic , these show less resilience and are outcompeted by slow growers .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"cell",
"physiology",
"carbohydrate",
"metabolism",
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"cardiovascular",
"anatomy",
"oxygen",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"carbohydrates",
"cell",
"metabolism",
"organic",
"compounds",
"glucose",
"glucose",
"metabolism",
"oxygen",
"metabolism",
"blood",
"vessels",
"chemistry",
"chemotaxis",
"biochemistry",
"cell",
"biology",
"organic",
"chemistry",
"anatomy",
"monosaccharides",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"metabolism",
"chemical",
"elements"
] |
2017
|
Blood vessel tortuosity selects against evolution of aggressive tumor cells in confined tissue environments: A modeling approach
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.